Interesting tool, and the pain point is real. Anyone who has run a small print job for a client knows the drill: customer sends a 400px JPEG logo pulled from their Facebook page, the printer bounces it, and someone ends up manually redrawing it at £40 an hour. If this takes that step away cleanly, there is a market for it.
A few thoughts and questions from the UK print side of things, in the spirit of the feedback you asked for.
What UK printers will actually want Most UK commercial printers and sticker/label houses have fairly standard artwork specs, and your tool will live or die on whether the output drops straight into their workflow without a "please resupply" email. Worth checking the exported PDF handles:
- 3mm bleed as the default (that is the norm over here, with 2mm safe zone inside the trim)
- CMYK colour space, not RGB. A lot of AI image generators spit out RGB files with colours that simply cannot be printed. If your converter silently keeps RGB values, users will get muddy prints and blame the printer
- A separate spot colour or named layer for the cut line (often "CutContour" in magenta) rather than a visible stroke on the same layer. Vinyl cutters and print houses using Roland or Mimaki kit expect that convention
- PDF/X-1a or PDF/X-4 export options, since some trade printers reject anything else
- Fonts converted to outlines, which sounds obvious but trips up plenty of tools
If it already does all that, say so loudly in your Store listing, because that is the thing a print buyer is scanning for.
Questions about the conversion itself Raster-to-vector tracing has been around for years, and the honest truth is that auto-tracing works beautifully on clean flat-colour logos and poorly on anything with gradients, soft shadows or fine detail. AI-generated images tend to be full of exactly that stuff. So:
- How does it cope with gradients and photographic textures? Does it simplify, mesh, or just produce thousands of tiny shapes?
- Is there any control over node count or smoothing, or is it one-click only?
- Can the user preview the trace against the original before exporting?
A side-by-side example or two in this thread would do more for you than any description.
The bit nobody likes talking about: who owns an AI logo Since you are pitching this at AI-generated logos, it is worth flagging for anyone reading. In the UK, copyright in computer-generated works is a grey and currently contested area, and the terms of service of the image generator matter a great deal. More practically, if a client wants to register that logo as a trade mark with the Intellectual Property Office, a vectorised version of something a generator produced from a one-line prompt may not be distinctive enough, and there is a real chance a very similar mark already exists somewhere.
That is not your tool's problem to solve, but a short note in your documentation along the lines of "check the licence terms of whatever generated the image before printing it commercially" would be responsible and would save your users grief.
Business model questions You have said free with unlimited conversions. Fair enough as a launch strategy, but forum members here tend to ask the obvious: what happens later? If people build it into a workflow and then hit a paywall six months down the line, that burns goodwill. Being upfront now about whether a paid tier is planned (and roughly what would sit behind it) will get you more serious users, not fewer.
Also, Windows only via the Microsoft Store rules out a big chunk of the design and print community, which skews heavily towards Mac. If a Mac or web version is on the roadmap, mention it.
Feature suggestions - Presets for common UK sticker sizes and standard label formats (A-series and the usual round and rectangular die sizes)
- An option to add registration marks and crop marks on the PDF
- Batch processing, since small print shops often have a folder of twenty client logos to deal with, not one
- White underprint layer option for clear or metallic vinyl, which is a common request for product labels
- A "print readiness" warning panel: low source resolution, RGB colours, text too small to cut, that sort of thing
Best of luck with it. Post a couple of real before-and-after examples, including an ugly one that it struggled with, and you will get far more useful feedback from this crowd than a polished pitch ever would.
