Seven days after launching the convenience store BOGO deals app, my App Store Connect dashboard showed 47 impressions, 3 product page views, and a perfectly round zero in the installs column.
I stared at that zero for an embarrassingly long time. Not angry, just confused — like I'd set up a lemonade stand, hung a sign, and watched 47 people walk past without even slowing down. The natural instinct was to blame everything at once: the icon, the screenshots, the description, the algorithm, Mercury in retrograde. But that's not how you fix anything. So I tried to actually read the funnel instead of just feeling bad about it.
Here's what the numbers broke down into, stage by stage.
Stage 1: 47 Impressions Is Not a Launch, It's a Whisper
First thing I had to accept — 47 impressions in 7 days is not a real signal about product quality. It's barely a signal at all. That number comes almost entirely from App Store search, and it means a small handful of people searched something vague like "convenience store deals" or "BOGO finder," my app showed up somewhere on the results page, and they registered as an impression.
For context: App Store Connect counts an impression only when at least half of your icon is visible on screen for at least one second. So 47 means maybe 80–100 people had a search results page loaded that happened to include me. That's not a crowd. That's a quiet Tuesday afternoon.
The benchmark I'd read about — loosely, somewhere around 2–5% impression-to-install rate for cold launches with no marketing — would have meant roughly 1 install from 47 impressions. Which is almost statistically consistent with zero. So the total number of impressions was the first problem, not the conversion.
47 impressions in a week isn't a funnel problem — it's a visibility problem wearing a funnel's clothes.
Stage 2: 3 Product Page Views Out of 47 (6.4% Tap-Through)
Out of 47 impressions, 3 people tapped through to the actual product page. That's roughly 6.4%. Industry average for a cold, unbranded app sits somewhere around 5–8%, so this wasn't broken — it was just tiny. The icon and title weren't repelling people. They just weren't attracting many in the first place.
I'd spent about two hours on the icon using a $0/month Figma free tier and some stock-ish illustrations. Nothing special, but nothing embarrassing either. The title was straightforward: category noun, utility verb, done. No clever wordplay that might confuse.
The 6.4% tap-through told me the icon and title were doing their job at a mediocre-but-acceptable level. This stage wasn't the leak.
Stage 3: 3 Views, 0 Installs (0% Product Page Conversion)
This is where it fell apart. Three people opened the product page and all three left without installing. A 0% product page conversion rate. That one actually stings a little, because it means something on the page itself — the screenshots, the description, the preview video I didn't make — failed to close.
I went back and looked at the screenshots with fresh eyes.
They were functional. Showed the UI. Listed features in the caption text. Completely forgettable. I'd built them the same way I'd built the app — efficiently, without thinking about whether a stranger would find them compelling in six seconds of scrolling. Screenshot 1 showed the app's main list view. Screenshot 2 showed a filter panel. Screenshot 3 showed a deal detail page.
What I hadn't shown: the moment of value. A user at a convenience store, three seconds from checkout, tapping the app and finding out the chips they're about to pay full price for are BOGO today. That story was invisible in my screenshots. I showed them a database. I should have shown them the win.
My screenshots were a feature list pretending to be a reason to care.
The description had the same problem. First paragraph: what the app does. Second paragraph: how it works. Third paragraph: a bullet list of features. I was writing documentation for users who'd already decided to install, not a pitch for users who hadn't.
Stage 4: What the Missing Reviews Cost Me
Zero installs also means zero reviews, which means the social proof section of the product page is a void. This compounds fast. The three people who visited my page saw no ratings, no review count, no indication that a single human had used this thing and survived. For a utility app asking to live on someone's phone, that silence is loud.
I knew this going in, abstractly. What I didn't expect was how quickly it would feel like a wall rather than a gap to fill over time. Getting the first 10 reviews is disproportionately hard compared to getting the next 90, and right now I had a product page that was quietly screaming "nobody trusts this yet."
Stage 5: The Keyword Audit I Should Have Done Before Launch
I ran a rough keyword audit post-launch using AppFollow on their $0 free tier (limited to 5 apps and delayed data, but enough for my purposes). The terms I'd optimized for were reasonably accurate descriptions of the app — but they had high difficulty and low volume for an unbranded new entry.
Turns out the search terms that actually bring traffic to apps like mine are more specific and more action-oriented than I'd assumed. Things like "7-eleven deals today" or "gs25 coupon" (I'm paraphrasing to keep things generic) — hyper-local, hyper-immediate queries. My keyword field was full of categorical terms that put me in competition with apps that had thousands of reviews. That was a positioning mistake I made before a single person ever searched.
I optimized for keywords that described what I built instead of keywords that matched what people were actually searching for.
Where the 47 Actually Came From
Breaking it down roughly: about 38 of the 47 impressions came from search, 6 from browse (probably "New Apps We Love" adjacent sections, where I briefly appeared as a new submission), and 3 from some referral source I can't identify — maybe someone shared a direct link once, maybe it's noise.
The 6 browse impressions converting to 0 page views is interesting. Browse traffic has lower intent by definition — people aren't looking for me, they're scrolling. That they didn't tap isn't surprising. The icon didn't hook anyone mid-scroll.
What I'm Actually Changing
I'm not rebuilding the app. The product works fine for what it does. What I'm changing, over the next two weeks:
New screenshots, built around a narrative arc instead of a feature list. I'm using the same Figma free tier, but spending the time I didn't spend the first time. The first screenshot will show a scenario, not a UI.
Keyword field gets rebuilt from scratch based on what AppFollow's free data shows about actual search volume for this category.
And I'm going to go find 10 people who might genuinely use this thing and ask them to use it — not review it, just use it — before I touch the listing again. Real usage first, ask for reviews second.
The 47 number doesn't depress me anymore. It's just a starting point with better documentation than most starting points get.
TL;DR: 47 impressions and zero installs isn't a mystery — the funnel broke at visibility first, then at screenshots that showed features instead of the moment of value, with keywords that competed where I couldn't win.
Next episode: I rebuild the screenshots from scratch and resubmit — tracking whether the product page conversion rate moves at all within the first 30 days.
Get the $31 Stack Prompt Pack — $19
The exact prompts, cost calculator, and n8n templates from this stack.
Top comments (0)