svgenerator.

why build this?

i needed svg files for my websites — icons, backgrounds, that sort of thing. the options were either hunting through free repositories and never finding exactly what i wanted, or paying for stock vectors that still needed tweaking. so i built an ai-powered svg generator that takes a text prompt and gives you a clean vector file.

the tool itself is fairly straightforward. you describe what you want, pick a style (line art, engraving, hand drawn, pixel art, etc.), and it generates an svg you can download and use. there's also an image-to-svg converter for when you already have a raster image you need vectorised. it works, it's live at svgenerator.org, and that's about as much as i want to say about the product side of things.

the more interesting part of this project has been the marketing.

finding a niche.

when i started looking at the svg generator space, it became pretty clear pretty quickly that the general market is saturated. there are dozens of ai svg tools, vectorisers, and converters already out there, many of them backed by companies with real marketing budgets. competing head-on for search terms like "ai svg generator" or "png to svg" didn't seem like a great use of time.

so i started digging into more specific use cases for svgs beyond web development. and i found the cricut and laser cutter community.

cricut machines and laser cutters use svg files as their input format, and it turns out there's a massive community of people who craft with these machines — vinyl decals, t-shirts, paper crafts, wood engravings. the thing is, they're not developers. they're buying or downloading svg files from various marketplaces, and a huge number of those files are broken. duplicate paths that make the blade cut the same line twice, open paths that cause the blade to lift mid-cut, gradients that confuse the cutting software entirely.

i spent time on the subreddits and facebook groups and it was the same story over and over — people uploading svgs to cricut design space and getting mangled cuts, asking each other for help fixing files, sharing workarounds. there's genuine demand for a tool that cleans up svg files specifically for cutting machines, and nobody is really serving it well.

validating with search data.

rather than just going on gut feeling from reddit posts, i did some proper research using google search traffic analytics. i wanted to know what the actual search volume looks like for cricut-related svg problems, what terms people are using, and whether any existing services are capturing that traffic.

the search volume was there. people are googling things like "why does my svg cut wrong on cricut," "fix svg for cricut design space," "svg too many layers cricut," and dozens of variations. what was more interesting was that when i looked at who was actually ranking for those terms, there wasn't much. the big svg tools don't mention cricut or silhouette or laser cutting on their sites because they're targeting developers and designers, not crafters. the cricut-specific sites are mostly marketplaces selling svg files, not tools for fixing them.

that gap is what made me decide to go further with it.

the pieter levels approach.

i'd been following pieter levels and his approach to building products — particularly the idea of validating demand before building the full thing. rather than spending weeks building a complete svg cleaning tool and then hoping people want it, i put up a waitlist that looks and feels like a working product.

the page lets you upload an svg, hit "fix svg," and watch a progress bar. then it asks for your email to send you the fixed file. what's actually happening is it's collecting emails and asking what tools people use (cricut, silhouette, laser cutter, illustrator, inkscape). nobody gets a fixed file yet — it's a waitlist. but the experience feels close enough to a real product that the signups reflect genuine intent rather than idle curiosity.

this lives at waitlist.svgenerator.org and gives me two things: a list of people who actually want this tool, and data on what machines they're using so i can prioritise which integrations to build first.

seo as a weapon.

the other side of this has been using the search traffic research to inform the main site's content. when i analysed competitor sites, i found svg-related search terms that none of them were including in their pages. things that people are actively searching for but that no existing tool has indexed against.

i made sure to include those terms naturally across svgenerator.org — in the faq, in the page copy, in the metadata. the idea is simple: if someone searches for a phrase and no competitor has that phrase on their site, but mine does, the search engine has a much closer match to serve. it's not about gaming anything, it's just about being the only result that actually talks about what the person is looking for.

i also put together a competitor seo research document covering the entire svg tool space — traffic numbers, keyword volumes, content strategies, and where the gaps are. that research is what gave me confidence that the cricut niche wasn't just anecdotally promising but had real, measurable search demand behind it.

what i've learned so far.

this is the first project where i've really focused on the marketing side from the start rather than treating it as an afterthought. i've made artwork and adverts using ai for pinterest and facebook, built a waitlist landing page to validate demand before writing backend code, and spent more time in google search console and analytics tools than in my code editor.

the biggest lesson is that finding a niche matters more than building a better mousetrap. the svg generator market is crowded, but the svg-fixer-for-cricut-users market barely exists. and the people in it are actively looking for help.

what's next?

the waitlist is collecting data. once there's enough signal on what people need most, i'll build the actual svg cleaning tool behind it. the generator itself is already live and working for general svg creation, but the cricut-focused fixer is where the real opportunity seems to be.