Predictive content diagnostics

ThinkAloud report

https://web-tracking.eu/ · purpose: Conversion

13%
of the first glance on an action
5
findings in total · 1 convergent · worst: Safety
Heatmap of where the eye lands firstFirst glance

In short

About this test — task, personas, method

Task: To understand if this cookieless, EU-hosted analytics solution can meet my website's tracking needs and simplify data collection without a cookie banner.

Personas: GDPR-focused Small Business Owner · Skeptical Marketing Manager · Time-pressed Web Developer · Content Creator Seeking Engagement Insights

Models: GLM-5.3-Flash + Claude-Haiku (a finding confirmed by both counts more strongly)

Method: synthetic personas read the page in a real browser and thought aloud; an attention model predicted the first glance.

Field 1 — First glance

Where the eye lands

red = high predicted attention***

13% of the predicted first glance lands on an action · 87% on text/decoration. This page must trigger an action — so the share is worth tracking over time.

Does the eye land on the action — or on the decoration? Ranked by predicted attention:***

  • Add to your site free5%
  • See how it works1%
  • Try free1%
  • Web-Tracking.eu1%
  • Documentation1%
  • Pricing1%
Coverage

The four kinds of friction

Works = the personas praised the field · problem = friction · nothing noted = neither praise nor friction.

Safety
dare I?
2
problems
Meaning
do I get why?
2
problems
Capability
can I manage it?
1
problem
Effort
can I be bothered?
7
works

Plus 1 defect(s) — real bugs (dead button / broken link) outside the four fields.

Recommendations

Improvement opportunities

Most important first. “Trade-off” = the new friction the fix itself may create.

Add Safety

1. Publish concrete plan prices and add-on costs on the comparison

Behaviour it should unlock: Compare plans and start signup with a known amount and a clear add-on bill. Leaving prices hidden costs the owner the two personas who already wanted to buy and left rather than guess what they would pay.

Finding: no prices shown despite pricing claims. Field is correctly tryghed (cost/commitment uncertainty), not missing education: they abandon comparison because they will not choose a plan they cannot price. Put numbers next to every pricing claim so comparison can finish.

Trade-off — new friction the fix itself may create: Public prices need upkeep and can cause sticker-shock drop-off or competitor matching.

Addresses: No prices shown despite pricing claims

Change Defect

2. Point See how it works at a live page that explains tracking mechanics

Behaviour it should unlock: Open how the product actually works from the homepage and keep exploring to judge fit. A dead or purely promotional path ends the visit; the owner loses that user before any trial.

Findings: the See how it works link is dead, and the CTA shows marketing, not mechanics. The dead link is a true defekt (broken control), not kan — do not add hints around a 404. The content failure is mening: they never learn cookieless collection, EU hosting, or why no cookie banner, so they cannot judge fit. Same CTA, working destination, mechanics instead of slogans.

Trade-off — new friction the fix itself may create: A real mechanics page must be maintained; if it is long, it becomes new besvaer.

Addresses: See how it works link is dead, How-it-works CTA shows marketing, not mechanics

Change Defect

3. Send Get started to account creation, not the homepage

Behaviour it should unlock: Create an account and enter the tool from Get started in one click. The loop means users never start; the owner loses signups those visitors were ready to complete.

Finding: Get started loops back to the homepage. Labeled kan, but the mechanism is defekt: the control does not navigate. Extra copy or help will not work; retarget the button to signup or onboarding.

Trade-off — new friction the fix itself may create: A working signup CTA can increase incomplete or bot accounts unless a light gate exists.

Addresses: Get started loops back to the homepage

Add Safety

4. Spell out what consumes the AI Analytics quota

Behaviour it should unlock: Start or buy without fearing a surprise quota hit. Unclear limits stall purchase; the owner loses buyers who wanted AI insights but would not risk an unknown cap.

Finding: AI Analytics quota is unexplained. Labeled mening, but the mechanism is tryghed: they hesitate because they cannot tell whether the insights they want count against the quota (overage/commitment risk), not because AI analytics lacks meaning. State what burns quota, what does not, and what happens at the limit.

Trade-off — new friction the fix itself may create: Quota rules add copy to scan and may expose tight limits, causing some to defer.

Addresses: AI Analytics quota is unexplained

Can be dispensed with

Elements the personas' reactions show do not carry their weight. The recommendation is to remove them.

Field 2–4 — Think-aloud

Where recipients drop off

SafetyStart here

No prices shown despite pricing claims

2/4 personas · severity 2.5/5 · 1 page(s)

Users cannot tell what they would pay or what is an add-on, so they abandon plan comparison and leave without signing up.

  • Skeptical Marketing Manager 3/5 GLM-5.3-Flash
    „'Simple pricing. No surprises' headline, but the tiers show only event limits — no actual monthly prices anywhere visible, so I can't evaluate the real cost"
  • Content Creator Seeking Engagement Insights 2/5 GLM-5.3-Flash
    „'Optimise is an add-on on top of the analytics' — as a novice I can't tell if the engagement features I want are included or cost extra, and no price is visible here."
/ · step 5

Divergence — only one persona

Expert lens — methodical review

Expert friction review

Methodical expert review of the page text and the personas' findings — not a persona simulation. Findings without persona confirmation are marked as hypotheses.

Workflows & repeated useEffortConfirmed by persona findings: Get started loops back to the homepage, See how it works link is dead

Primary CTAs waste a click every visit

Anyone who comes to evaluate or buy hits Get started or See how it works and does not advance. That failed click repeats every session and never reaches install, a dashboard, or a second use. Repeated dead-end clicks cost more than a weak onboarding.

Behaviour it blocks: Blocks the evaluate-then-sign-up loop on every session; the owner pays in repeat bounce instead of activation.

Decision points (RACE)SafetyConfirmed by persona findings: No prices shown despite pricing claims

Leaving beats plan choice when price is hidden

The page claims pricing but shows no amounts or add-on costs. Under purchase pressure the cheap move is to close the tab, not guess spend. The desired action (pick a plan) is not the cheapest action.

Behaviour it blocks: Blocks plan comparison and paid signup; the owner loses buyers who already had intent.

Decision points (RACE)MeaningConfirmed by persona findings: How-it-works CTA shows marketing, not mechanics, See how it works link is dead

How-it-works loses to closing the tab

The control that should explain cookieless collection delivers marketing or nothing. Users cannot judge fit against a cookie-banner stack, so bounce wins. The wanted action is more expensive than leaving.

Behaviour it blocks: Blocks the does-this-replace-my-tracker decision; the owner loses the evaluators this page is for.

System feedback loopsMeaningConfirmed by persona findings: AI Analytics quota is unexplained

AI Analytics quota fails the so-what test

A quota is shown with no rule for what consumes it and no benchmark. If the user does nothing, the number still does not imply a next step. The figure is meaning-friction, not guidance.

Behaviour it blocks: Blocks trial and purchase of AI insights; the owner loses upgrades that stall on unclear risk.

What's missingCapabilityConfirmed by persona findings: Get started loops back to the homepage

No working entry to signup or the product

Users expect Get started to open an account or snippet. That destination does not function, and no other start-here path is available. The conversion flow has no terminal action.

Behaviour it blocks: Blocks completing signup; ready buyers cannot create an account.

What's missingMeaningConfirmed by persona findings: How-it-works CTA shows marketing, not mechanics

No proof of cookieless, no-banner tracking

The job is to confirm EU-hosted cookieless analytics can drop a cookie banner. There is no mechanic, event list, or cue that collection works without a consent UI. Fit cannot be verified.

Behaviour it blocks: Blocks the core fit check; the owner loses sites shopping specifically to remove banners.

Implementation frictionDefectConfirmed by persona findings: Get started loops back to the homepage

Get started is a homepage loop

The main conversion control returns users to the same page. Nothing was submitted, so there is no undo and no recovery into a real signup. A self-loop is a dead end, not a start.

Behaviour it blocks: Blocks account creation at the last click; CTA intent is discarded.

Implementation frictionDefectConfirmed by persona findings: See how it works link is dead

See how it works is a dead end

The explanation link does not open. A click that promises a path and fails is more expensive than scrolling. Exploration stops on the homepage.

Behaviour it blocks: Blocks product understanding; visitors leave instead of continuing.

Remove before you add — verdict per element

The expert's mandatory verdict on every significant element: does it carry its weight? "Remove" means remove — not improve.

ElementVerdictWhyTrade-off
Get started KeepIt is the only named conversion control; removing it would hide the action this flow exists to complete.—
See how it works RemoveA dead or marketing-only link spends a click and never explains the product, so it does not earn its place.Evaluators lose a labeled education path; do not refill the slot with campaign copy.
pricing claims Cut downPlan talk without amounts or add-ons burns trust and does not help a buy decision.Sender loses vague we-have-plans social proof; that is cheaper than the current abandon.
AI Analytics quota Cut downA cap with no counting rule or benchmark adds hesitation without changing the next step.Softens quota as an upgrade lever for the sender.
Strengths

What worked

works · MeaningConfirmed by both models

Headline instantly matches core needs

4/4 personas

The homepage headline and value proposition immediately signal cookieless, EU-hosted analytics with no banner and no credit card, so the product feels made for the visitor.

  • GDPR-focused Small Business Owner GLM-5.3-Flash
    „Headline and subtext immediately confirm the three things I care about: no cookie banner, EU data residency, no credit card to try it"
  • GDPR-focused Small Business Owner Claude-Haiku
    „Headline immediately signals the solution matches my need: cookieless, no banner, EU-based. Clear value proposition."
  • 4 more quotes
    • GDPR-focused Small Business Owner Claude-Haiku
      „Headline and problem statements directly address my concerns: GDPR risk with Google Analytics, data leaving Europe, consent banner filtering. Feels relevant and trustworthy."
    • GDPR-focused Small Business Owner Claude-Haiku
      „Clear value proposition addressing all my pain points: GDPR compliance, no cookie banner friction, EU hosting, transparent pricing with free tier, and emphasis on simplicity ('takes 3 minutes'). Builds confidence."
    • Skeptical Marketing Manager Claude-Haiku
      „Headline directly addresses my core problem: GA data gaps due to consent banners. Value prop is clear and specific to EU context."
    • Time-pressed Web Developer GLM-5.3-Flash
      „The headline instantly tells me the core value — no cookie banner — and 'Built in Denmark. Hosted in Germany' pre-answers my EU-compliance question in one line."
works · MeaningConfirmed by both models

Pain points mirror users' real problems

4/4 personas

Named objections around consent-banner data loss, GDPR/US-transfer risk, and GA complexity make the product feel personally relevant rather than generic.

  • GDPR-focused Small Business Owner GLM-5.3-Flash
    „Naming the US-server GDPR risk and the consent-banner data loss shows they understand my exact legal and data-accuracy problem"
  • Skeptical Marketing Manager GLM-5.3-Flash
    „The pain points mirror my exact situation — consent banner data gaps and the US data transfer risk — so the product immediately feels relevant rather than generic marketing."
  • 4 more quotes
    • Skeptical Marketing Manager Claude-Haiku
      „The three pain points directly mirror my concerns as a skeptical marketing manager — consent filtering, GDPR risk, and GA complexity. Feels like they understand my actual problems, not just selling a generic alternative."
    • Time-pressed Web Developer GLM-5.3-Flash
      „The pain points are exactly my clients' objections — shows they understand the developer/agency use case"
    • Time-pressed Web Developer Claude-Haiku
      „Clear problem statement (GDPR risks, data leaving EU, complexity) directly addresses my compliance concerns as a developer"
    • Content Creator Seeking Engagement Insights GLM-5.3-Flash
      „The pain points speak my language — 'you just want to know how many visitors you have' nails exactly why GA overwhelms me; feels like they actually get a novice content creator"
works · EffortConfirmed by both models

Risk-free setup in minutes

4/4 personas

No password, no credit card, a free plan, and a one-line script that takes a few minutes removes every excuse not to try it.

  • GDPR-focused Small Business Owner GLM-5.3-Flash
    „'No password, no credit card' and 'Set up in 3 minutes' — removes every excuse and makes trying it feel risk-free"
  • GDPR-focused Small Business Owner Claude-Haiku
    „The 3-minute setup (no password, no credit card, one script line) directly removes friction for my concern about implementation complexity."
  • 4 more quotes
    • GDPR-focused Small Business Owner Claude-Haiku /auth/login
      „Clear, friction-free signup: no credit card required, free plan, Google option available, pre-filled test email ready to go"
    • Skeptical Marketing Manager GLM-5.3-Flash
      „The 3-minute setup with no password, no credit card, and a single script line removes the usual trial friction — and page change tracking with 14-day before/after speaks directly to a real GA workflow pain."
    • Time-pressed Web Developer GLM-5.3-Flash /auth/login
      „'Free plan. No credit card. Takes 60 seconds.' — instantly removes signup risk and effort concerns"
    • Time-pressed Web Developer Claude-Haiku
      „Setup process is straightforward and minimal: email-only signup, no password/card required, one-line script paste. Matches my need for easy integration."
works · SafetyConfirmed by both models

Concrete EU and cookieless compliance claims

3/4 personas

Specific statements about no cookies, Germany hosting, Schrems II, and data never leaving the EU confirm the legal compliance users came to verify.

  • GDPR-focused Small Business Owner GLM-5.3-Flash
    „'No cookies needed — no consent banner required' and '100% EU hosted in Germany, no Schrems II issues' directly confirm the legal compliance I came to verify, and 'one script tag, ~3 minutes' confirms it's easy to implement"
  • GDPR-focused Small Business Owner Claude-Haiku
    „Clear, direct answers to my main concerns: GDPR compliance (no cookies, EU-hosted, Schrems II compliant), simple implementation (one script tag), and optional GA4 forwarding gives me control."
  • 3 more quotes
    • Skeptical Marketing Manager GLM-5.3-Flash
      „They openly disclose that optional GA4 forwarding sends data to Google (US) while their own data stays in the EU — honest, concrete hosting details (Germany, Schrems II)."
    • Time-pressed Web Developer GLM-5.3-Flash
      „'No cookies needed — no consent banner required' concretely confirms the cookieless promise, and 'Add to your site — takes 3 minutes' signals a lightweight integration"
    • Time-pressed Web Developer GLM-5.3-Flash
      „'Your data never leaves the EU' plus built in Denmark / hosted in Germany — konkrete hosting-fakta, som er det præcise compliance-svar, jeg kan videregive til min kunde"
works · Safety

Honest competitor comparisons

3/4 personas

Tables and copy that admit weaknesses and even recommend rivals for some use cases build credibility with skeptical buyers.

  • Skeptical Marketing Manager GLM-5.3-Flash /pricing
    „'Considering another tool?' section promising an honest comparison including 'where they beat us' — admitting weaknesses is a strong credibility signal for a skeptical buyer like me"
  • Skeptical Marketing Manager GLM-5.3-Flash /compare
    „The table openly admits weaknesses ('including where we lose') and marks competitors' partial capabilities honestly rather than trashing everything — that builds credibility with me as a skeptical buyer"
  • 3 more quotes
    • Skeptical Marketing Manager GLM-5.3-Flash /compare
      „The 'Honest take' section explicitly recommends competitors (including GA4) for use cases where they fit better — that directly addresses my wariness of over-hyped solutions and builds real trust. Plus the €5/mo entry price finally gives me a concrete cost anchor."
    • Time-pressed Web Developer GLM-5.3-Flash /compare
      „'Honest take: when to pick what'-afsnittet angiver åbent, hvornår konkurrenterne er det bedre valg — det gør sammenligningen troværdig i stedet for at virke som markedsføring"
    • Content Creator Seeking Engagement Insights GLM-5.3-Flash /pricing
      „The 'Considering another tool?' section openly admits where competitors beat them — that kind of honesty makes me trust the whole pricing page more"
works · SafetyConfirmed by both models

Transparent free-tier pricing

2/4 personas

A clearly listed free tier with all core features and simple paid progression answers cost questions and signals no hidden fees.

  • GDPR-focused Small Business Owner GLM-5.3-Flash
    „Free tier with 10,000 events/month and 1 website directly answers my 'what does a small site cost' question — the answer is zero."
  • GDPR-focused Small Business Owner Claude-Haiku
    „Pricing is transparent and simple — free tier available, clear tier progression, no hidden costs mentioned"
  • 1 more quotes
    • Content Creator Seeking Engagement Insights GLM-5.3-Flash
      „'Simple pricing. No surprises' with a free tier listing 'All core features' directly answers my worry that the engagement features might be a paid add-on"
works · CapabilityConfirmed by both models

Obvious next-step paths

2/4 personas

A clearly labelled path to how it works and a direct signup flow with no extra steps make it obvious how to proceed.

  • GDPR-focused Small Business Owner Claude-Haiku /auth/login
    „Clear, direct signup flow matching my goal — no unnecessary steps, free plan confirmed, no credit card required"
  • Content Creator Seeking Engagement Insights GLM-5.3-Flash
    „'See how it works' is right there in the hero — a clearly labelled path to exactly the explanation I came for, no hunting required"
works · Safety

FAQ kills billing and legal anxiety

2/4 personas

Specific answers on ePrivacy, hard event-limit stops with no overages, prorating, and a DPA make committing feel safe and predictable.

  • Skeptical Marketing Manager GLM-5.3-Flash /pricing
    „The FAQ answers my exact objection with specifics: cookieless justified via ePrivacy Directive Article 5(3), and hitting the event limit just pauses recording with no overage charges — credible, not marketing fluff"
  • Skeptical Marketing Manager GLM-5.3-Flash /pricing
    „Event limit FAQ promises a hard stop with no overage charges and keeps dashboard/historical data accessible — predictable billing with no gotchas"
  • 2 more quotes
    • Content Creator Seeking Engagement Insights GLM-5.3-Flash /pricing
      „The FAQ directly answers practical worries — hitting the event limit just pauses recording with no overage charges, and I can switch plans anytime with automatic prorating. That makes choosing a plan feel low-risk."
    • Content Creator Seeking Engagement Insights GLM-5.3-Flash /pricing
      „The FAQ kills my billing anxiety — 'no overage charges', instant plan switching, and EU-only data storage with a DPA on request all make it feel safe to commit"
works · Effort

Step-by-step migration guides

2/4 personas

Per-competitor switching guides, including Google Analytics, directly address the effort of leaving an existing tool.

  • Skeptical Marketing Manager GLM-5.3-Flash /compare
    „Per-tool step-by-step migration guides, including one specifically for Google Analytics — directly addresses my switching-effort concern instead of leaving me to figure it out alone"
  • Time-pressed Web Developer GLM-5.3-Flash /compare
    „Step-by-step migration guides for each competitor directly address the switching/maintenance concern"

Isolated praise

Raw data

Full transcript

Everything the personas wrote, turn by turn — the summaries above are clustered out of exactly these words. Open a persona to read all of it.

GDPR-focused Small Business Owner · GLM-5.3-Flash · 5 steps · done
GDPR-focused Small Business Owner · Claude-Haiku · 9 steps · done
Skeptical Marketing Manager · GLM-5.3-Flash · 12 steps · step-budget
Skeptical Marketing Manager · Claude-Haiku · 12 steps · step-budget
Time-pressed Web Developer · GLM-5.3-Flash · 11 steps · done
Time-pressed Web Developer · Claude-Haiku · 12 steps · step-budget
Content Creator Seeking Engagement Insights · GLM-5.3-Flash · 12 steps · step-budget
Content Creator Seeking Engagement Insights · Claude-Haiku · 12 steps · step-budget
Notes

* First glance (~3 sec): where the eye instinctively lands. Says nothing about comprehension — the think-aloud part measures that.

** Note — the two instruments are coupled here. The ranking next to the heatmap was given to the personas as their first impression, so they sense the page in human order rather than the page's code order. This means that when the think-aloud part mentions one of the elements from that ranking, it is not independent confirmation — the persona was told. Only the first turn is guided by that ranking; everything after it is the persona's own.

*** There is no universal target for how much of the first glance should land on an action — compare with your own page before and after a change.