What We Do
Webappski is an AEO studio in Gdynia, Poland: we make your product the answer ChatGPT, Perplexity, Gemini, and Claude give buyers in your category. Work starts with a measured visibility audit, then continues as a done-with-you plan or a monthly retainer — the same playbook we run on our own products, TypelessForm and Typelessity.
AEO — AI Search Visibility
We make your product appear in ChatGPT, Perplexity and Gemini answers. We run the method on our own product first and publish every dated measurement — the drops as well as the peaks.
Featured in AI Answers
Webappski publishes its own AI-visibility numbers instead of describing them — every dated run, drops included. On the 2 September 2026 run our product TypelessForm was present in 10 of 12 answer cells (Visibility Index 83/100). It read 12 of 12 on 11 July and 6 of 12 on 1 September. No single point is a claim about today.
Audited track record
We publish every run of our own AI visibility — including the ones that go down
Webappski re-measures TypelessForm every month against a fixed basket of 12 buyer questions across ChatGPT, Gemini, Claude and Perplexity. Every point here is a real aeo-platform run made with the open-source tool we sell — open any point and read the dated report behind it, including the ones where we lost ground.
Last measured run, 2026-09-02: present in 10 of 12 AI answers
This curve is not monotonic and is not a standing state — read every point with its date. The left end is also not strictly comparable to the right: the question set only settled into its current three per engine by 11 Jun, and the engines shifted their own models underneath us over that span (Gemini 2.5-pro → 3.5-flash in May, Claude Sonnet 4.6 → 5 between the June and July runs, Gemini 3.5 → 3.7-flash before September). The 100 on 11 Jul was a real full-basket run; presence then went 92 on 13 Aug, 50 on 1 Sep and 83 on 2 Sep, and we plot all of it. An earlier 10 Jun run also touched 100, but on a narrower 3-engine grid (no Perplexity) over 9 questions — we don't plot that one; on the full 4-engine basket it recomputes honestly to 83. Measured on each engine's API surface via our own keys — a reproducible proxy, not a guarantee of what the consumer apps show a person; Google AI Overviews, AI Mode and Microsoft Copilot are not in the basket.
Every cell is a real aeo-platform run (cited queries of 3). Claude sat at 0 until June and has been named in all three ever since; ChatGPT went to 0 on 1 Sep and back to 3 the next day; Perplexity and Gemini swing run-to-run — that volatility is the honest picture AI engines give. New runs add a column on the right; each links to its own hosted report.
Built and Promoted by Us
Webappski builds its own AI products and makes them visible in AI search. TypelessForm (one-shot voice form filling, an npm-published widget live since March 2026) and Typelessity (a conversational booking widget) run on the same AEO playbook we sell to clients — and every number we publish for them is measured, not estimated.
Our Process
Webappski runs AEO in five tracked steps, each tied to a measurable number. First a baseline visibility audit across all four major AI engines, then gap analysis against named competitors, structural content fixes (answer capsules, FAQ schema, crawler access, server-side rendering), authority signal building, and monthly monitoring published openly.
Audit
We check your visibility across ChatGPT, Perplexity, Gemini, and Claude. You get a scored report showing where you stand.
Strategy
Custom optimization plan for your product and niche. We identify the highest-impact opportunities.
Execution
We implement changes across your site and the wider web. You see progress in a monthly visibility re-measure.
Monitor
Monthly reports showing citation growth, AI visibility trends, and competitive movement.
Latest from the blog
The Webappski blog documents how a product gets cited by AI: measured visibility runs, what actually moved the numbers, and what failed. Posts are written from our own aeo-platform data — the TypelessForm case, teardowns of free visibility checkers, and notes from the playbook we run for clients.
How Do You Know a Change in Your AI Visibility Score Is Real? Our Own Report Called a 42-Point Fall It Could Not Support
On 1 September 2026 the tracker we build reported a 42-point fall for our own product and listed the rivals apparently standing in its place. None of them had displaced anything, and about 17 of the 42 points are demonstrably the instrument — the rest is unattributed. This is the reading discipline that should have caught it, the guards that exist, and the axis that had none.
Sep 2, 2026
Read More →
Do ChatGPT, Gemini and Claude Cite the Same Sources? 616 Hosts, 14 Shared by All Three
The follow-up our four-language snapshot promised, narrowed to one language: 267 saved Polish-language answers, 39 identical questions per engine, and the first source-overlap metric we have published — with the coverage asymmetry that partly explains it.
Sep 1, 2026
Read More →
Why Did My AI Visibility Score Drop? We Checked 18 AI Visibility Tools — Most Only Show That It Did
Tracking AI visibility across dated runs stopped being rare in 2026: 17 of the 18 tools we checked in August do it, and at least three of them do it free. Explaining which question and which engine caused the move is still rare.
Aug 26, 2026
Read More →