🥇 Run a UX Incentive Design Audit — UXID
You get a 0–100 score that shows exactly where your page loses momentum — vague CTAs, buried trust signals, missing reciprocity, and fake urgency. UXID extracts 15 deterministic signals from your HTML. No AI in the score itself. Run both SEO and AEO modes to see the gap.
Both is the default. SEO emphasises click-through and dwell incentives. AEO emphasises answerability and citation-worthiness.
Discover what UX incentive design is — and why it matters
UX incentive design is the practice of reducing perceived risk at the exact moment a visitor decides whether to act. A button that says "Submit" asks for effort; a button that says "Start my 14-day free trial" promises a specific outcome. The difference is not cosmetic — it changes what the visitor believes they are agreeing to.
The signals UXID measures fall into four groups. Core action checks whether the CTA describes the value (not the effort) and whether it is visible above the fold. Trust checks whether social proof sits next to the button (proximity matters more than volume) and whether guarantees or privacy statements address hesitation at the point of decision. Commitment checks whether the ask matches the visitor's level of interest and whether free value is offered before the sale. AEO checks whether AI engines can extract and cite the page — lists, sourced statistics, clear definitions, and structured data.
Every score is deterministic. The AI panel only suggests fixes; it never changes the number.
Fix these 5 signals first
- CTA specificity reduces perceived risk. Vague buttons ("Learn More") force the visitor to imagine what happens next, and they imagine a form, a sales call, or a mailing list. Specific copy ("Download the 50-Page Guide") closes that gap. Across hundreds of tests, specific CTAs outperform vague ones by 15–35% because the visitor no longer has to guess.
- First-person framing increases ownership. "Start my free trial" consistently beats "Start your free trial" in A/B tests. The mechanism is psychological: "my" puts the visitor in their own voice; "your" keeps it as something being asked of them. Replications land smaller, but the direction rarely changes.
- Value-first copy answers the question before it is asked. The first 150 words are not a place to describe the company. They are a place to state the outcome. "You get a 0–100 score" is a value statement. "We are a platform that..." is a feature statement. Value-first openings keep the visitor reading; feature-first openings lose them.
- Social proof must sit next to the CTA, not on a separate page. The Journal of Consumer Psychology found that proof timing matters more than volume. When a visitor is deciding whether to click, proof adjacent to the button reduces uncertainty in the same glance. Moving three testimonials next to the primary CTA lifted clicks by roughly 20% compared with parking them in a separate section.
- Urgency only works when the consequence is real. Fake countdown timers that reset on page reload are now regulated as dark patterns in multiple jurisdictions. Ethical urgency means the deadline is specific ("Sale ends March 15"), the inventory count is verifiable, and the consequence of waiting is material. If none of those are true, the score stays neutral — not a failure.
Read the UXID methodology FAQ
Is the score AI-generated?
No. The 15 signals are extracted from the fetched HTML via deterministic DOM/regex rules. The AI panel is optional and only suggests fixes — it never contributes to the score itself.
Why do CTA visibility and contrast look approximate?
Because we audit static HTML. True above-fold position and rendered contrast require a live browser. UXID flags the structural proxies and labels them clearly — it doesn't invent precision it doesn't have.
What's the difference between UXID and Quit Risk?
Quit Risk measures friction (what makes people leave). UXID measures incentives (what makes people act). They're complements — run both on the same URL to see the full picture.
Why does AEO mode weigh lists and sourced stats higher?
AI engines assemble answers from fragments. A list is already fragmented — discrete, parallel, liftable. Prose has to be taken apart first. In a 2026 analysis of 18 pages that AI assistants actually cited in a financial category, 89% contained a list. That was the only near-universal signal. Structured data appeared on 67% of cited pages, named authors on 39%, and FAQ schema on 0%. A list is a floor, not an edge: not having one makes citation harder, but having one does not guarantee citation. The differentiator sits in the shape of the content, not the presence of a single tag.
Why is reciprocity scored separately from social proof?
They solve different problems. Reciprocity — giving something of genuine value before asking for action — creates a psychological obligation that increases response rates. Social proof — showing that others have already acted — reduces uncertainty at the moment of decision. Cialdini's research on reciprocity found that restaurant servers who gave an unexpected second mint saw tips rise 23%. In marketing, free tools, ungated content, and actionable resources work the same way: they establish goodwill before the conversion request. The key is that the value must feel genuine, not transactional. A lead magnet that feels like bait triggers skepticism, not obligation.
Why does ethical urgency carry a neutral score when no urgency is detected?
Because most pages do not need urgency. A documentation page, a pricing page, or a blog post can be entirely honest and still score 60 on urgency because the detector found no countdown or deadline phrase. Scoring it as a failure would punish pages for not using a tactic that may not apply to them. The signal only moves above neutral when a specific, real deadline is present (85), or below neutral when an artificial pattern is detected (30).
See the 15 signals UXID measures
Each signal is auto-detected from the page HTML — no AI, no guesswork. Expand any card to read the full explanation, examples, and fixes.
Learn how the score is weighted (Both / SEO / AEO)
Both averages all 15 signals. AEO signals are weighted slightly lower since they only matter for AI citation channels.
SEO weights CTA specificity (×1.5), action-oriented headings (×1.4), and value-first framing (×1.4) higher. AEO signals drop to ×0.3–0.4.
AEO weights lists (×1.5), sourced stats (×1.5), definitional clarity (×1.4), FAQ structure (×1.3), and structured data (×1.2) higher. CTA signals drop to ×0.6–0.9.
Signals that are not detected on the page are excluded from the weighted average entirely — they don't count as failures. Switching modes re-weights instantly using already-extracted signals, without re-fetching the page.
Read the UXID methodology FAQ
What is UXID and how does it score?
UXID is a free UX incentive design audit. It extracts 15 deterministic signals from page HTML — CTA specificity, value-first framing, trust anchors, reciprocity, social proof proximity, ethical urgency, action-oriented headings, plus AEO signals — and combines them into a 0–100 score. Nothing in the score is AI-generated; the optional AI panel only suggests fixes.
Why are some signals 0 or 'not detected'?
Not every page has every signal. When a signal genuinely isn't present, the card shows 'not detected' rather than a fake zero — it is excluded from the weighted score, not treated as a failure.
What's the difference between UXID and Quit Risk?
Quit Risk measures friction (what makes people leave). UXID measures incentives (what makes people act). They're complements — run both on the same URL to see the full picture.
What is the difference between Both, SEO and AEO modes?
Both averages all 15 signals. SEO mode weights CTA specificity, action-oriented headings, and value-first framing higher — for organic click-through and dwell time. AEO mode weights lists, sourced stats, definitional clarity, FAQ structure, and structured data higher — for AI citation-worthiness. Changing mode re-weights instantly without refetching.
Why does the tool say 'Static HTML signals only'?
UXID analyses the fetched HTML, not a rendered browser session. Above-fold position and true rendered contrast can only be approximated from structural proxies. The tool labels this limitation clearly. For true rendered positioning, pair with Quit Risk or Lighthouse Plus.
Can I use UXID on any URL?
Any publicly accessible page — no login required. Some pages block automated fetches. If the tool says 'Empty or blocked response', try a different page.
Traffic Torch
User experience analysis and engagement optimization audit tools.
The UXID tool is built to identify whether a page provides a clear, credible reason for visitors to act. This is developed through analysis of real-world pages, combining modern engagement metrics, conversion psychology, and AEO citation signals used by AI search engines.
Ylia Callan – Lead Developer
Musician • Author • Researcher • Web Designer • TAFE NSW Australia • Creator of privacy-first UX design web tools. Bridging creativity, consciousness exploration, and modern technology to build meaningful, performant, and user-centric experiences.
Server Stacks
Traffic Torch runs on epic, scalable infrastructure: Hosted via GitHub for blazing-fast static delivery and effortless updates. Protected and accelerated globally by Cloudflare with edge CDN, security, and performance optimizations.
Backed by real-world expertise in UX research, engagement analytics, and search performance.
Published: 2026-10-20
Last updated: 2026-10-20
