DEV Community: Steven Snell

AI Photo Editing Tips and Tools

Steven Snell — Fri, 21 Aug 2026 12:37:15 +0000

Last year I spent $20 on Fiverr to get a background removed from a headshot. The turnaround was two days. The result was fine. Then someone mentioned remove.bg, and I did the same job in about four seconds for free.

That was the moment I realized the errand was gone. Not simplified, not cheaper. Gone.

I've been comfortable with image tools for a long time. I know my way around Photoshop. I use Canva regularly. I'm not a designer, but I've never been the type to panic when someone hands me a PSD file. So when AI started showing up inside the tools I already used, I didn't have a dramatic conversion moment. It was more like the water got warmer without anyone turning the dial.

Background removal is basically solved

This is the clearest win. Tools like remove.bg, PhotoRoom, and Clipdrop have turned what used to be a careful, tedious masking job into something you don't even think about anymore. Hair edges, transparent objects, low-contrast scenes. I still remember spending real time with the pen tool on those kinds of cutouts. Now I upload the image, and it's done before I've finished reaching for my coffee.

PhotoRoom has become especially useful for anyone doing ecommerce or product content. It handles the removal and then lets you drop the subject onto AI-generated backgrounds with matched lighting and shadows. Remove.bg is still my go-to when I just need a clean cutout and nothing else, because it does that one thing extremely well and doesn't try to upsell me on a full editing suite while I'm there.

For most content work, any of these will get you there. The differences only show up on the hard stuff like wispy hair or semi-transparent fabric. For a blog featured image or a social post, it genuinely does not matter which one you pick.

Typing your way through an edit

The stranger shift is generative editing. Nano Banana, Google's image model inside Gemini, lets you upload a photo and describe what you want changed in plain English. "Remove the person in the background." "Change the wall color to blue." "Add rain to the window." You type it like you're leaving a note for a retoucher, and the AI just does it.

I've used Photoshop's Generative Fill for similar tasks, and it's impressive in a different way. It's surgical. You select an area, describe what should go there, and it fills it in with something that usually looks convincing at normal viewing distance. For fixing a distracting element in a photo or extending a background to fit a wider crop, it saves real time.

But here's the thing I keep coming back to. When I use these tools, the work doesn't feel like editing anymore. It feels like directing. I'm not making decisions with my hands the way I do when I'm adjusting curves or painting a mask. I'm writing short instructions and evaluating the output. Which is, honestly, closer to what I do all day as a writer anyway. Maybe that's why it feels so natural, and also why it feels slightly off. The skill I spent years building in Photoshop, the muscle memory around selections and layers and blend modes, just matters less now for a lot of the tasks I used to use it for.

The upscaling and cleanup side

Then there's the restoration and enhancement category. Topaz Photo AI is the standout here. Feed it a noisy, soft, or low-resolution image and it will denoise, sharpen, and upscale it in ways that genuinely look good. I've rescued client photos that I would have told them to reshoot two years ago. Luminar Neo does similar work with a broader creative toolkit on top, things like sky replacement and automatic portrait retouching that lean more toward stylistic choices than corrections.

These tools sit in a different part of the workflow than the generative stuff. They're fixing problems rather than inventing content. And for that reason, I trust them more instinctively. When Topaz sharpens a blurry photo, I can see what it did and judge whether it worked. When Nano Banana adds rain to a window that never had rain, I'm relying on the AI's taste as much as my own.

The part nobody talks about

Here's what actually changed for me, and I don't see this discussed enough. Because I can handle all of this myself now, I'm expected to. One client I work with used to have an in-house person who'd take my article text and build the featured image. At some point last year they stopped doing that and started asking me to supply finished images with the draft. No conversation about it. No adjustment to the rate. The capability just silently became part of the job description.

I'm not bitter about it. The tools make it fast enough that it doesn't wreck my day. But I do think writers and content people should pay attention to this pattern of expanding expectations. Every tool that makes something easy also makes it expected. That's not a reason to avoid the tools. It is a reason to think about where your rate reflects the full scope of what you're delivering.

Where I've actually landed

I use remove.bg or PhotoRoom weekly. Canva's AI features have replaced a lot of what I used to open Photoshop for. I reach for Nano Banana when I want to experiment with something visual and I'm not sure what I want yet. And I still open Photoshop when the job actually requires precision, because nothing else gives you that level of control when you need it.

The overall feeling is positive, with a caveat. These tools made me faster and more self-sufficient, and I like both of those things. But they also compressed a set of skills that used to take years to develop into something anyone can do with a text prompt. If you're someone who built value around being the person in the room who could handle images, that compression is worth sitting with for a minute.

Has your job quietly absorbed skills you never originally signed up for?

Building a Client Referral Pipeline

Steven Snell — Wed, 19 Aug 2026 11:42:52 +0000

I went through my invoices at the end of last year and noticed something I probably should have noticed sooner. My three highest-paying projects had all come from the same person. A former client I'd done a website content overhaul for, maybe eight months earlier. She'd recommended me to a colleague, who recommended me to someone else, who then looped me in on a bigger contract.

I hadn't asked her to do any of that. I didn't have a referral program. I didn't even have a system for staying in touch with past clients. I'd just done decent work, and she happened to think of me at the right moment.

That "happened to" part is what eventually bothered me.

The Advice That's Almost Right

You hear it constantly in freelancing circles: do good work and referrals will come. And it's not wrong. Nobody refers a freelancer who missed deadlines and delivered sloppy copy. Quality is the floor, not the strategy.

But here's what that advice leaves out. Your former client might have loved working with you. They might even want to recommend you. The problem is that the moment has to line up. Someone in their network has to mention needing a writer, and your name has to surface in their memory right then, and they have to feel confident enough in what you do to explain it on your behalf. That's a lot of conditions, and you're leaving all of them to chance if you don't do anything after the project wraps.

I left them to chance for years. It worked sometimes. It could have worked a lot more.

What I Actually Started Doing

The biggest shift was embarrassingly simple. When a project ended well, I started saying the words out loud. Something like, "If anyone in your network ever needs content or copywriting help, I'd love an introduction." That's it. No pitch deck, no referral link, no incentive structure. Just a clear, low-pressure mention that I'm open to it.

I was surprised how much difference that made on its own. I think most clients genuinely don't consider referring you unless the idea is planted. They're busy. They liked working with you, but they're not walking around thinking about your pipeline.

The second thing was staying visible to past clients without being annoying about it. I started doing a loose quarterly check-in. Not a newsletter, not a "just touching base" email that screams sales follow-up. I'd reply to something they posted on LinkedIn, or send a two-line email asking how a project turned out. Sometimes I'd share an article that related to something we'd worked on together. The goal was just to stay in the recent-memory part of their brain, so that when someone did ask them for a recommendation, I wasn't a name they had to dig for.

The third thing took me longer to figure out, but it might matter the most. I gave people language to describe what I do. This sounds weird until you realize that most of your past clients, if asked to recommend a writer, would struggle to explain your specialty. "He's a good writer" doesn't get someone to reach out. But "He does long-form content for SaaS companies and he's fast" gives their contact enough to act on. I started including a one-liner in my wrap-up emails that made it easy to pass along. Something they could literally copy and paste into a message.

I also started tracking where my leads actually came from. Nothing fancy. A spreadsheet column that logged the source for every inquiry. After about six months, the data was clear enough to change how I spent my time. Referrals closed faster, paid better, and almost never haggled on my rates. Cold leads from job boards were the opposite on every count.

What Didn't Work

I tried a referral incentive once. Ten percent off a future project if someone sent a client my way. It felt wrong almost immediately. The kind of work I do isn't transactional in that way, and framing it like a bounty cheapened the relationship. The client I offered it to seemed slightly uncomfortable, and I never brought it up again.

I also learned that timing matters more than I expected. Asking for referrals too early, before a project is fully delivered and the client has seen results, puts people in an awkward spot. They want to help, but they're not ready to vouch for you yet. I wait until the work is done and they've had a chance to actually use it before I mention anything about introductions.

Where Things Stand Now

Referrals aren't my only source of work, and I wouldn't want them to be. Relying on a single channel is its own kind of risk. But when I look at the projects I've enjoyed most and the clients who were easiest to work with, almost all of them came through a referral. The fit is just better when someone who knows your work sends someone your way.

The whole system, if you can even call it that, takes maybe an hour a month. A few check-ins, a few wrap-up emails with the right sentence included, and a spreadsheet I glance at quarterly. Nothing about it is sophisticated. I just had to actually start doing it instead of hoping it would happen on its own.

Have you built any kind of referral system into your freelance work, or has it mostly been accidental? I'm curious whether other people found the same things worked for them.

Cold Emails That Actually Get Replies (From Someone Who Sent a Lot of Bad Ones First)

Steven Snell — Tue, 18 Aug 2026 17:09:52 +0000

About two years into freelancing, I sat down on a Sunday night and sent 40 cold emails. I'd found a template on some blog, swapped in my name and a few details, and blasted it out to marketing directors I'd pulled from LinkedIn. I felt productive. I felt like I was doing the work.

I got one reply. It was someone telling me to take them off my list.

That was my introduction to cold outreach, and it almost convinced me the whole approach was broken. It took a while to figure out that the approach wasn't the problem. My emails were.

What bad cold emails actually look like

Here's the thing I didn't understand early on. I thought my emails were personalized because I used the person's name and mentioned their company. But read one of those emails out loud, swap in any other company name, and it still makes perfect sense. That's not personalization. That's a mail merge.

The emails I was sending all had the same bones. A compliment I didn't really mean, a paragraph about my services that read like a brochure, and a vague ask to "hop on a call." Every freelancer sending cold emails was writing some version of this. Every recipient had seen it before. I was wallpaper.

The shift that actually mattered

I didn't have some big epiphany. I just got tired of wasting time on emails nobody read and started being pickier about who I contacted. Instead of 40 emails on a Sunday, I'd send maybe five or six across a whole week. But I'd spend real time on each one.

Before writing anything, I'd go read something the person had actually published. A blog post on their company site, a LinkedIn article, even a social media thread. I was looking for a gap, something their content was missing that I could speak to. Not in a "gotcha" way, more like "I noticed you're covering X but haven't touched Y, and I've written about that for other clients." That framing changed everything.

The reply rate went from essentially zero to something real. Not a flood, but enough that cold email stopped feeling like shouting into a void. Two of my longest-running client relationships started from emails like that. One of them told me months later that she replied because it was obvious I'd actually read her work, and that almost never happened.

A few things that matter more than you'd expect

Subject lines are the whole game and I ignored them for way too long. I used to write subject lines like "Freelance Content Writer Available" which is basically begging to be archived. What worked better was something specific and short, tied to whatever I'd noticed about their content. Something that read like a person wrote it, not a pitch deck.

Brevity matters too. My early emails ran four or five paragraphs. Nobody wants to read a cover letter from a stranger in their inbox at 9 a.m. I got better results when the whole email fit on a phone screen without scrolling. Three short paragraphs, tops. Say what you noticed, say what you can do about it, and make a small ask.

And the ask itself is worth thinking about. "Can we hop on a 30-minute call?" is a big ask from someone you've never met. "Would it be okay if I sent over a couple of ideas for your blog?" is small enough that saying yes costs almost nothing. I started getting replies the moment I stopped asking people to invest time in me before I'd earned it.

What I still get wrong

I'm not going to pretend I've solved cold email. I still overthink the opening line sometimes and spend twenty minutes on a sentence that probably doesn't matter. My follow-up game is inconsistent. Some weeks I'm disciplined about it, other weeks I completely forget, and by the time I remember, the window has closed.

I also still occasionally send one that's too long because I get excited about the angle and forget that the recipient doesn't care about my enthusiasm. They care about their own problems. Every time I catch myself writing a long cold email, I know I've drifted back into writing for myself instead of for the person reading it.

The uncomfortable truth about volume

A lot of cold email advice leans hard on volume. Send more, automate more, scale it up. That advice might work if you're selling software, but freelancers aren't selling a product. You're selling yourself, and "yourself at scale" is a contradiction. The moment a cold email feels automated, it's dead.

Fewer emails, more thought per email. That's the whole thing. It's slower and it doesn't feel as productive on a Sunday night, but it's the version that actually led to work.

Has cold outreach worked for you, or have you found a better way to land freelance clients? I'm genuinely curious what's working right now.

Email Marketing for Freelancers Who Keep Putting It Off

Steven Snell — Tue, 18 Aug 2026 17:07:33 +0000

Every freelancer eventually hears the same advice: start an email list. And most of them nod, agree it sounds smart, then never do it. Not because they don't believe it works, but because the guides that explain how are written for marketing teams with budgets, content calendars, and dedicated designers. When you're one person trying to finish a client project by Thursday, "build a welcome sequence with segmented automations" sounds like homework from a class you didn't sign up for.

The good news is that email marketing for a solo freelancer looks nothing like what those guides describe. You don't need a funnel. You don't need a content upgrade. You need a way to stay visible to people who already know your work, so that when they need someone, you're the name that comes to mind.

A small list is the whole point

There's a widespread assumption that email lists need to be big to be useful. That makes sense if you're selling a product to strangers. But freelancers aren't doing that. You're maintaining relationships with a few dozen people who've worked with you, considered working with you, or know someone who might.

Fifty subscribers who actually open your emails will generate more leads over a year than a thousand people who signed up for a freebie and forgot about you. The economics of freelance email are entirely different from ecommerce or media. A single reply from the right person on your list can turn into a project worth more than months of social media effort.

So stop thinking about "growing" your list in the way marketers talk about it. Think about it as keeping a short roster of real contacts warm.

What you actually send

This is where most freelancers stall. They picture themselves having to produce a polished newsletter with sections, a header graphic, and a consistent publishing schedule. That's optional. It can also be a plain email that takes twenty minutes to write.

A few things that work well at this scale: a quick take on something happening in your niche, a link to something useful you found that week, a brief update on what you've been working on (without naming clients, if that's sensitive), or a short lesson from a recent project. You can mix these however you want. One email might be two paragraphs and a link. The next might be a few hundred words on a specific problem you solved.

Frequency matters less than showing up at all. Monthly is fine. Every two weeks is fine. The worst frequency is "once, enthusiastically, then never again," which is what happens when people commit to weekly before they've figured out what to say.

Picking a tool without the spreadsheet

The tool decision paralyzes people who should be writing their first email instead. Here's the short version based on where things stand right now.

Kit (formerly ConvertKit) has the most generous free tier for a solo creator. You can have up to 10,000 subscribers on the free plan with unlimited sends. The catch is that automations are extremely limited at that level, but for a simple newsletter you won't need them. Beehiiv gives you 2,500 subscribers for free, also with unlimited sends, and includes a hosted website for your archive. Buttondown is minimal and clean, free up to 100 subscribers, with paid plans starting at $9 a month if you outgrow it. Mailchimp still exists, but its free plan has been cut to 250 contacts and 500 sends per month, which barely covers a monthly email to a modest list.

For a freelancer just starting, Kit or Beehiiv are the obvious choices because you won't hit a paywall for a long time. Pick one, set it up in an afternoon, and move on. You can always migrate later.

Getting the first subscribers without a lead magnet

The lead magnet industrial complex has convinced people that nobody will subscribe to anything without receiving a free PDF first. That's true when you're trying to attract cold traffic from Google. It's not true when you're emailing people who already know you.

Your first fifteen to twenty subscribers come from asking. Past clients. People you've collaborated with. Contacts from LinkedIn who you've had actual conversations with. Someone who emailed you about a project that didn't work out. You already have these relationships. You just need to tell people the list exists and give them an easy way to join.

A one-line mention in your email signature works. A short post on LinkedIn or wherever you're active works. If you have a portfolio site, a signup form on it works. None of this requires a landing page with testimonials and a countdown timer. You're not launching a product. You're inviting people you know to hear from you occasionally.

The actual payoff

A freelancer's newsletter doesn't need to be a content product. It doesn't need sponsorships, paid tiers, or a growth strategy. Its job is simpler and arguably more valuable: keeping you in the peripheral vision of people who might hire you or refer you.

Freelancers lose work to invisibility more often than to competition. Someone needed a writer last month, and they went with whoever came to mind first. That's it. An email that shows up every few weeks with something genuinely useful is enough to be that person. The entire system can be one tool on a free plan, a couple of hours a month, and a list that never gets bigger than your actual professional network. That's not a limitation. That's the design.

Has anyone here tried running a simple newsletter alongside client work? I'm curious what made it stick or what made you quit.

How to Create Strong Product Listings for E-Commerce Websites

Steven Snell — Sun, 16 Aug 2026 21:26:27 +0000

A client sold a canvas weekender bag. Nice thing, made well, sold steadily at markets and in her little shop. Online it just sat there. Traffic came in, people looked, almost nobody bought. She figured the copy was the problem and hired me to fix it.

So I wrote her some copy. Really wrote it. I had a line about the bag being the kind of thing you'd hand down to someone, and I was proud of it. Conversions didn't move.

Second pass, I leaned harder into the maker story, the workshop, the whole hand-stitched angle. A little bump, maybe, hard to say. Still not what either of us wanted.

The third version is the one that worked, and it's slightly embarrassing how simple it was. I stopped trying to make anyone feel something and started answering the questions they were clearly already asking. How big is it actually? Will it fit a laptop? What happens if I hate it when it shows up, and where does the thing even ship from? None of that is clever. All of it is what stood between a curious person and their credit card.

That reframed how I think about these pages entirely. A product page earns its keep by removing doubt. Someone's already interested or they wouldn't be on the page, so your job is to clear the small pile of hesitations sitting between them and checkout. Most of those hesitations are boring, practical, and completely answerable.

Here's what I've picked up since, mostly the hard way.

Write to the question, not the product

The instinct is to describe what you're selling. Dimensions, materials, features, all of it. That's fine, but it buries the reader in your priorities instead of theirs.

Flip it. For whatever you sell, there's a short list of things a buyer needs to know before they'll commit, and they're usually thinking it in the form of a worry. "Is this going to be too small." "Will it work with the thing I already own." "Is that a real photo or a rendered one." Write the copy that answers those head-on, and put the most common worry near the top where a scanner will catch it.

I keep a rough habit now. Before writing a line of a page, I list the five questions I'd ask if I were about to spend my own money on the thing. Then I make sure the page answers all five before it says anything pretty.

Cut the adjectives that don't carry weight

Writers love adjectives on product pages because they feel like the job. Words like premium, luxurious, thoughtfully designed. Ask what any of those actually tell a buyer and the answer is nothing much. They're the seller talking to themselves.

"Premium leather" means less than "the leather softens and darkens with use, so it looks better at year three than day one." One is a claim. The other hands someone a picture of owning the thing. Specifics do the persuading, and vague praise mostly takes up space. If anything it makes people trust you slightly less, because everyone says it.

Let the images answer what words can't

I'm a writer, so this one took me a while to admit. On a product page the pictures often do more work than the copy, and the two should be splitting the labor on purpose.

The photo's job is to handle what text is bad at. Scale is the big one. A bag floating on a white background tells me nothing about size, so show it on someone's shoulder, or next to a laptop so the proportions read instantly. Show the thing in use, in a real hand, in real light. If there's a detail people fixate on, the zipper quality or how a fabric actually drapes, get a shot of exactly that. A single honest photo of the part someone's unsure about can close a sale that three paragraphs couldn't.

And don't let the pure hero shot be your only image. It's beautiful and it's useless for deciding anything.

Respect how people actually read the page

Nobody reads a product page top to bottom like an essay. They scan, jump to the price, bounce down to reviews, hunt for shipping, come back up. The page has to survive being read out of order.

So headers and short chunks beat long dense paragraphs. And the practical stuff, sizing and shipping and returns, can't be hidden three clicks deep behind a "Details" tab. Every extra click you make someone take to answer a basic question is a small invitation to leave.

Returns especially. A clear, generous, easy-to-find return policy kills more purchase anxiety than almost anything you can write, because it quietly tells people the cost of being wrong is low.

Reviews belong in this conversation too. People trust other buyers more than they'll ever trust you, and a review that names the exact thing someone's worried about does a kind of persuading you just can't manage from the seller's chair. Make them easy to find, and don't scrub the mildly critical ones. A wall of nothing but five stars reads as fake, and people can smell it.

The unglamorous version of the job

If I had to compress all of this into one working rule, it's that a good product page is an act of anticipation. You sit where the buyer sits, list every reason they might hesitate, and answer each one before they have to go looking. The copy that does that rarely wins awards. It just works.

The clever line about handing the bag down? I kept it. Moved it near the bottom, after all the real questions were answered. Turns out a nice sentence lands a lot better once someone already knows the thing fits their laptop and ships in two days.

What's the last product page that actually talked you out of a purchase, and do you know which part of it did the talking?

Link Building Tips for E-Commerce Businesses

Steven Snell — Fri, 14 Aug 2026 10:36:49 +0000

A few years ago, I picked up a client who sold saltwater aquarium supplies. Real store, real products, the kind of thing hobbyists pour an alarming amount of money into. He came to me with a simple ask. He wanted links pointing at his category pages, the ones for protein skimmers and reef lighting, because those were the pages he needed to rank.

So we tried. I found relevant blogs and forums, wrote the outreach, pitched the pages. Nobody bit. Not one person wanted to link to a page that existed to sell them a $400 skimmer, and honestly, why would they? A category page is a storefront. People link to things that help them, and a storefront doesn't help anyone who isn't already shopping.

That took me longer to accept than I'd like to admit. The pages you most want links to are usually the pages no one has any reason to link to.

Once I stopped fighting that, the work got easier. We quit chasing links to money pages and started building the pages people actually reference, then used internal links to pass that authority down to the products. That's most of the strategy right there. Everything below is a variation on it.

Build the reference page your niche is missing

Here's where I'd spend the bulk of my effort if I were starting an ecommerce link campaign today.

Every niche has a question people ask constantly and answer badly. For the aquarium store it was cycling a new tank, the weeks-long process of getting the water chemistry right before you add fish. Beginners get it wrong, kill their fish, and feel terrible. There were forum threads about it going back a decade, but no single clear, calm, thorough guide that a person could send to a friend.

So we made one. A real reference, long and specific, with the mistakes people make and how to sidestep them. It took genuine time to write and it read like someone who'd actually done it, because I interviewed the client until it did.

That page earned more links in six months than every product page combined. Other hobby sites linked to it. A couple of forum moderators pinned it. One university extension program cited it. And every one of those pages linked internally to the skimmers and test kits a person would need to actually do the thing, which is how the ranking value found its way to the products anyway.

The trick is simple. Find the thing your customers need to understand before they buy, and make the best explanation of it that exists. Patient and genuinely useful. Those pages age well, and people keep pointing at them for years.

Use the data you're already sitting on

Ecommerce stores generate data constantly, and most of them never do anything with it.

You know what sold, what got returned, what people searched for on your site, which regions buy what, how your prices drifted over a year. Journalists and bloggers love that kind of thing because they can't get it anywhere else. A quick post along the lines of "we looked at 12,000 orders and here's what people actually buy for their first reef tank" is the sort of original angle a writer will happily credit you for.

You don't need a research department. You need one honest number nobody else has, and a clear writeup of what it means. I've watched a single modest data post out-earn months of ordinary outreach, because it gives other people something to say.

Get the easy links you're ignoring

Some of the simplest links in ecommerce are sitting in your own supply chain.

If you stock other brands, a lot of those manufacturers keep "where to buy" or authorized retailer pages, and they'll list you if you ask. Suppliers, trade associations, local business groups, the makers whose products you carry. These won't move mountains, but they're relevant, they're real, and most stores never bother to claim them. Send the emails. It's an afternoon of unglamorous work that quietly pays off.

Be a quotable source

The old HARO era is over. Cision shut it down at the end of 2024, then the brand came back under Featured, and the whole thing splintered into a handful of tools like Qwoted and Source of Sources. The tactic survived the shakeup, though. Reporters still need experts on deadline, and they still credit those experts with links.

If your client or their team actually knows the niche, they can answer relevant queries and land links from publications you could never pitch cold. The wins go to whoever replies fast, leads with a specific quote a writer can drop straight in, and skips the filler. For a store owner with real expertise, it's some of the highest-authority linking you can get for the price of a few thoughtful emails a week.

What I stopped doing

I'll save you the money I wasted. Mass guest-post placements on sites that exist only to sell links did nothing that lasted. Buying "niche edits" was worse. Blasting resource-page requests to 300 addresses got me a 1% response rate and a bad taste in my mouth. The stuff that worked was slower and more human, and it kept working long after the shortcuts stopped.

If I had to compress all of it into one sentence, it's this. Make things worth linking to, then make sure your product pages are one internal click away. The rest is just outreach and patience.

What's the most useful non-commercial page on your store right now? If you can't name one, that's probably where I'd start.

Invoicing Tips for Freelancers

Steven Snell — Thu, 13 Aug 2026 10:30:53 +0000

For a long time, my invoices read like an apology.

I'd tack on a line about "no rush" or "whenever works." I'd bury the total somewhere in the middle of a paragraph instead of putting it where anyone could actually find it. On one invoice, early on, I wrote "sorry to send this over" right at the top, like asking to be paid for finished work was an imposition I needed forgiveness for.

Clients paid late. I figured that was just the job.

It took me an embarrassing while to connect those two things. The invoice wasn't some neutral bit of admin that happened after the real work ended. It was the last thing a client experienced from me, and I was closing out every single project on a note of hesitation.

The invoice is the last thing they feel

Think about how a project actually ends. You do great work, you deliver, maybe there's a warm email thread where the client tells you they're thrilled. Good feelings all around.

Then you send an invoice, and if that invoice is confusing or sheepish or vague, it drags the whole experience down a notch on the way out the door. The client's final impression of working with you isn't the polished deliverable anymore. It's the awkward money moment you created at the end.

I don't think most freelancers see it this way. We treat the invoice as a chore to get through, so it comes out rushed and half-hearted, and then we wonder why getting paid feels like pulling teeth. The client picks up on all of it. An invoice that looks like you weren't sure you were allowed to send it teaches people to treat it as optional.

What my bad invoices were actually communicating

Here's the uncomfortable part. When I went back and looked at the invoices from my first few years, the problem wasn't the numbers. The numbers were fine. It was everything around them.

"No rush, whenever's convenient" told the client the deadline was theirs to set. So they set it wherever they liked, which was usually much later than I'd hoped. The apology at the top told them I felt bad about the whole transaction, which is a strange feeling to hand someone right when you want them to take you seriously and open their wallet. And the missing or buried due date meant there was nothing concrete for anyone to act on.

Every one of those choices came from the same place, which was a quiet discomfort with asking for money. I dressed it up as being easygoing and low-maintenance. Clients read it as uncertainty.

The freelancers I know who get paid fast and without drama aren't harder or colder about it. They're just clear. Their invoices state what the work was, what it costs, when it's due, and how to pay. No hedging, no throat-clearing, no emotional weather report about how they feel sending it.

Rewriting the whole thing

So I rewrote mine, and honestly the changes were small. That's the part that annoyed me most, that it was this easy and I'd spent years not doing it.

First, I put the due date on there and made it specific. Not "net 30 whenever" but an actual date, in plain sight. A real date gives a client something to plan around. A vague window gives them permission to forget.

Second, I cut every apologetic word. All of it. The invoice now says what the work was and what's owed, and then it stops. It turns out you don't need a soft cushion of "so sorry to bother you" language around a number the client already agreed to pay.

Third, I made the line items readable to a normal human. Instead of one lump labeled "content services," I break it into what I actually delivered, in words the client would recognize from our conversations. This does two useful things. It reminds them of the value they got, and it heads off the "wait, what is this for?" email that stalls payment for a week.

And fourth, I stopped treating the payment terms as negotiable body language. Deposit up front for bigger projects, balance on delivery, due dates that mean something. When I present that as simply how I work, clients treat it as normal. Back when I presented it as a nervous request, they treated it as a nervous request.

The part that surprised me

I expected clients to push back on the more direct invoices. Some corner of my brain was braced for someone to reply, "wow, a little pushy with the due date, aren't we?"

Nobody did. Not once.

What actually happened is that people paid faster, and the money conversations that used to feel fraught just got boring, which is exactly what you want them to be. A clear invoice reads as competence. It signals that you've done this a hundred times, that you have a process, that you're a professional who expects to be paid on schedule because that's how professionals operate. Clients respond to that. They pay competent-looking invoices before they pay the sheepish ones, because the sheepish ones feel skippable.

I used to think the way to keep clients happy was to be maximally accommodating right up to the last second, including about money. I had it backwards. Being clear and matter-of-fact about payment is part of being good to work with. It removes a source of confusion and low-grade tension from the relationship, and it lets the project end the way it should, on the strength of the work rather than on an awkward ask.

The invoice is the last handshake. Make it a firm one.

How do your invoices read right now, if you go back and look at them cold? Mine didn't say what I thought they said.

5 Ways E-Commerce Businesses Can Use QR Codes

Steven Snell — Wed, 12 Aug 2026 10:49:26 +0000

A while back I bought a fairly expensive pour-over kettle, and there was a small QR code printed on the inside flap of the box. I scanned it. I'm not sure what I expected, maybe a setup guide or something about the temperature settings. What I got was the brand's homepage, the same page I'd already been on to buy the thing.

That sums up my problem with how most ecommerce brands treat QR codes. The code works fine. The scan does what it's supposed to. Then it drops you somewhere that has nothing to do with why you scanned in the first place.

I write content for a living, a lot of it for ecommerce clients, so I tend to think about this from the angle of what the person on the other end wanted in that moment. A QR code is a tiny promise. Someone bothered to pull out their phone and point it at your packaging, which is more effort than most people give any piece of marketing. The least you can do is send them somewhere built for that.

Here are five uses I've seen earn the scan.

1. Put a guide behind the code, not a store

The reflex is to link the code to your shop so people buy again. Skip that. The most valuable thing you can hand someone right after they've opened a product is help using it.

If you sell a coffee grinder, the scan should land on a two-minute video showing the grind settings for different brews. Sell skincare, and it's a short routine walkthrough. And if the product is genuinely fiddly, put the real instructions there, written in plain language, instead of cramming them onto a folded paper insert that nobody unfolds anyway.

I've written a fair number of these post-purchase guides for clients, and the good ones have something in common. They answer the first real question a customer has, fast, without a wall of marketing copy sitting in front of it. The reorder nudge can come later, once you've been useful. Lead with the sale and you've quietly told the customer the code was for you, not for them. That's the whole game with this one, and it's the reason I'd spend more time here than on anything else on this list. Get the destination right and the code stops feeling like a gimmick and starts feeling like service.

2. Catch the review while they're happy

The best time to ask for a review is the moment someone is holding the thing they just bought and liking it. That window closes fast.

A QR code on an insert card, pointing straight to a dead-simple review or photo-upload flow, catches people at the peak of that feeling. No login wall, no five-step form. Point, scan, tap a couple of stars and a sentence, done. I've watched a small insert like that lift review counts more than a whole sequence of follow-up emails, mostly because the emails land three days later when the excitement has already faded.

3. Kill the support ticket before it happens

A good share of "how do I..." emails come from people who never opened the manual, and honestly, fair. Manuals are miserable.

Print a code that opens a short FAQ or a setup clip and you head off a chunk of those questions before they reach your inbox. Less glamorous than the others, sure. But if you're a small team answering your own support, every ticket you prevent is time handed back to you. I'd rank this one higher than most people would.

4. Make reordering consumables a single scan

If you sell anything people run out of, the coffee or the dog food, put a reorder code right on the product itself, on the bag or bottle that stays on the shelf rather than the box that got recycled the day it arrived.

The moment someone notices they're low is the moment they'll act. If reordering is one scan instead of a dig through old emails for your site, you catch that impulse before it fades. This is about as close to a no-brainer as anything here.

5. Send print and pop-up traffic to a page that knows where it came from

If you're running a pop-up shop, a market stall, a print ad, or a campaign insert, the code shouldn't dump everyone onto the homepage. Build a landing page for that specific source and speak to it directly. Someone scanning at your booth already knows who you are. Someone scanning a magazine ad has no idea yet. Sending both to the same front page throws away the difference between them.

It doesn't have to be elaborate. A headline that matches whatever they were just looking at, plus one clear next step, will beat your generic homepage nearly every time.

The part that actually matters

Every one of these comes down to the same handful of seconds. You've got someone's attention right after they scan, and where you send them decides whether that attention was worth anything at all. A homepage throws it away. A page built for the moment turns a bit of printed ink into something people are glad they used.

Scanned a code lately that actually sent you somewhere useful? I'm curious what the good ones look like out in the wild, because I keep running into the bad ones.

5 of My Favorite AI Tools for Graphic Design

Steven Snell — Tue, 11 Aug 2026 12:44:40 +0000

As a writer, I always need images for my articles and blog posts. In the past, I would use stock photos, but AI has changed the game. Now, I can create custom graphics, edit and customize stock photos, or create my own photorealistic images with AI.

Here are the five tools I keep going back to, plus the honest reason each one stays on the list. Fair warning before we start. Not one of these made me a designer. They just let me describe what I want in normal words and hand me something usable, which is the only kind of design work my brain will tolerate.

Canva

I open this one first, every single time. Canva stopped being a template library a while ago and turned into the thing I use for maybe eighty percent of what I need. I type a rough description into Magic Design and it hands me a full layout I can push around until it stops looking generic. The background remover has saved me on more headshots and product shots than I can count. When I need the same graphic sized for a post header and a social card, one click reshapes it.

The AI images it spits out are fine. Good enough for a lot of jobs, though not the ones where the visual has to carry the whole piece. And if you lean on it hard, the monthly credits run dry faster than you'd expect, usually right when you're on deadline. Pro runs somewhere around fifteen dollars a month depending on which price you catch them on, and the free tier is genuinely enough to see whether it fits how you work. For a writer who just needs to stop shipping ugly posts, this is the whole game most days.

Nano Banana

This is what I use for cover images now. It's Google's image generation baked into Gemini, and the naming is a mess because there are a few versions floating around with slightly different price tags. Ignore that part. What matters is that it reads a long, fussy, paragraph-length description the way a person would, and gives back something that actually matches what I pictured. I can write out the mood, the lighting, and the framing I want for a 16:9 header, and it holds most of that together.

You can try it free with daily limits right inside the Gemini app, which is how I started. For blog covers where I want something photoreal and specific, this has quietly become my default.

Ideogram

Every other tool on this list falls apart the second you ask for readable words inside the image. Ideogram is the one that solved that. If I need a quote card or a thumbnail with an actual headline sitting on it, this is where I go. The text comes out clean instead of the usual alphabet soup.

The free tier gives you around ten prompts a day, and paid plans start near seven dollars, so it's a cheap thing to keep in your back pocket for the handful of posts where the words have to live in the picture.

Midjourney

I'll be honest about my history with this Midjourney. I bailed on it early because everything ran through Discord and I could not make myself care enough to learn the etiquette. Then they shipped a real web app, and I came back. When a post genuinely deserves a striking image, Midjourney still makes the best-looking output I can get out of a text prompt. It's the one I reach for when the visual is the whole point.

The catch is money and mechanics. There's no free tier anymore, plans start at ten dollars, and you're really buying GPU time rather than a set number of images, which took me a minute to wrap my head around. For someone publishing a few posts a week, the lowest plan burns through fast. I don't run it as my daily driver. I save it for the pieces that earn it.

Recraft

This is the odd one out, and I'll admit it's more tool than I strictly need. Recraft makes vector graphics, the kind that stay sharp at any size, and it's weirdly good at holding a consistent style across a whole batch. I started using it when I wanted a matching set of simple headers for a series so my posts didn't look like five strangers made them. If visual consistency across a run of articles matters to you, it's worth a look. If it doesn't, skip it without guilt.

What actually ties these together

Not one of these is a design education. That's the point. The ones that stuck are the ones that let me stay a writer, describe an image in a sentence or two, and get back something I'm not embarrassed to publish. The flashier a tool wants me to get, the less I use it.

What are you all using to make your covers? I'm always hunting for the one that finally makes this painless.

Writing to Get Cited by AI Is a Different Skill Than Writing to Rank in Google

Steven Snell — Mon, 10 Aug 2026 21:32:35 +0000

Type a question into Google right now and there's a decent chance you never leave the search page. The answer sits right there, generated on the spot, with maybe two or three source links tucked into the bottom of it. Ten blue links used to compete for a click. Now one paragraph competes for a citation.

That shift matters more than most content advice has caught up with. Ranking on page one used to be the finish line. Increasingly, the finish line is getting pulled into an answer that someone reads and never clicks through from at all. And getting pulled into that answer takes a different kind of writing than getting ranked ever did.

What Google Actually Rewarded

For twenty years, ranking well meant reverse-engineering an algorithm that was trying to guess what a human typed and wanted. That produced a specific kind of writing, one built around keyword placement and phrasing that matched whatever a person typed into the box. Length mattered too, since word count signaled thoroughness to an algorithm even when the extra words were just padding.

None of that was really about the words themselves. It was about satisfying a system that stood between the writer and the reader, on the assumption that satisfying the system was the only way to reach the reader at all.

What AI Systems Do Instead

An AI answer engine isn't ranking pages. It's extracting claims. It reads through a pile of sources, pulls out the sentences that most directly answer the question, and stitches them into a response. Nobody scrolls past that response to see where it came from unless they specifically want to check.

That changes what counts as good writing in a fairly specific way. A sentence that gets pulled out of a paragraph and dropped into someone else's answer either holds up on its own or it doesn't. If a claim only makes sense next to the three sentences before it, it never gets picked. If it depends on a "however" two paragraphs earlier to be accurate, it gets misquoted or skipped entirely.

Writing for extraction means writing sentences that survive being ripped out of context. That's a real skill, and it isn't the one a decade of on-page SEO trained people to build.

The Habits This Actually Changes

A few things shift once extraction, not ranking, becomes the target.

Answer the question first. Content built for search engines often built up to the point slowly, partly because length signaled depth to the algorithm and partly because writers were taught to hook readers before delivering value. An AI system isn't hooked by anything. It wants the clearest possible statement of the answer, and it wants it early.

Make claims that are true in isolation. A lot of content hedges by design: "it depends," "in some cases," "generally speaking." That hedging protects the writer, but it also makes a sentence useless to something trying to extract a clean answer. Compare "email marketing can be effective depending on your list and industry" with "a small, engaged list will outperform a large, cold one." The second version might be wrong in some situation somewhere. That's fine. It's specific enough to be useful, and specific enough to get pulled into an answer. Vague ones get passed over simply because there's nothing to grab onto.

Let one section stand on its own. A page written for a human reader can build an argument across ten paragraphs, each one leaning on the last. A page written to be extracted needs sections that work as standalone units, so any one of them could be the piece an AI system pulls without losing its meaning. That's a structural change as much as a sentence-level one, and it's probably the hardest habit to build, since most writers were trained to make paragraphs depend on each other.

Where This Overlaps With What Already Worked

SEO and AI-answer writing overlap more than they conflict. Clarity was always the goal, even when the tactics for getting there were clumsy. Specificity has always outperformed vagueness in both worlds. Expertise still shows, whether a machine is extracting a sentence or a person is scanning a page for the part that answers their question.

The overlap is bigger than the split, and what changed was emphasis more than values. Writing that was already clear and direct probably works fine in both systems. Writing that leaned on length and keyword density to get by is the writing that's now exposed.

What to Actually Do With This

The practical move is to write the clearest possible answer first and think about structure after that, rather than chasing a new checklist for AI citation or throwing out SEO altogether. Say the thing plainly, make the claim specific enough that it could be wrong, and don't make a reader, or a machine, dig through three paragraphs of throat-clearing to find out what you actually think.

That was probably good advice before any of this started. It just wasn't the advice that got rewarded. Now it is.

Have you noticed which of your own pages show up in AI answers, and whether they're the ones you expected?

Digital PR isn't dead, but most of what gets called digital PR now is just cold outreach with a nicer name

Steven Snell — Tue, 28 Jul 2026 15:16:28 +0000

An email landed in my inbox last week with the subject line "Digital PR opportunity for your site." I opened it because I'm curious about how these things are written now, the same way you might slow down for a fender bender.

It was three sentences. The sender loved my recent article (they didn't say which one). They had a piece on their client's blog that would be "a perfect resource" for my readers. Would I consider adding a link? That was it. No story, no angle, nothing a person would want if you stripped the link out of the request. Just an ask wearing a blazer.

I get some version of that email a few times a week now. Most of them use the phrase "digital PR" somewhere, and almost none of them are doing digital PR in any sense I'd recognize.

What the term used to buy you

Years ago I worked on a campaign for a client in the personal finance space. We had a genuinely interesting internal dataset about how people in different age brackets actually budgeted, not what they said they did in surveys. The whole job was turning that into something a journalist would want to write about on its own merits. We built the angle, cleaned up the numbers, wrote a tight summary, and pitched reporters who covered money and behavior.

The links came, but that was almost a side effect. The reason it worked is that we handed writers something they could use. A hook. A stat they hadn't seen. A reason for the piece to exist that had nothing to do with our client's SEO goals.

That's the version of digital PR I still believe in. It's slow, it's expensive, and it fails a lot. You spend real time on the story before you send a single email, and even then a lot of journalists ignore you. When it lands, though, you've earned something. A mention in a publication whose domain you could never buy your way onto, sitting next to a data point that was worth publishing.

How the label got hollowed out

Here's my honest read on what happened. "Link building" started to sound spammy, because a lot of it was. Google spent years telling everyone that buying and trading links was against the rules, and the phrase picked up a smell. So the industry needed a cleaner word for the same budget line, and "digital PR" was sitting right there, borrowed from an actual discipline that predates any of this.

The rebrand worked a little too well. Now the term covers everything from a genuine data-driven media campaign down to a mail-merged request to swap links, and everything in between calls itself the same thing. When a word stretches to cover that much ground, it stops meaning anything. The good practitioners get lumped in with the person sending forty identical emails an hour, and the person sending forty emails an hour gets to borrow the credibility of the good practitioners. Everybody loses except the second group.

The tell is whether there's a story

You can usually spot the difference in about four seconds, and it comes down to one question. Is there something here a journalist would actually want, or is the link the whole point?

Real digital PR leads with the thing worth covering. The pitch is built around a story, a survey nobody else ran, an angle on a trend that a reporter can turn into 800 words without doing your job for you. The link is the natural consequence of being useful, not the reason for the email.

Cold outreach dressed up as PR leads with the ask. You can feel it. There's a compliment that could apply to any site, a vague claim that some page is a "great resource," and then the request. Strip out the flattery and you're left with "please link to me." No story. Nothing a person would take if the link weren't attached.

A second tell is whether the email required any research. Did they read the thing they're complimenting? Could this exact message have gone to a thousand other sites with the URL swapped? If the answer is yes, it doesn't matter what they call it. It's a numbers game with a PR costume on.

Where I've landed on it

I'm not here to say digital PR is fake or dead. The good version is more effective than almost anything else I've seen for earning links that hold their value, precisely because it's hard and most people won't do the work. That's the whole point. The difficulty is the moat.

What frustrates me is watching a useful discipline get diluted by people who kept the label and skipped the labor. They figured out that "digital PR" opens more inboxes than "link exchange," and they were right, and now the term is worth a little less every month because of it.

So when I see "digital PR opportunity" in a subject line these days, I assume the worst until the email proves me wrong. Usually it doesn't. Occasionally one does, and those are the people I actually write back to.

If you do outreach for a living, I'm curious where you've landed. Do you still call it digital PR, or has the term gotten too muddy to be worth using? And if you're on the receiving end like me, what's the tell that makes you delete on sight?

The SEO advice of the past ten years doesn't work the way it used to

Steven Snell — Mon, 27 Jul 2026 21:02:32 +0000

I still remember the spreadsheet. Keyword, search volume, difficulty score. Pick the keyword, hit the target density, write the meta description, done. I built entire client strategies around that sheet for years, and it worked well enough that I never questioned it too hard.

I opened an old version of that spreadsheet last month while helping a client plan content for the fall. Half the columns didn't matter anymore. Not in a dramatic, the-sky-is-falling way. Just quietly, the way a tool stops getting used and you don't notice until you go looking for it.

The keyword still matters, the strategy around it doesn't

I'm not going to tell you keywords are dead. People still type words into search boxes, and Google still uses them to figure out what a page is about. That part hasn't changed.

What's changed is everything I used to do around the keyword. Density checks feel almost silly now. Nobody's counting how many times "best content management system for small business" appears on a page, least of all an AI model summarizing five sources into one paragraph. It's reading for meaning, not repetition. Stuff a keyword in five times and you're not helping your odds, you're just writing worse.

I used to treat the first 100 words as prime real estate for the primary keyword. These days I treat them as prime real estate for the actual answer. Those aren't always the same sentence anymore, and that distinction is doing a lot of work I didn't used to think about.

Backlinks matter less

For most of the last decade, if you'd asked me the single highest-leverage thing a freelancer could do for a client's SEO, I'd have said links without much hesitation. Get mentioned somewhere credible, get linked from a real site in your niche, and rankings would follow eventually.

Links still count for something. I'm not throwing that out. But I've watched pages with a thin backlink profile show up in AI-generated answers ahead of pages that would've dominated the old rankings, purely because the thin page answered the actual question faster and more clearly. That would've been almost unthinkable to me in 2018.

The thing that got more valuable, not less: being right

The skill that's mattered most through this shift isn't a new one. It's accuracy. Old-school SEO tolerated a certain amount of fluff, because a human skimming a page for the gist of it would forgive some padding on the way to the good part.

An AI system summarizing your page for someone else has no patience for that. If a claim is vague or unsupported, it either gets skipped or, worse, misrepresented in the summary. I've had a client's page get quoted by an AI answer tool with a stat that was technically in the article but stripped of the context that made it accurate. That's on us as writers as much as it's on the tool. If the sentence can't stand completely on its own, it's going to get pulled out and used that way regardless.

I've started writing every important claim as if it might get lifted out of the page entirely and shown to someone with zero surrounding context. That's a much higher bar than "does this read okay in the flow of the article," and it's changed how I structure paragraphs more than any algorithm update ever did.

What I've actually kept

Clear headers. Direct answers near the top instead of three paragraphs of throat-clearing before the point. Writing for a specific reader with a specific problem instead of trying to cover every angle of a topic in one piece. Those habits predate any of this AI search stuff by years, honestly. They were just good writing practices that happened to also be good SEO practices, and now they're good AEO practices too. The label changed. The underlying skill didn't.

What I dropped was the mechanical layer sitting on top of that: the density targets, the exact-match keyword phrasing, the strategy meetings about anchor text. That stuff feels like it belonged to a different internet.

I don't think anyone fully knows what's coming next here, and I'm skeptical of people who talk like they do. But if you're a writer wondering whether ten years of SEO knowledge is now useless, I don't think it is. I think about a third of it quietly expired and nobody sent the memo. The rest was just good writing wearing an SEO costume, and it turns out that part travels fine.