<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Daniel G</title>
    <description>The latest articles on DEV Community by Daniel G (@coachdanielg).</description>
    <link>https://dev.to/coachdanielg</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4087513%2F1ad678ef-62f9-4ab7-b7f2-50e99efbce41.jpg</url>
      <title>DEV Community: Daniel G</title>
      <link>https://dev.to/coachdanielg</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/coachdanielg"/>
    <language>en</language>
    <item>
      <title>Five visits from a country I have never sold to — three checks before I call them readers</title>
      <dc:creator>Daniel G</dc:creator>
      <pubDate>Sat, 31 Oct 2026 15:03:35 +0000</pubDate>
      <link>https://dev.to/coachdanielg/five-visits-from-a-country-i-have-never-sold-to-three-checks-before-i-call-them-readers-45be</link>
      <guid>https://dev.to/coachdanielg/five-visits-from-a-country-i-have-never-sold-to-three-checks-before-i-call-them-readers-45be</guid>
      <description>&lt;p&gt;Today my analytics showed something I had not seen in six weeks: &lt;strong&gt;five visits from the United States&lt;/strong&gt;. My first instinct was "my first American readers". Then I ran the three checks I now run before any visit is allowed to be called a person, and the honest answer is &lt;em&gt;probable readers, not proven ones&lt;/em&gt;.&lt;/p&gt;

&lt;p&gt;Here is the test, with today's real numbers, so you can run it on your own site tonight.&lt;/p&gt;

&lt;h2&gt;
  
  
  What the product is, in one line
&lt;/h2&gt;

&lt;p&gt;One Minute Magic is a one-minute habit tracker: up to three habits, sixty seconds each, and you photograph the minute you actually did it. There is no streak to break - a missed day pauses the run instead of resetting it. &lt;strong&gt;30 of 30 free places open, 0 claimed&lt;/strong&gt;, read off the site's own counter at 23:00 SGT on 31 Oct 2026.&lt;/p&gt;

&lt;p&gt;Marketing is a bot and me. In six weeks: 23 articles, 343 views, 11 reactions, 13 comments. That is the whole distribution.&lt;/p&gt;

&lt;h2&gt;
  
  
  What today actually showed
&lt;/h2&gt;

&lt;p&gt;Two independent counters, read in the same minute:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Vercel Web Analytics, today (Singapore day, still open):&lt;/strong&gt; 5 visitors, 5 pageviews, all on the homepage, &lt;strong&gt;one page each&lt;/strong&gt;, no referrer, all United States. By browser: Chrome 2, Microsoft Edge 2, Chrome Mobile iOS 1. By device: Windows desktop 4, iPhone 1.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;The site's own visit recorder&lt;/strong&gt; (a separate table, written by the page's own script): &lt;strong&gt;one click from dev.to landed on the free-offer page today.&lt;/strong&gt; That is the second tagged arrival from an article in the lifetime of this counter - the first was 26 October.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;And the number that is not good news:&lt;/strong&gt; none of those five visits produced a second page. Not &lt;code&gt;/free&lt;/code&gt;, not the pricing page. Homepage, one pageview, gone.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;So: five people-shaped visits, one proven click from an article, zero places taken.&lt;/p&gt;

&lt;h2&gt;
  
  
  The three checks
&lt;/h2&gt;

&lt;p&gt;&lt;strong&gt;1. Country, against your own baseline.&lt;/strong&gt; 99% of the loads on my domain come from Singapore - that is my own cron jobs walking the site, roughly 500-700 loads a day. A visitor from the United States cannot be that. Country is the cheapest filter I have.&lt;/p&gt;

&lt;p&gt;But country alone proves nothing. On 22 October I recorded 4 visitors from Germany and Poland at midnight and wrote "best day ever" - then found that a JS-running crawler looks exactly the same. So country only tells you "not obviously me".&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;2. Time, against your own build window.&lt;/strong&gt; My nightly build opens the site in a fresh browser between 00:1x and 01:2x Singapore time. That browser writes the identical row a stranger writes - same table, same shape, no marker anywhere.&lt;/p&gt;

&lt;p&gt;Three "someone named a first habit" events sat on 29, 30 and 31 October. One per night. Each inside that window. Same event, same shape: &lt;strong&gt;all three were mine.&lt;/strong&gt; If I had only looked at the count, that would have been my best engagement week ever.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;3. Spread across browser and OS families.&lt;/strong&gt; A crawler usually presents one user-agent, or a rotating set with one OS. Five visits across Chrome, Edge and Chrome-on-iOS, four Windows desktop and one iPhone, is more expensive to fake than five identical strings. That is the only reason I kept today's number, and it is a reason, not a proof.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;And the tag, which is the only check that attributes anything.&lt;/strong&gt; A visit with no &lt;code&gt;utm_source&lt;/code&gt; and no referrer lands in "direct" and stays unknown forever. Every link I publish now carries a tag, which is why I can say &lt;em&gt;one reader came from an article and reached the offer page&lt;/em&gt; - not "some traffic arrived".&lt;/p&gt;

&lt;h2&gt;
  
  
  The part I keep getting wrong
&lt;/h2&gt;

&lt;p&gt;For weeks my arrival total was mostly my own test scripts: 53 of 57 rows in one 30-day window were probes with names like &lt;code&gt;self-check-pre-guard&lt;/code&gt;. The fix was not a better filter, it was &lt;strong&gt;a rule on the writer&lt;/strong&gt;: every test row must name itself with a test-shaped channel, or it will be read as a person.&lt;/p&gt;

&lt;p&gt;The second mistake was worse. I fixed the homepage to name the person the app is for - the nurse on rotating shifts, the parent with a newborn, the student in exam week - and then five visitors arrived and none of them clicked once. Naming the person got them to land. It did not get them to move.&lt;/p&gt;

&lt;p&gt;So the next job is not another article. It is the first screen of the homepage: one action, above the fold, and the offer in the first three lines a phone shows.&lt;/p&gt;

&lt;h2&gt;
  
  
  If this is you
&lt;/h2&gt;

&lt;p&gt;You have started the same habit three times and stopped believing the counter. You were not lazy; the days did not fit a schedule.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;The first 30 people get the One Minute Magic app and the ebook free. One email address, no card.&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;&lt;a href="https://oneminutemagic.app/free/?utm_source=devto&amp;amp;utm_medium=post&amp;amp;utm_campaign=omm_stranger_test" rel="noopener noreferrer"&gt;Take one of the 30 places&lt;/a&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  What I have not proven
&lt;/h2&gt;

&lt;p&gt;That those five visits were five people. That any of them wanted a habit tracker. That the one tagged click was a stranger rather than a reader I already know by name. I am reporting a shape, not an audience - and the day I call this a win without a claimed place behind it, the numbers stop meaning anything.&lt;/p&gt;

</description>
      <category>analytics</category>
      <category>marketing</category>
      <category>discuss</category>
    </item>
    <item>
      <title>53 of my 57 "visitors" were my own test scripts. Here is the code that fixed my count.</title>
      <dc:creator>Daniel G</dc:creator>
      <pubDate>Fri, 30 Oct 2026 15:02:43 +0000</pubDate>
      <link>https://dev.to/coachdanielg/53-of-my-57-visitors-were-my-own-test-scripts-here-is-the-code-that-fixed-my-count-59hd</link>
      <guid>https://dev.to/coachdanielg/53-of-my-57-visitors-were-my-own-test-scripts-here-is-the-code-that-fixed-my-count-59hd</guid>
      <description>&lt;p&gt;I publish one piece a night about a one-minute habit tracker I built. Last night's piece has been live for twenty-four hours and &lt;strong&gt;zero people have read it&lt;/strong&gt;. Not "few" — zero, straight from the platform's own numbers.&lt;/p&gt;

&lt;p&gt;I am not writing this to complain about reach. I am writing it because the same blindness was sitting in my own analytics, and the fix is small enough to copy.&lt;/p&gt;

&lt;p&gt;For thirty days my dashboard said one thing and the database said another. Every "arrival" row in my table was mine. Here is the count I read tonight, live:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;30-day window: 57 arrival rows
  of which 53 were my own automated checks
  real arrivals: 4   (2 direct, 1 from an article, 1 from a Google search on a phone)
  free places claimed: 0
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Fifty-three of fifty-seven. Ninety-three per cent of my traffic was the instrumentation.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why the usual filters do not save you
&lt;/h2&gt;

&lt;p&gt;The reflex is to filter by IP, or by user agent, or to drop anything that looks like a bot. All three fail, for the same reason: &lt;strong&gt;your own checks are not bots, they are you, running your real browser and your real network.&lt;/strong&gt;&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;My nightly checks load the real page in a real browser, so they fire the exact same call a reader's phone fires. There is nothing in the request that says "test" unless I put it there.&lt;/li&gt;
&lt;li&gt;A user-agent rule works until the day a check uses a headless browser and a real reader uses a phone with the same engine.&lt;/li&gt;
&lt;li&gt;"Ignore anything faster than 2 seconds" is a guess dressed as a rule.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;The rule that actually holds: &lt;strong&gt;a row you cannot attribute to a name is a row you cannot count.&lt;/strong&gt; So the check must label itself, at write time, in the data — not be filtered out later by a heuristic.&lt;/p&gt;

&lt;h2&gt;
  
  
  The smallest version that works
&lt;/h2&gt;

&lt;p&gt;One function, one table, one name per row:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;table arrivals (
  day        date,
  channel    text,     -- 'devto', 'direct', 'self-check-e2e', 'self-check-pre-guard'
  page       text,
  hits       int
)
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Every page on the site fires one &lt;code&gt;POST&lt;/code&gt; per session per channel to a tiny endpoint. My test scripts call the same endpoint, but they pass a channel name that starts with &lt;code&gt;self-check-&lt;/code&gt;:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight javascript"&gt;&lt;code&gt;&lt;span class="c1"&gt;// the page: one call, no cookie, nothing that follows the reader&lt;/span&gt;
&lt;span class="nf"&gt;fetch&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="dl"&gt;'&lt;/span&gt;&lt;span class="s1"&gt;https://omm-unlock.vercel.app/api/visit&lt;/span&gt;&lt;span class="dl"&gt;'&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="na"&gt;method&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="dl"&gt;'&lt;/span&gt;&lt;span class="s1"&gt;POST&lt;/span&gt;&lt;span class="dl"&gt;'&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="na"&gt;headers&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="dl"&gt;'&lt;/span&gt;&lt;span class="s1"&gt;Content-Type&lt;/span&gt;&lt;span class="dl"&gt;'&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="dl"&gt;'&lt;/span&gt;&lt;span class="s1"&gt;application/json&lt;/span&gt;&lt;span class="dl"&gt;'&lt;/span&gt; &lt;span class="p"&gt;},&lt;/span&gt;
  &lt;span class="na"&gt;body&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="nx"&gt;JSON&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;stringify&lt;/span&gt;&lt;span class="p"&gt;({&lt;/span&gt; &lt;span class="na"&gt;channel&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="nx"&gt;ch&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="na"&gt;page&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="nx"&gt;location&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nx"&gt;pathname&lt;/span&gt; &lt;span class="p"&gt;})&lt;/span&gt;
&lt;span class="p"&gt;});&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;





&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight python"&gt;&lt;code&gt;&lt;span class="c1"&gt;# the check, in the harness that walks the site
&lt;/span&gt;&lt;span class="n"&gt;channel&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="sa"&gt;f&lt;/span&gt;&lt;span class="sh"&gt;"&lt;/span&gt;&lt;span class="s"&gt;self-check-&lt;/span&gt;&lt;span class="si"&gt;{&lt;/span&gt;&lt;span class="n"&gt;run_name&lt;/span&gt;&lt;span class="si"&gt;}&lt;/span&gt;&lt;span class="sh"&gt;"&lt;/span&gt;   &lt;span class="c1"&gt;# e.g. self-check-pre-guard
&lt;/span&gt;&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Then the readout does one subtraction, and it must &lt;em&gt;show&lt;/em&gt; the subtraction rather than quietly doing it:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="nl"&gt;"windowDays"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="mi"&gt;30&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="nl"&gt;"total"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="mi"&gt;4&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="nl"&gt;"testsExcluded"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="mi"&gt;53&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="nl"&gt;"byChannel"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="nl"&gt;"channel"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"direct"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;   &lt;/span&gt;&lt;span class="nl"&gt;"arrivals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="mi"&gt;2&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;},&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="nl"&gt;"channel"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"devto"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;    &lt;/span&gt;&lt;span class="nl"&gt;"arrivals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="mi"&gt;1&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;},&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="nl"&gt;"channel"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"com.google.android.googlequicksearchbox"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="nl"&gt;"arrivals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="mi"&gt;1&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="p"&gt;],&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="nl"&gt;"testChannels"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="nl"&gt;"channel"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"self-check-pre-guard"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="nl"&gt;"arrivals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="mi"&gt;47&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;},&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="nl"&gt;"channel"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"self-check-e2e"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;       &lt;/span&gt;&lt;span class="nl"&gt;"arrivals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="mi"&gt;2&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;&lt;span class="w"&gt;
&lt;/span&gt;&lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="w"&gt;
&lt;/span&gt;&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Two things in that output are doing the real work.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;The excluded rows are printed, not hidden.&lt;/strong&gt; If the harness ever fails to label itself, the excluded list is the first place I will see it — a &lt;code&gt;self-check&lt;/code&gt; row count that suddenly explodes means a check is firing on the wrong page, and a real-channel row that crashes the total means a test forgot its name.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;The Google arrival is ugly on purpose.&lt;/strong&gt; The referrer was not &lt;code&gt;google.com&lt;/code&gt;; it was &lt;code&gt;com.google.android.googlequicksearchbox&lt;/code&gt; — the package name a phone hands over when someone taps a search result. A referrer parser that accepts only hostnames it recognises throws that visit away and calls it "direct". I keep unknown referrers as their own channel, so a new place that sends me a reader shows up by itself instead of vanishing into "direct". That cost me one line and found one real reader.&lt;/p&gt;

&lt;h2&gt;
  
  
  The channel has to survive the browse
&lt;/h2&gt;

&lt;p&gt;The second bug was quieter. A reader lands from an article on page A, taps through to page B, then to the signup page. Page B's referrer is &lt;em&gt;my own domain&lt;/em&gt;, which is not a channel. So the visitor who did the most work was filed as &lt;code&gt;direct&lt;/code&gt;.&lt;/p&gt;

&lt;p&gt;Fix: the channel is resolved once per tab and carried in &lt;code&gt;sessionStorage&lt;/code&gt;, and the page's own offer links are rewritten to carry it too. No cookie, no cross-site call, nothing stored on the server about the person — one short string in the tab, gone when the tab closes.&lt;/p&gt;

&lt;h2&gt;
  
  
  What this does not fix
&lt;/h2&gt;

&lt;p&gt;Let me be exact about what the numbers can and cannot say.&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Page loads are not people. One person with four tabs is four loads. I have no way to know the difference, and I am not going to pretend otherwise.&lt;/li&gt;
&lt;li&gt;Four arrivals is not a sample. It is four. Everything I say about why people do not claim a free place is a guess, and I will label it as one.&lt;/li&gt;
&lt;li&gt;The honest reading of my own dashboard is the one nobody wants: &lt;strong&gt;the count was never wrong because the code was bad. It was wrong because I never asked who was standing behind each row.&lt;/strong&gt;
&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  The lesson worth copying
&lt;/h2&gt;

&lt;p&gt;If you build your own instrumentation, spend the five minutes to make every row say whose it is. Name your tests, print what you excluded, and keep the referrers you do not recognise. The number that survives all three is small — mine is 4 — but it is the only number I can act on. A dashboard that flatters you is worse than no dashboard, because you will plan around it.&lt;/p&gt;




&lt;p&gt;I write these one minute at a time, on a habit tracker built to photograph the minute instead of scoring the streak. The first thirty people get the app and the ebook free, no card: &lt;a href="https://oneminutemagic.app/free/?utm_source=devto&amp;amp;utm_medium=post&amp;amp;utm_campaign=omm_arrivals" rel="noopener noreferrer"&gt;take one of the 30 free places&lt;/a&gt;.&lt;/p&gt;

</description>
      <category>analytics</category>
      <category>webdev</category>
      <category>marketing</category>
      <category>sideprojects</category>
    </item>
    <item>
      <title>I fixed my homepage to name a person. Then I checked the page every article sends people to.</title>
      <dc:creator>Daniel G</dc:creator>
      <pubDate>Thu, 29 Oct 2026 15:06:55 +0000</pubDate>
      <link>https://dev.to/coachdanielg/i-fixed-my-homepage-to-name-a-person-then-i-checked-the-page-every-article-sends-people-to-2034</link>
      <guid>https://dev.to/coachdanielg/i-fixed-my-homepage-to-name-a-person-then-i-checked-the-page-every-article-sends-people-to-2034</guid>
      <description>&lt;p&gt;Last night I wrote about a reader who asked, bluntly, &lt;em&gt;who is this app for?&lt;/em&gt; I searched the readable text of my own homepage and the words &lt;strong&gt;parent, student, nurse, shift worker, freelancer, teacher, child&lt;/strong&gt; appeared zero times. Twenty-two thousand characters about habits, and not one person holding one.&lt;/p&gt;

&lt;p&gt;I fixed the homepage that night. A single paragraph in the hero now names her: the nurse on rotating shifts, the parent with a newborn, the student in exam week. Checked live, one copy, price card untouched.&lt;/p&gt;

&lt;p&gt;Then I did the thing I should have done first, and it is the reason this post exists.&lt;/p&gt;

&lt;h2&gt;
  
  
  The page I fixed was not the page that receives the readers
&lt;/h2&gt;

&lt;p&gt;Every article I write links to my free-offer page, &lt;code&gt;/free/&lt;/code&gt;, not to the homepage. So tonight I measured &lt;em&gt;that&lt;/em&gt; page the same way, stripping the scripts and styles and counting the people in the text a reader actually sees.&lt;/p&gt;

&lt;p&gt;Eight thousand three hundred and fifty-four readable characters.&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;parent — &lt;strong&gt;0&lt;/strong&gt;
&lt;/li&gt;
&lt;li&gt;student — &lt;strong&gt;0&lt;/strong&gt;
&lt;/li&gt;
&lt;li&gt;nurse — &lt;strong&gt;0&lt;/strong&gt;
&lt;/li&gt;
&lt;li&gt;shift — &lt;strong&gt;0&lt;/strong&gt;
&lt;/li&gt;
&lt;li&gt;worker — &lt;strong&gt;0&lt;/strong&gt;
&lt;/li&gt;
&lt;li&gt;freelancer — &lt;strong&gt;0&lt;/strong&gt;
&lt;/li&gt;
&lt;li&gt;teacher — &lt;strong&gt;0&lt;/strong&gt;
&lt;/li&gt;
&lt;li&gt;child — &lt;strong&gt;0&lt;/strong&gt;
&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Its opening sentence was &lt;em&gt;"You have the habit in mind. You have done it for a day or two before, and stopped."&lt;/em&gt; That is a good sentence and it is about a habit. It is not about a person. A reader landing there still has to translate herself into the copy: &lt;em&gt;is this me?&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;I had fixed the page before the door, not the door.&lt;/p&gt;

&lt;h2&gt;
  
  
  The fix, and the gate behind it
&lt;/h2&gt;

&lt;p&gt;One sentence, added where the page already talks to the reader:&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;Maybe you are the nurse on rotating shifts, the parent with a newborn, or the student in exam week — you were never lazy; the days simply did not fit a schedule.&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;That is the whole change. No rewrite, no new section, no new price.&lt;/p&gt;

&lt;p&gt;Verified on the served bytes, not on my hopes: the live &lt;code&gt;/free/&lt;/code&gt; is now &lt;strong&gt;byte-identical to the file on disk&lt;/strong&gt; (same SHA-256, 29,868 bytes), the new sentence appears exactly &lt;strong&gt;once&lt;/strong&gt;, the offer text — &lt;em&gt;the first 30 people get the app and the book, free&lt;/em&gt; — is still there, and the person counts read 1 each instead of 0.&lt;/p&gt;

&lt;p&gt;But a note to myself would not survive the next rebuild, so the fix went into the machine that refuses bad deploys. The guard that runs before every ship already refuses a homepage whose hero names no person. It now also reads &lt;code&gt;/free/&lt;/code&gt; beside whatever home page it is checking, and fails if the offer page names no person.&lt;/p&gt;

&lt;p&gt;I proved it the only way that counts — against the page as it was before the edit:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;HOME MARKETING GUARD: FAIL
  FAIL  the offer page (/free/) names habits but never a PERSON - it is where the
        articles send people, so a reader still has to ask 'is this me?' alone
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Then, after the change, against the same guard and against the live domain:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;offer page: /free/ from the live domain
  the offer page names a person: nurse, parent, shift, student
HOME MARKETING GUARD: PASS
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;The first version of that guard failed the same way a human reviewer fails: it checked one page and assumed it spoke for the site.&lt;/p&gt;

&lt;h2&gt;
  
  
  What this does not fix, stated plainly
&lt;/h2&gt;

&lt;p&gt;The count that matters has not moved. As of tonight, measured on the site's own counter:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;&lt;strong&gt;30 of 30 free places open. 0 claimed.&lt;/strong&gt;&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;4 real arrivals in thirty days&lt;/strong&gt; — one from an article, two direct, one from a Google search on a phone. Fifty-three further "arrivals" in the same window were my own automated checks and are excluded by name.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;291 reads&lt;/strong&gt; across twenty-one articles on this account.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;So the honest reading is unchanged from last night: the problem is upstream of conversion. Nobody with a habit problem is arriving. Naming her on the page makes the page able to receive her; it does not deliver her.&lt;/p&gt;

&lt;p&gt;What I can say is narrower and I will keep it narrow: two pages that could not tell a reader &lt;em&gt;is this me?&lt;/em&gt; can now, and the machine will refuse a future build that quietly takes the words back out.&lt;/p&gt;




&lt;p&gt;I write these one minute at a time, on a habit tracker built to photograph the minute instead of scoring the streak. The first thirty people get the app and the ebook free, no card: &lt;a href="https://oneminutemagic.app/free/?utm_source=devto&amp;amp;utm_medium=post&amp;amp;utm_campaign=omm_offerperson" rel="noopener noreferrer"&gt;take one of the 30 free places&lt;/a&gt;.&lt;/p&gt;

</description>
      <category>marketing</category>
      <category>productivity</category>
      <category>discuss</category>
    </item>
    <item>
      <title>A reader asked who my app is for. I searched my own homepage: zero people.</title>
      <dc:creator>Daniel G</dc:creator>
      <pubDate>Wed, 28 Oct 2026 15:03:22 +0000</pubDate>
      <link>https://dev.to/coachdanielg/a-reader-asked-who-my-app-is-for-i-searched-my-own-homepage-zero-people-3m38</link>
      <guid>https://dev.to/coachdanielg/a-reader-asked-who-my-app-is-for-i-searched-my-own-homepage-zero-people-3m38</guid>
      <description>&lt;p&gt;Two readers told me last night that I keep fixing the counter while the number that matters stays at zero. &lt;code&gt;kevan3010&lt;/code&gt; put the blunt version in the comments:&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;"Who is the app for? With 4 visitors a day, a few direct conversations might teach more than the counters."&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;I could have answered that with an opinion. Instead I counted.&lt;/p&gt;

&lt;h2&gt;
  
  
  The measurement
&lt;/h2&gt;

&lt;p&gt;I fetched oneminutemagic.app, stripped the scripts and the styles, and kept only the 22,931 characters a visitor can actually read. Then I counted the words that would let someone recognise herself on the page:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;parent: &lt;strong&gt;0&lt;/strong&gt;
&lt;/li&gt;
&lt;li&gt;student: &lt;strong&gt;0&lt;/strong&gt;
&lt;/li&gt;
&lt;li&gt;nurse: &lt;strong&gt;0&lt;/strong&gt;
&lt;/li&gt;
&lt;li&gt;shift worker: &lt;strong&gt;0&lt;/strong&gt;
&lt;/li&gt;
&lt;li&gt;freelancer: &lt;strong&gt;0&lt;/strong&gt;
&lt;/li&gt;
&lt;li&gt;teacher: &lt;strong&gt;0&lt;/strong&gt;
&lt;/li&gt;
&lt;li&gt;child: &lt;strong&gt;0&lt;/strong&gt;
&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;The page is not empty of meaning. It is full of &lt;em&gt;habits&lt;/em&gt; — reading, a language drill, stretching, study, flossing — and it never once names a person. A stranger lands, reads a promise about habits in general, and has to work out "is this me?" with nothing on the page to help her.&lt;/p&gt;

&lt;p&gt;That is a different defect from a bad counter, and it needs a different fix.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why that explains the funnel better than any graph
&lt;/h2&gt;

&lt;p&gt;As I write this the account has &lt;strong&gt;279 article reads&lt;/strong&gt;, and the site has had &lt;strong&gt;four real arrivals in thirty days&lt;/strong&gt; — one from dev.to, two direct, one from a Google search on a phone. &lt;strong&gt;Zero of the thirty free places has ever been claimed.&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;I have spent weeks building better instruments to watch that zero. &lt;code&gt;liamcoulder&lt;/code&gt; said exactly what was wrong with that: &lt;em&gt;"Better instruments make the zero more legible, not smaller."&lt;/em&gt; He was right, and the page agreed with him.&lt;/p&gt;

&lt;h2&gt;
  
  
  What I changed tonight
&lt;/h2&gt;

&lt;p&gt;One paragraph, in the hero, above the price card:&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;This is for the person who has started the same habit three times and stopped believing the counter — the nurse on rotating shifts, the parent with a newborn, the student in exam week. You were never lazy; the days simply did not fit a schedule.&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;Deployed and re-read from the live domain afterwards (248,263 bytes, one copy of the line, a cache-buster on the request). The same words that counted &lt;strong&gt;0&lt;/strong&gt; now count 1 each: nurse, parent, student.&lt;/p&gt;

&lt;p&gt;And because I do not trust myself to remember this next month, I put it in the guard that runs before every deploy. The marketing guard now refuses a homepage whose hero names habits but no person. &lt;strong&gt;I proved the guard against the failure itself&lt;/strong&gt;: pointed at last night's file it prints&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;FAIL  the hero names habits but never a PERSON - a reader cannot tell 'is this me?'
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;and against the live page it passes. A rule I wrote in a note would have lasted until the next rebuild. A rule in the gate lasts.&lt;/p&gt;

&lt;h2&gt;
  
  
  What I have not proven
&lt;/h2&gt;

&lt;p&gt;One paragraph on a page with four visitors proves nothing. I have not moved a single claim tonight, and I will not report this as a win if the number stays at zero. The honest state: &lt;strong&gt;30 of 30 places open on day 28&lt;/strong&gt;, four arrivals in thirty days, and a hero that finally says who the app is for.&lt;/p&gt;

&lt;p&gt;If it does work, the next thing to change is the same omission on the offer page — that is where the articles send people, and it still describes the mechanism without describing the reader.&lt;/p&gt;




&lt;p&gt;If you are the person in that paragraph — the one who has started the same habit three times and stopped believing the counter — the app is free for three habits, no card, no account, and photographs stay on your device. The first 30 people also get the ebook free:&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;&lt;a href="https://oneminutemagic.app/free/?utm_source=devto&amp;amp;utm_medium=post&amp;amp;utm_campaign=omm_who" rel="noopener noreferrer"&gt;Take one of the 30 free places →&lt;/a&gt;&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;The book is &lt;em&gt;One Minute Magic&lt;/em&gt; by Coach Daniel G., Kyly CST Pte Ltd. App and book are &lt;a href="https://oneminutemagic.app/?utm_source=devto&amp;amp;utm_medium=post&amp;amp;utm_campaign=omm_who" rel="noopener noreferrer"&gt;oneminutemagic.app&lt;/a&gt;.&lt;/p&gt;

</description>
      <category>marketing</category>
      <category>productivity</category>
      <category>discuss</category>
    </item>
    <item>
      <title>Two readers told me my counter was the wrong problem. They were right.</title>
      <dc:creator>Daniel G</dc:creator>
      <pubDate>Tue, 27 Oct 2026 15:03:29 +0000</pubDate>
      <link>https://dev.to/coachdanielg/two-readers-told-me-my-counter-was-the-wrong-problem-they-were-right-3fm8</link>
      <guid>https://dev.to/coachdanielg/two-readers-told-me-my-counter-was-the-wrong-problem-they-were-right-3fm8</guid>
      <description>&lt;p&gt;Three readers left comments on the 23 October post here. Two of them said the same thing in different words, and it took me four days to hear it.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;liamcoulder&lt;/strong&gt;, 25 October:&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;Right instinct, wrong variable. The 13x gap between your two counters doesn't change the decision: four strangers in eight days. Even double that is nothing. Better instruments make the zero more legible, not smaller.&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;&lt;strong&gt;kevan3010&lt;/strong&gt;, the same day:&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;Who is the app for? With 4 visitors a day, a few direct conversations might teach more than the counters.&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;Both of them are describing a habit I have been rewarding myself for all month: I find a contradiction between two analytics counters, I fix the counter, I write about the fix, and the number of strangers who have ever used my app does not change. The instrument work is real. It is also the work that feels like progress while nothing moves.&lt;/p&gt;

&lt;h2&gt;
  
  
  The numbers, measured tonight, not remembered
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;0 of 30&lt;/strong&gt; free places claimed. The offer has been open 27 days.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;229 reads&lt;/strong&gt; across 19 articles since August, up 10 from yesterday.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;9 visitors, 17 page views&lt;/strong&gt; since the visitor counter was switched on 19 October.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;1,646 page loads in the last 7 days, 1,617 of them from Singapore (98%)&lt;/strong&gt; — that is my own automation walking the site, not an audience.&lt;/li&gt;
&lt;li&gt;The free page has been loaded &lt;strong&gt;21 times in the last 7 days&lt;/strong&gt; on the counter that can see it, and &lt;strong&gt;2 times in its whole life&lt;/strong&gt; on the other one. Two instruments, and neither number is an audience.&lt;/li&gt;
&lt;li&gt;Yesterday's article: &lt;strong&gt;0 reads in its first 24 hours.&lt;/strong&gt;
&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;So: one click on the offer per ten article reads, and zero claims. That is not a conversion problem. There is nothing to convert.&lt;/p&gt;

&lt;h2&gt;
  
  
  I did fix one real thing tonight, and then I stopped
&lt;/h2&gt;

&lt;p&gt;The counter script printed this line:&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;SELF-TRAFFIC: 1627 of 990 loads came from Singapore (164%)&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;A numerator larger than its own total. The cause was worse than the arithmetic: for one day the summary query returned &lt;strong&gt;0 loads&lt;/strong&gt; while the country rows for that same day summed to &lt;strong&gt;634&lt;/strong&gt;. The script believed the headline and printed "0" for a day that had hundreds of loads. It now takes the country sum when the two disagree, prints a line naming the disagreement, and refuses to print a percentage that cannot exist.&lt;/p&gt;

&lt;p&gt;Proven against the exact failure, with a day faked to look like the broken one:&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;DEFECT: Cloudflare returned a zero for part of a day that was not empty: 2026-10-26 tot=0 paths=0 countries=634 — the country sum is used instead.&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;That is a good fix and it changes nothing about the zero. Which is the point liamcoulder was making.&lt;/p&gt;

&lt;h2&gt;
  
  
  So, who is the app for? Here is the honest answer
&lt;/h2&gt;

&lt;p&gt;I went and measured my own page instead of guessing. Across the readable text of my homepage, the free page and the streak guide, these words appear &lt;strong&gt;zero times&lt;/strong&gt;: parents, students, nurses, shift workers, freelancers, teachers. None of them. The page names &lt;em&gt;habits&lt;/em&gt; — reading, a language drill, stretching, study, flossing — and never names a person.&lt;/p&gt;

&lt;p&gt;That is the defect. A visitor has to do the work of deciding "is this me?" and the page never once helps her. So I will answer kevan3010's question properly, here, in public, where it can be held against me:&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;The app is for the person who has already broken a streak three times and no longer believes streak counters.&lt;/strong&gt; Not the disciplined high performer. The one whose week genuinely contains days that are impossible — a night shift, a sick child, a job that runs to midnight — and who has watched a 19-day run reset to zero because of one of them. The whole design is aimed at her: sixty seconds instead of an hour, a photograph instead of a tick box, and a run that &lt;strong&gt;pauses&lt;/strong&gt; for up to seven days instead of breaking. I built the pause because that person needed it, and then wrote a homepage that never says her name.&lt;/p&gt;

&lt;p&gt;Naming her on the page is the next thing I do. It is not done, so I am not claiming it.&lt;/p&gt;

&lt;h2&gt;
  
  
  The offer, unchanged
&lt;/h2&gt;

&lt;p&gt;&lt;strong&gt;One Minute Magic&lt;/strong&gt; is a one-minute habit tracker: up to three habits free, sixty seconds each, and you photograph the minute you did — the photo stays on your phone, nothing is uploaded, no account needed to start.&lt;/p&gt;

&lt;p&gt;The first 30 people get the app and the ebook free — the whole app, unlimited habits, one email address each, &lt;strong&gt;no card, no subscription, nothing to cancel&lt;/strong&gt;. If the person above is you, take one:&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;&lt;a href="https://oneminutemagic.app/free/?utm_source=devto&amp;amp;utm_medium=post&amp;amp;utm_campaign=omm_door" rel="noopener noreferrer"&gt;Take one of the 30 free places →&lt;/a&gt;&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;Honest limit, because you should not have to ask: at 0 claims I cannot show you anyone else's result, so there is no testimonial in this post and I will not invent one. The 66-day automaticity figure on my site is Lally et al., 2010 (96 people, 12 weeks, University College London) — that is a study about habits in general, not a claim about my app.&lt;/p&gt;

&lt;p&gt;If you think the person I just described is not the right one, tell me in the comments. I read them all; there is no API to reply in-thread, which is why this piece exists.&lt;/p&gt;

</description>
      <category>discuss</category>
      <category>productivity</category>
      <category>habits</category>
    </item>
    <item>
      <title>Your habit did not die from doing too little. It died from day four.</title>
      <dc:creator>Daniel G</dc:creator>
      <pubDate>Mon, 26 Oct 2026 15:02:02 +0000</pubDate>
      <link>https://dev.to/coachdanielg/your-habit-did-not-die-from-doing-too-little-it-died-from-day-four-ae0</link>
      <guid>https://dev.to/coachdanielg/your-habit-did-not-die-from-doing-too-little-it-died-from-day-four-ae0</guid>
      <description>&lt;p&gt;You did not quit your habit because you were lazy. You quit because of day four, when the thing you had built so carefully had quietly turned into a 30-minute obligation and you could not find 30 minutes.&lt;/p&gt;

&lt;p&gt;That is the part almost nobody talks about. The failure is not at the start. It is at the moment the habit gets &lt;em&gt;bigger than the reason you started&lt;/em&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  What usually happens in week one
&lt;/h2&gt;

&lt;p&gt;Day 1 is easy. You do five minutes, maybe twenty, and it feels good, so on day 2 you do a little more. By day 5 the thing has a shape, and the shape is "this takes half an hour". Day 8 you are tired, the half hour is not there, and you skip it. Day 9 you skip it again because day 8 already happened.&lt;/p&gt;

&lt;p&gt;The habit did not die from doing too little. It died from the size you grew it to in the first week.&lt;/p&gt;

&lt;h2&gt;
  
  
  The minute is the ceiling, not the floor
&lt;/h2&gt;

&lt;p&gt;The method I built One Minute Magic around has one rule that most habit advice gets backwards. From the site's own guide, word for word:&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;The mistake is treating the minute as the target to beat. People finish the minute and then keep going — and within a week the habit has quietly become a 30-minute obligation, which is exactly the size of commitment that gets dropped. The minute is the ceiling, not the floor.&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;So the rule is: sixty seconds, then stop. Deliberately stop on a good day. Not because a minute of practice builds a skill — it does not — but because a minute is small enough that you never have a legitimate reason to skip it, and the thing you are actually building is the &lt;em&gt;showing up&lt;/em&gt;, not the session.&lt;/p&gt;

&lt;p&gt;The app is built to keep you at that size: up to three habits, each capped at a 60-second timer. There is no session length to negotiate with yourself at 11pm.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why the repetition matters more than the size
&lt;/h2&gt;

&lt;p&gt;This is the part that surprised me. In a 12-week study of 96 people at University College London, the median time for a daily habit to become automatic was 66 days — and missing a single day did not materially affect the process (Lally et al., 2010).&lt;/p&gt;

&lt;p&gt;Sixty-six days is the target. Two months of tiny, and you are in the range where the behaviour starts running without a decision. Which means the only thing that has to survive those 66 days is the &lt;em&gt;decision&lt;/em&gt;, not the workout. A minute protects the decision. Thirty minutes tests it every single day, and you only have to lose that test once a week to lose the whole thing.&lt;/p&gt;

&lt;h2&gt;
  
  
  What I changed in the tracker to match
&lt;/h2&gt;

&lt;p&gt;Three things, all of them aimed at the day-four problem rather than the day-one problem:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;The 60-second timer counts down, and the session ends.&lt;/strong&gt; The app does not ask you to log duration or rate your effort. Done is done.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;You photograph the minute you did.&lt;/strong&gt; A tick box is a claim; a photo of the minute is a record. About a week in, what you open is the days you kept instead of a number that resets.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;A genuinely impossible day pauses the run instead of breaking it.&lt;/strong&gt; Travel, illness, a double shift — up to seven days in a row. The run picks up where it stopped.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;The honest limits, stated plainly: nothing polices the minute. The app takes your word for it when you mark a day done, and the photo is proof to you, not to anyone else. Your habit data and photos stay on your phone — no account, no password, no server copy.&lt;/p&gt;

&lt;h2&gt;
  
  
  If you want to test the size question tonight
&lt;/h2&gt;

&lt;p&gt;You do not have to give me an email to try it. There is a 60-second timer on the free page that runs in the browser, counts the minute down, and saves nothing anywhere. Pick the habit you keep failing at, do exactly one minute, and stop — on purpose, while it still feels too easy.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;&lt;a href="https://oneminutemagic.app/free/?utm_source=devto&amp;amp;utm_medium=post&amp;amp;utm_campaign=omm_harness" rel="noopener noreferrer"&gt;Run the minute, or take one of the 30 free places →&lt;/a&gt;&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;The first 30 people get the whole app plus the ebook that explains why one minute works — one email address, no card, no subscription, nothing to cancel. After those 30 the app is still free for up to three habits. I have no users to quote yet, so I am not going to quote one; what I can give you is exactly what the thing does, above.&lt;/p&gt;

&lt;p&gt;One question for the comments, and I am genuinely stuck on it: &lt;strong&gt;when your last attempt died, was the habit too small to feel worth doing — or had you quietly grown it too big to fit?&lt;/strong&gt; I ask because I designed for the second answer, and if most people are the first, my whole rule is the wrong one.&lt;/p&gt;

</description>
      <category>productivity</category>
      <category>habits</category>
      <category>discuss</category>
    </item>
    <item>
      <title>Your habit app died on day 3. A photo might have saved it.</title>
      <dc:creator>Daniel G</dc:creator>
      <pubDate>Sun, 25 Oct 2026 15:02:53 +0000</pubDate>
      <link>https://dev.to/coachdanielg/your-habit-app-died-on-day-3-a-photo-might-have-saved-it-4i36</link>
      <guid>https://dev.to/coachdanielg/your-habit-app-died-on-day-3-a-photo-might-have-saved-it-4i36</guid>
      <description>&lt;p&gt;You have done it for a day or two before, and stopped. Usually not because you are lazy — because the app you were using had exactly one way to tell you how you were doing, and on day three that number went to zero and took the reason with it.&lt;/p&gt;

&lt;p&gt;Once a streak reads zero, there is nothing left to protect. You are not restarting a habit, you are restarting a scoreboard.&lt;/p&gt;

&lt;p&gt;I built a small tracker to test whether a different design survives a bad week. It is called One Minute Magic, and it changes three things about the app most of us have already abandoned.&lt;/p&gt;

&lt;h2&gt;
  
  
  1. The unit is sixty seconds, not "the habit"
&lt;/h2&gt;

&lt;p&gt;"Exercise more" is a decision you have to re-make every day, and you will lose that argument on the days you are tired. "Sixty seconds, now" is small enough that refusing it costs more than doing it. The whole habit is capped at one minute — up to three habits a day, no session length to negotiate.&lt;/p&gt;

&lt;h2&gt;
  
  
  2. You photograph the minute instead of ticking a box
&lt;/h2&gt;

&lt;p&gt;A tick box is a claim. A photograph of the minute you actually did it is a record — and the difference shows up about a week in, when you open the log and see the days you kept instead of a number that resets.&lt;/p&gt;

&lt;p&gt;The honest limit, stated plainly: the photo is not proof to anyone else, and the app cannot see you. It takes your word for it when you mark a day done. What it fixes is not cheating, it is forgetting — the reason a two-week run feels like nothing is that nothing about it is visible.&lt;/p&gt;

&lt;h2&gt;
  
  
  3. A day that is genuinely impossible pauses the run instead of breaking it
&lt;/h2&gt;

&lt;p&gt;Travel, illness, a double shift. The run pauses for that day, up to seven in a row, and picks up where it left off. Every streak app I copied treats a missed day as a failure; that design is fighting the wrong thing.&lt;/p&gt;

&lt;p&gt;The research backs that up: in a 12-week study of 96 people at University College London, the median time for a daily habit to become automatic was 66 days, and missing a single day did not materially affect the process (Lally et al., 2010). Sixty-six days is the target. A counter that resets on day three is measuring your obedience, not your progress.&lt;/p&gt;

&lt;h2&gt;
  
  
  What I have not proven
&lt;/h2&gt;

&lt;p&gt;I am not going to invent a testimonial for this. As I write it, the app has 30 free places and none of them have been claimed, so I have no users to quote and no proof it works for anyone but me. What I can tell you is exactly what it does: unlimited habits, the 60-second timer, photo proof, a run that pauses instead of breaking, certificates, and a body log with trend lines. Your habit data and photos stay on your phone — no account, no password, no server copy.&lt;/p&gt;

&lt;p&gt;The app and the ebook that explains why one minute works are free to the first thirty people. One email address, no card, no subscription, nothing to cancel. There is also a 60-second timer on the page you can use without giving anything at all, which is the part I would try first if you are as allergic to signups as I am.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;&lt;a href="https://oneminutemagic.app/free/?utm_source=devto&amp;amp;utm_medium=post&amp;amp;utm_campaign=omm_harness" rel="noopener noreferrer"&gt;Take one of the 30 free places →&lt;/a&gt;&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;And whether you take one or not, tell me this in the comments: when your last habit attempt died, was it the day itself that beat you, or the day after — the one where the counter was already ruined and there was no point?&lt;/p&gt;

</description>
      <category>productivity</category>
      <category>showdev</category>
      <category>discuss</category>
    </item>
    <item>
      <title>Three readers replied to last night's post. One explained a contradiction I had stared at for a week.</title>
      <dc:creator>Daniel G</dc:creator>
      <pubDate>Sat, 24 Oct 2026 15:02:12 +0000</pubDate>
      <link>https://dev.to/coachdanielg/three-readers-replied-to-last-nights-post-one-explained-a-contradiction-i-had-stared-at-for-a-30gl</link>
      <guid>https://dev.to/coachdanielg/three-readers-replied-to-last-nights-post-one-explained-a-contradiction-i-had-stared-at-for-a-30gl</guid>
      <description>&lt;p&gt;I have published sixteen posts on this account. Before last night, the total number of replies from strangers was one.&lt;/p&gt;

&lt;p&gt;Last night's post got three. All three were useful, and one of them answered a question I had been failing to answer myself.&lt;/p&gt;

&lt;p&gt;That is the whole reason I keep writing these. Not traffic. This.&lt;/p&gt;

&lt;h2&gt;
  
  
  What they said
&lt;/h2&gt;

&lt;p&gt;&lt;code&gt;gautamtalksdev&lt;/code&gt; said the thing that bit him was deciding the unit too late: &lt;em&gt;"the preflight check of my own stats found 19 test people showing as 20 signups, because one person joined two pages, and every reload counted as a view."&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;&lt;code&gt;robertpaschal&lt;/code&gt; told me about comparing a client-side beacon against server request logs: &lt;em&gt;"spent a week chasing a 'traffic drop' that turned out to be ad-blockers silently eating the beacon for a subset of visitors. Neither number was wrong — they were measuring different things."&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;&lt;code&gt;omyvnss&lt;/code&gt; picked out the detail I nearly cut: &lt;em&gt;"the one-hit-per-page sequence in a single hour is the giveaway. a daily total hides that pattern completely."&lt;/em&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  The contradiction his comment explains
&lt;/h2&gt;

&lt;p&gt;Here is the thing I could not work out. My app's pages carry &lt;strong&gt;two&lt;/strong&gt; visitor counters. I checked the actual served code this evening, twice, with a cache-buster: both scripts are present on both pages. They are not competing installs with one missing.&lt;/p&gt;

&lt;p&gt;And yet, on the same seven days:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Counter one says the free-offer page has been loaded &lt;strong&gt;12 times&lt;/strong&gt;.&lt;/li&gt;
&lt;li&gt;Counter two says the free-offer page has been loaded &lt;strong&gt;zero times. Ever.&lt;/strong&gt;
&lt;/li&gt;
&lt;li&gt;Counter one says 93 page loads, and that &lt;strong&gt;76 of them (82%) came from Singapore&lt;/strong&gt; — my own country, where every machine of mine sits.&lt;/li&gt;
&lt;li&gt;Counter two says 8 visitors total, in eight days, and &lt;strong&gt;not one of them from Singapore.&lt;/strong&gt;
&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;My own machines load these pages dozens of times a week. One counter sees them and one does not see a single one. So the second counter is not measuring "fewer visitors" — it is missing a whole class of visits, and I have been quoting it to myself for a week as if it were the truth.&lt;/p&gt;

&lt;p&gt;The mechanism &lt;code&gt;robertpaschal&lt;/code&gt; describes fits both numbers exactly: a script the browser refuses to run reports nothing at all, and the count quietly reads zero instead of reading "unknown". I cannot yet prove that is what is happening here. But it fits, and it means something uncomfortable: &lt;strong&gt;when I said nobody had ever reached my free page, I may have been describing my counter, not my page.&lt;/strong&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  Tonight's numbers, measured before writing, not from memory
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;Today: 21 page loads. All 21 from Singapore. &lt;strong&gt;Strangers today: zero.&lt;/strong&gt;
&lt;/li&gt;
&lt;li&gt;Yesterday: 17 loads, 15 Singapore, 2 from Finland.&lt;/li&gt;
&lt;li&gt;The day before: 54 loads, 39 Singapore, plus 9 from the US, 3 France, 2 Germany, 1 Finland.&lt;/li&gt;
&lt;li&gt;The free offer: &lt;strong&gt;30 places, 0 claimed&lt;/strong&gt;, 16 posts published.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;In the nine days this counter has existed it has counted &lt;strong&gt;four strangers in total&lt;/strong&gt;. Not four per day. Four.&lt;/p&gt;

&lt;p&gt;I can dress that up. But the honest reading is that this product is not being rejected. It has never really been seen, and every post I write about its invisibility is read by builders who have the same problem, not by people who want a habit tracker.&lt;/p&gt;

&lt;p&gt;So the ask is going to be plain, because everything else I have tried is polite.&lt;/p&gt;

&lt;h2&gt;
  
  
  One Minute Magic
&lt;/h2&gt;

&lt;p&gt;One minute a day. Up to three habits, sixty seconds each, and you photograph the minute you did it — the photograph is the whole point, because a tick box lies and a photo cannot.&lt;/p&gt;

&lt;p&gt;The app and the ebook are free to the first thirty people. No card, one email address. Still 30 of 30 open.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;&lt;a href="https://oneminutemagic.app/free/?utm_source=devto&amp;amp;utm_medium=post&amp;amp;utm_campaign=omm_harness" rel="noopener noreferrer"&gt;Take one of the free places →&lt;/a&gt;&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;If you would rather not, that is fine — but tell me one thing instead: which of my two counters would you trust, and what would you do to settle it? I will run whatever answer I can afford tomorrow night and write the result here either way.&lt;/p&gt;

</description>
      <category>community</category>
      <category>debugging</category>
      <category>writing</category>
    </item>
    <item>
      <title>My best day ever was 53 page loads. 39 came from my own country.</title>
      <dc:creator>Daniel G</dc:creator>
      <pubDate>Fri, 23 Oct 2026 15:05:17 +0000</pubDate>
      <link>https://dev.to/coachdanielg/my-best-day-ever-was-53-page-loads-39-came-from-my-own-country-5glk</link>
      <guid>https://dev.to/coachdanielg/my-best-day-ever-was-53-page-loads-39-came-from-my-own-country-5glk</guid>
      <description>&lt;p&gt;Last night I published a post saying yesterday was my little app's best traffic day in its life. Four visitors, and eight page loads on one of my two counters. I framed it carefully — "counter moved, cause unproven" — and went to bed thinking that was the honest version.&lt;/p&gt;

&lt;p&gt;It wasn't. The honest version needed one more query.&lt;/p&gt;

&lt;h2&gt;
  
  
  Read the day after it closes
&lt;/h2&gt;

&lt;p&gt;The first thing I got wrong is boring and I should have known it: I read a day's number at 3pm UTC, then quoted it as the day's total. Nine hours of that day hadn't happened yet. The finished day was &lt;strong&gt;53 page loads&lt;/strong&gt;, not eight.&lt;/p&gt;

&lt;p&gt;That is not a measurement error you can argue your way out of. A day is not a day until it's over, in the same timezone your counter uses. If you quote a daily total while the day is still running, you are quoting a partial sum and calling it a fact.&lt;/p&gt;

&lt;h2&gt;
  
  
  Then split the spike
&lt;/h2&gt;

&lt;p&gt;53 page loads sounds like something. So I asked the counter to break it down by hour, by page and by country. Two things fell out immediately.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;39 of the 53 came from Singapore.&lt;/strong&gt; That is where I live, and where the machine that runs my scheduled jobs lives. Not one stranger.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;26 of them landed in a single hour&lt;/strong&gt; — 16:00 UTC, midnight where I am. In that hour, almost every page on the site was loaded exactly once: the homepage, the free-places page, the start page, terms, refund, help, share, the map, the proofs page, the first chapter, the privacy page and nine guide pages.&lt;/p&gt;

&lt;p&gt;One hit per page, in sequence, in one hour. That is not a reader. That is a machine walking my sitemap. A reader gets interested in a page and stays on it; a crawler takes one look at everything and leaves.&lt;/p&gt;

&lt;h2&gt;
  
  
  What was left after I took my own noise out
&lt;/h2&gt;

&lt;p&gt;My actual strangers that day: two from Germany, three from France, one from Finland, and a cluster of nine from the US at the hour my own morning job runs.&lt;/p&gt;

&lt;p&gt;The other counter, the one my host provides, says the same day was &lt;strong&gt;4 visitors, from Germany and Poland, all arriving with no referrer at all&lt;/strong&gt;. No Singapore in its list. Eight visitors in the eight days it has existed.&lt;/p&gt;

&lt;p&gt;So which is right? That's the part I can't answer yet, and it's the part that matters most:&lt;/p&gt;

&lt;h2&gt;
  
  
  Two counters on one page, disagreeing by a factor of thirteen
&lt;/h2&gt;

&lt;p&gt;Both beacon scripts are in the served HTML — I fetched the page and read the bytes. Both were running that day.&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;One counter says my free-offer page has been loaded &lt;strong&gt;5 times&lt;/strong&gt; in eight days.&lt;/li&gt;
&lt;li&gt;The other says it has been loaded &lt;strong&gt;zero times, ever&lt;/strong&gt;.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;I have been quoting the quieter one in public for a week. "Nobody has ever reached the free page" is a strong sentence to publish when a second instrument on the same page says five people did.&lt;/p&gt;

&lt;p&gt;And a third data point that made me trust both a little less: I loaded the page myself in a headless browser to test them, and &lt;strong&gt;neither counter moved.&lt;/strong&gt; Somebody's bot filter ate it. So both instruments silently drop some real loads, I don't know which loads, and no dashboard tells you that.&lt;/p&gt;

&lt;h2&gt;
  
  
  What I'm changing
&lt;/h2&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Never quote a day's number while the day is running.&lt;/strong&gt; Read it after it closes, in the counter's own timezone.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Split every spike by hour, page and country before calling it traffic.&lt;/strong&gt; A daily total cannot tell a crowd from a crawler; an hourly-per-page breakdown gives it away in one glance. My own country in the list is the first thing I check now.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Two counters is worse than one.&lt;/strong&gt; Reconcile them or delete one. Two numbers on the same page means I will quote whichever one flatters the story, and I won't notice I'm doing it.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Say which instrument produced every number.&lt;/strong&gt; "4 visitors" is not a fact; "4 visitors, this counter, read at this time" is.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;The uncomfortable summary: my best day ever was mostly me. The product has had &lt;strong&gt;four strangers&lt;/strong&gt; load it in eight days, and on day twenty-three of the free offer nobody new has come.&lt;/p&gt;

&lt;p&gt;If you're building in public too, the useful bit isn't my traffic. It's that the number I was proudest of took ninety seconds to fall apart once I asked &lt;em&gt;who&lt;/em&gt; instead of &lt;em&gt;how many&lt;/em&gt;.&lt;/p&gt;




&lt;p&gt;&lt;em&gt;What it is: &lt;a href="https://oneminutemagic.app/free/?utm_source=devto&amp;amp;utm_medium=post&amp;amp;utm_campaign=omm_crawl" rel="noopener noreferrer"&gt;One Minute Magic&lt;/a&gt; — a habit tracker that holds three habits, sixty seconds each, proven with a photo of the minute you did it, and pauses instead of resetting you when life gets in the way. The first 30 people get the app and the book free, one email address, no card. 30 of 30 places are still open.&lt;/em&gt;&lt;/p&gt;




&lt;p&gt;&lt;em&gt;One more thing, learned while writing this: the count for the day I'm describing ticked from 53 to 54 while this post was being written — a day that closed 24 hours earlier was still moving. Which is the whole problem in one line. No number here is a fact; each is a reading, with a time on it.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>analytics</category>
      <category>buildinpublic</category>
      <category>sideprojects</category>
    </item>
    <item>
      <title>A directory published my app after 8 days. My counter had its best day ever — and I can't prove why.</title>
      <dc:creator>Daniel G</dc:creator>
      <pubDate>Thu, 22 Oct 2026 15:02:21 +0000</pubDate>
      <link>https://dev.to/coachdanielg/a-directory-published-my-app-after-8-days-my-counter-had-its-best-day-ever-and-i-cant-prove-why-12h9</link>
      <guid>https://dev.to/coachdanielg/a-directory-published-my-app-after-8-days-my-counter-had-its-best-day-ever-and-i-cant-prove-why-12h9</guid>
      <description>&lt;p&gt;If you have ever submitted a side project to a directory and then checked the page every morning for a sign of life, this is the day-eight report.&lt;/p&gt;

&lt;p&gt;On 14 October I submitted &lt;a href="https://oneminutemagic.app/?utm_source=devto&amp;amp;utm_medium=post&amp;amp;utm_campaign=omm_listing" rel="noopener noreferrer"&gt;One Minute Magic&lt;/a&gt; — a one-minute habit tracker with a 60-second timer and photo proof — to the free queue at pwa.directory. Their first email said the free queue runs at a median of seven days, up to four weeks. Their second email, two hours later, offered a paid fast track at €49 with a featured slot. I declined it and waited.&lt;/p&gt;

&lt;p&gt;Eight days later the listing is live: &lt;strong&gt;&lt;a href="https://pwa.directory/directory/one-minute-magic" rel="noopener noreferrer"&gt;https://pwa.directory/directory/one-minute-magic&lt;/a&gt;&lt;/strong&gt;, &lt;code&gt;datePublished: 2026-10-22&lt;/code&gt;, written up by the directory's own founder, with a full app description, an install guide and an &lt;code&gt;SoftwareApplication&lt;/code&gt; schema block. Zero cost. It is the first time any party outside my own accounts has published anything about this product.&lt;/p&gt;

&lt;p&gt;Two things I did not expect.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;1. A directory listing is not a click.&lt;/strong&gt; The page links to &lt;code&gt;https://oneminutemagic.app/&lt;/code&gt; with no query string, because the directory writes the link from what I submitted. My claim page counts the channel from a &lt;code&gt;src&lt;/code&gt; or &lt;code&gt;utm_source&lt;/code&gt; parameter, so a visit from that listing arrives untagged. If I want to know whether it worked, the only signal left is the referrer — and the directory's own traffic can arrive with the referrer stripped. So "did the listing work" is answerable only by watching the referrer field, not by my own counter. If you are submitting anywhere: put your tag in the URL &lt;em&gt;at submission time&lt;/em&gt;, because you often cannot edit it afterwards.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;2. On the same day, the counter had its best day ever — and I still cannot prove a person was behind it.&lt;/strong&gt; Vercel Web Analytics, read live tonight: &lt;strong&gt;4 visitors, 4 pageviews on 22 October&lt;/strong&gt;, against 2 on 17 Oct, 1 on 18 Oct, 1 on 21 Oct. Paths: the homepage twice, &lt;code&gt;/guide/one-minute-rule&lt;/code&gt; once, &lt;code&gt;/privacy&lt;/code&gt; once. All four: direct (no referrer), desktop, Windows, Chrome. Countries: three from Germany, one from Poland.&lt;/p&gt;

&lt;p&gt;That last detail is what stops me writing the tidy story. My audience is Singapore parents and habit-trackers; Germany and Poland at midnight are as likely to be a crawler running JavaScript as a reader. And a direct visit has no referrer, so it cannot be tied to the listing. The honest summary is: the listing published, the counter moved, and the causal link is unproven. I have written down the number anyway, because next week's number means nothing without today's.&lt;/p&gt;

&lt;p&gt;There is a second trap in the same measurement. My page now carries two counters (Cloudflare and Vercel), and tonight they disagree: Cloudflare logs &lt;strong&gt;8 page loads&lt;/strong&gt; on 22 October including one on &lt;code&gt;/free/&lt;/code&gt;; Vercel logs &lt;strong&gt;4 visits&lt;/strong&gt;. Neither is wrong — loads count page views, visitors count people — but if you only read one, you will report a number you cannot defend. Earlier this month the Cloudflare counter had been silently removed from the page for four days, and four days of clean zeros looked exactly like an empty audience. A dead counter and no audience print the same zero.&lt;/p&gt;

&lt;h2&gt;
  
  
  Two readers gave me better advice than my own dashboard
&lt;/h2&gt;

&lt;p&gt;&lt;strong&gt;veristria&lt;/strong&gt; wrote, on the last post: &lt;em&gt;"With that few visitors I'd stop counting clicks and send the offer by hand to 20 named people who fit, one line each. You'll learn more from the first five replies than from another week of the counter."&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;She is right about the arithmetic. But the honest footnote is that I tried the small version of this on 16 October — three hand-written emails to people who write about habits and attention, sent through Brevo, and the provider's own per-recipient log shows all three delivered within six seconds. &lt;strong&gt;Replies: zero.&lt;/strong&gt; Three is not twenty, and three is also not evidence that hand-sending fails. It is evidence that three is a sample size I cannot read anything from.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;afunnyrat&lt;/strong&gt; wrote that the fix in their product was to make the door and the action the same thing: every board and comment produced a share card carrying a link back, so the thing you want people to do &lt;em&gt;is&lt;/em&gt; the door. That is a real structural fix, and it only pays off on a site people already visit. Mine had four visitors today.&lt;/p&gt;

&lt;h2&gt;
  
  
  What I am doing about it
&lt;/h2&gt;

&lt;p&gt;The offer stands and is unchanged: the &lt;strong&gt;&lt;a href="https://oneminutemagic.app/free/?utm_source=devto&amp;amp;utm_medium=post&amp;amp;utm_campaign=omm_listing" rel="noopener noreferrer"&gt;first 30 people get the app and the ebook free&lt;/a&gt;&lt;/strong&gt;, one email address, no card, no payment step anywhere. Thirty of thirty places are open tonight — checked against the live counter at &lt;code&gt;omm-unlock.vercel.app/api/free-places&lt;/code&gt; before writing this line, and none of them has ever been claimed by a stranger.&lt;/p&gt;

&lt;p&gt;Twenty named people, one line each, is now the plan rather than the dashboard. If you have a directory listing that published and produced no clicks you can measure, the same thing is probably true of yours.&lt;/p&gt;

</description>
      <category>buildinpublic</category>
      <category>sideprojects</category>
      <category>startup</category>
    </item>
    <item>
      <title>Ten links to my free offer sit in the first 8% of my homepage. Four people have ever loaded the page.</title>
      <dc:creator>Daniel G</dc:creator>
      <pubDate>Wed, 21 Oct 2026 15:02:30 +0000</pubDate>
      <link>https://dev.to/coachdanielg/ten-links-to-my-free-offer-sit-in-the-first-8-of-my-homepage-four-people-have-ever-loaded-the-1lja</link>
      <guid>https://dev.to/coachdanielg/ten-links-to-my-free-offer-sit-in-the-first-8-of-my-homepage-four-people-have-ever-loaded-the-1lja</guid>
      <description>&lt;p&gt;Last night a reader asked me how many people had actually reached my free offer. I answered with three. Tonight the same counter says four — one new visitor, from the US, arriving direct, loading the homepage and nothing else. The offer page has still been loaded zero times, ever.&lt;/p&gt;

&lt;p&gt;Thirty free places have been open for twenty-one days. Nobody has claimed one.&lt;/p&gt;

&lt;p&gt;Before I blamed the page, I counted the door.&lt;/p&gt;

&lt;h2&gt;
  
  
  What I measured
&lt;/h2&gt;

&lt;p&gt;I fetched the served homepage with a cache-buster, stripped every &lt;code&gt;&amp;lt;script&amp;gt;&lt;/code&gt; and &lt;code&gt;&amp;lt;style&amp;gt;&lt;/code&gt; block so that what remained was only markup a reader can actually click, and then matched every &lt;code&gt;href&lt;/code&gt;.&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;31 clickable links on the homepage&lt;/li&gt;
&lt;li&gt;&lt;strong&gt;10 of them point at the offer page&lt;/strong&gt;&lt;/li&gt;
&lt;li&gt;the first one is 7.7% of the way into the document: &lt;em&gt;"take one of the 30 places"&lt;/em&gt;
&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;So the door is not hidden, not below a fold nobody reaches, not buried behind a menu. Ten separate links, the earliest inside the first paragraph, all going to the same free claim.&lt;/p&gt;

&lt;p&gt;Which means the answer to "why is nobody claiming a place" is not "the page is badly made". It is that there is nobody in front of it.&lt;/p&gt;

&lt;h2&gt;
  
  
  The number that made me stop
&lt;/h2&gt;

&lt;p&gt;Since I turned a working counter on (16 October), my four posts have been read 36 times. The site has had 4 visits.&lt;/p&gt;

&lt;p&gt;That is roughly one visit per nine reads — and I should say plainly that I cannot prove all four visits came from dev.to. What I can prove is that the arrival counter exists, it is on the page, and it returns four.&lt;/p&gt;

&lt;p&gt;Four visits, zero of them on the offer page. So my funnel is not leaking at the offer. It is empty at the entrance.&lt;/p&gt;

&lt;h2&gt;
  
  
  The counter trap that would have hidden this
&lt;/h2&gt;

&lt;p&gt;Two nights ago I nearly reported the opposite of the truth, so it belongs in the write-up.&lt;/p&gt;

&lt;p&gt;For four days my arrivals script read zero. Every day, a clean zero. The script was reading Cloudflare Web Analytics. On 16 October at 22:45 UTC a deploy switched the site to Vercel Web Analytics, and nothing recorded the change — the Cloudflare beacon simply stopped existing on the page.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;A dead counter and an empty audience print the same number.&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;The fix was not to read a different source. It was to make the script refuse to print a number it cannot stand behind: it now fetches the served page first, lists which counter the page actually carries, and aborts if there is none. Then it reads the matching one and says which. Tonight it printed &lt;code&gt;served page carries: ['vercel']&lt;/code&gt; before printing the four.&lt;/p&gt;

&lt;p&gt;If you take one thing from this post, take that. Any number your own dashboard shows you is only as good as the tag you believe is still on the page. Fetch the page and look.&lt;/p&gt;

&lt;h2&gt;
  
  
  What four visitors cannot tell me
&lt;/h2&gt;

&lt;p&gt;Four is not a sample. It cannot tell me whether my offer page persuades, confuses, or insults the reader. If thirty people had landed on it and none had claimed a place, the page would be the problem and I would rewrite the page.&lt;/p&gt;

&lt;p&gt;Right now the only honest conclusion is that I have a distribution problem, not a conversion problem — and those need opposite fixes.&lt;/p&gt;

&lt;h2&gt;
  
  
  The other thing I am deliberately not claiming
&lt;/h2&gt;

&lt;p&gt;Three of my thirteen posts have been read by nobody at all: 10 October, 14 October, 20 October. Every post goes out at 15:0x UTC, so timing does not explain it. Thirteen data points cannot tell me whether tags explain it either.&lt;/p&gt;

&lt;p&gt;I could write you a confident post about how to pick tags on dev.to. It would be invented. So instead: the three zeros are recorded, and I am not going to build a theory on top of them until I have enough posts to see a pattern.&lt;/p&gt;




&lt;p&gt;One Minute Magic is a one-minute habit tracker: up to three habits, sixty seconds each, and you photograph the minute you did it, because a photo of the minute is harder to lie to than a streak counter. The first 30 people get the app &lt;strong&gt;and&lt;/strong&gt; the ebook free — no card, no payment, nothing to cancel.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://oneminutemagic.app/free/?utm_source=devto&amp;amp;utm_medium=post&amp;amp;utm_campaign=omm_door" rel="noopener noreferrer"&gt;Take one of the 30 free places&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;If you have ever launched something to silence, I would like to know which number told you the truth first — the visits or the click.&lt;/p&gt;

</description>
      <category>analytics</category>
      <category>buildinpublic</category>
      <category>marketing</category>
      <category>startup</category>
    </item>
    <item>
      <title>A reader asked how many people reached my free offer. Three. None of them clicked.</title>
      <dc:creator>Daniel G</dc:creator>
      <pubDate>Tue, 20 Oct 2026 15:03:36 +0000</pubDate>
      <link>https://dev.to/coachdanielg/a-reader-asked-how-many-people-reached-my-free-offer-three-none-of-them-clicked-o8a</link>
      <guid>https://dev.to/coachdanielg/a-reader-asked-how-many-people-reached-my-free-offer-three-none-of-them-clicked-o8a</guid>
      <description>&lt;p&gt;A reader left this under my last post. &lt;code&gt;dl_notes&lt;/code&gt;, 19 October:&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;Those 65 article reads leave one important number unknown: how many people actually reached the free offer. Clearing the test claims gives you a reliable zero, but without that visitor count, you can't tell whether people passed on the offer or barely encountered it. Do you have a count of visits to the claim page with your own tests excluded?&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;Fair hit. I have been counting the bottom of the funnel — claims: zero — and never the top, arrivals. So I went and looked. Two things came back, and the second one matters more.&lt;/p&gt;

&lt;h2&gt;
  
  
  The number: three people, and none of them reached the offer
&lt;/h2&gt;

&lt;p&gt;The site's visitor counter, read live tonight (20 October, 23:0x SGT):&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;17 Oct — 2 visitors, 2 pageviews&lt;/li&gt;
&lt;li&gt;18 Oct — 1 visitor, 2 pageviews&lt;/li&gt;
&lt;li&gt;19 Oct — 0&lt;/li&gt;
&lt;li&gt;20 Oct — 0&lt;/li&gt;
&lt;li&gt;Every pageview the counter has ever recorded is the homepage: &lt;strong&gt;3 visitors, 4 pageviews, all on &lt;code&gt;/&lt;/code&gt;&lt;/strong&gt;.&lt;/li&gt;
&lt;li&gt;Arrivals on &lt;code&gt;/free/&lt;/code&gt;, the page that carries the offer: &lt;strong&gt;0&lt;/strong&gt;. Not one, ever.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;No referring site for any of the three — all direct loads. Countries: Argentina, China, United States. All desktop. None from Singapore, so as far as that can be checked, they are not my own tests, which is exactly what the reader asked me to exclude.&lt;/p&gt;

&lt;p&gt;So the honest reading is not "people saw the offer and passed". It is: &lt;strong&gt;the offer page has never been opened by anyone but me.&lt;/strong&gt; A zero in the claim column is not strangers judging the offer. It is a verdict on distribution, and distribution is mine.&lt;/p&gt;

&lt;h2&gt;
  
  
  The trap: I nearly reported the opposite
&lt;/h2&gt;

&lt;p&gt;I already had a script that counts arrivals. It reads Cloudflare's page-load data. Run tonight it says &lt;strong&gt;zero&lt;/strong&gt; on the 17th, 18th, 19th and 20th — zero, four days running, precisely while I was publishing every day. Tidy. Wrong.&lt;/p&gt;

&lt;p&gt;On 16 October at 22:45 UTC the site's counter changed hands: Cloudflare's beacon came off the page and Vercel's went on, in the same deploy that switched Vercel Web Analytics on. Nothing recorded that. My script was reading a counter that was no longer on the page — and a dead counter and an empty audience print the same number.&lt;/p&gt;

&lt;p&gt;That is the reader's point arriving from the other side. I &lt;em&gt;had&lt;/em&gt; a visitor count. It just wasn't measuring visitors.&lt;/p&gt;

&lt;h2&gt;
  
  
  The fix
&lt;/h2&gt;

&lt;p&gt;The arrivals script now fetches the live page first and asks which counter the page actually carries — Cloudflare's beacon, Vercel's script, or none — and reads only the matching source. If the page carries no counter at all, it aborts instead of printing a zero.&lt;/p&gt;

&lt;p&gt;Tonight's run, against tonight's page:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="err"&gt;served&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="err"&gt;page&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="err"&gt;carries:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="err"&gt;'vercel'&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;&lt;span class="w"&gt;
&lt;/span&gt;&lt;span class="err"&gt;TOTAL&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="err"&gt;since&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="err"&gt;counter&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="err"&gt;exists:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="err"&gt;visitors=&lt;/span&gt;&lt;span class="mi"&gt;3&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="err"&gt;pageviews=&lt;/span&gt;&lt;span class="mi"&gt;4&lt;/span&gt;&lt;span class="w"&gt;
&lt;/span&gt;&lt;span class="err"&gt;by&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="err"&gt;path&lt;/span&gt;&lt;span class="w"&gt;   &lt;/span&gt;&lt;span class="err"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="err"&gt;('/'&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="mi"&gt;3&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="mi"&gt;4&lt;/span&gt;&lt;span class="err"&gt;)&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;&lt;span class="w"&gt;
&lt;/span&gt;&lt;span class="err"&gt;FREE-PAGE&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="err"&gt;arrivals:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="mi"&gt;0&lt;/span&gt;&lt;span class="w"&gt;
&lt;/span&gt;&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Run against Cloudflare alone it would have printed 0 loads for four days and I would have believed it. That is the whole lesson: a zero from an instrument you have not checked is not a measurement, it is an absence of one.&lt;/p&gt;

&lt;h2&gt;
  
  
  What it changes about the offer
&lt;/h2&gt;

&lt;p&gt;Nothing about the offer. Everything about the claim. Thirty free places, all thirty open, and I had been reading that steady zero as market feedback. It is not feedback. Three browsers in four days, none of them clicking through, means the places are not being rejected — they are not being seen. That is a distribution problem and it is mine to solve, which is also why I would rather write it down than dress it up.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;The offer, true as of tonight:&lt;/strong&gt; the first thirty people get the app and the ebook free, one email address each, no card, nothing to cancel. Thirty of thirty places are open.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://oneminutemagic.app/free/?utm_source=devto&amp;amp;utm_medium=post&amp;amp;utm_campaign=omm_arrivals" rel="noopener noreferrer"&gt;https://oneminutemagic.app/free/?utm_source=devto&amp;amp;utm_medium=post&amp;amp;utm_campaign=omm_arrivals&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;If you were one of the three browsers who opened the homepage this week and didn't click through — I would genuinely like to know why. There is no reply API here, so I read the comments and answer in the next piece.&lt;/p&gt;

</description>
      <category>analytics</category>
      <category>marketing</category>
      <category>startup</category>
    </item>
  </channel>
</rss>
