Skip to content

Conversion intelligence for agencies

Find the conversion gapsthat help youwin the client.

GapStudy analyses a business, shows the evidence behind each problem, and turns those gaps into proposal angles you can actually use.

Run a Free Gap Study→

No card required · No client login · Evidence included

Study consolePublic-site analysis

Up to five public pages. No client login. No form submissions.

Compare a public competitor (optional)

Ready for a real study

Page copy, headings, action links, forms and trust references become a traceable evidence record.

PUBLIC EVIDENCEPage HTML & copyCTA destinationsForms & labelsTrust referencesSelected competitors
01Example output

Every finding arrives with the proof attached.

No vague scores and no unexplained recommendations. Every finding shows what was observed, where it was found and how the agency can use it.

Fictional example reportNorthbeam Legal
Study EX-2026-014 · captured 14 Mar 2026 (example)
Finding GS-01

Pricing requires a sales conversation before visitors are given any indication of cost.

Severity
High
Confidence
87 / 100 · example score
Surface
Pricing page
Evidence type
Visible page structure
Page reference
/pricing
Observation

All pricing actions lead to the same sales enquiry.

Modelled

Conversion impact is modelled until first-party analytics are connected.

Evidence chain

The pricing page presents four service tiers. Every tier resolves to the same “Book a call” action. No price, range or starting figure is visible anywhere on the page.

Preview mode — demonstration output. No live scan was run. Commercial implications are qualitative; no revenue or conversion outcome is measured.

02How it works

From a prospect URL to a stronger proposal.

Four steps. The product does the inspection; you do the persuasion.

01

Enter the business

Add a public website and an optional competitor. Choose the business you want to understand.

02

Capture the evidence

Inspect bounded public pages, headings, links, forms and proof references.

03

Rank the gaps

Review findings ordered by severity, with explainable confidence and visible limitations.

04

Build the pitch

Copy the recommended fix, shape the scope and share an evidence-linked report.

Direct observations and modelled conclusions are displayed differently, so you always know what is proven and what is inferred.

Observed Modelled
03What it analyses

Six surfaces. One commercial question.

Where is the customer journey losing confidence, clarity or momentum — and can the agency prove it?

01

Message and Positioning

  • Page titles
  • Heading structure
  • Offer copy for manual review
  • Action labels and destinations
02

Trust and Evidence

  • Testimonial references
  • Case-study links
  • Review references
  • Logo alt-text references
03

Offer and Pricing

  • Visible pricing text
  • Package references
  • Guarantee language
  • Published offer information
04

Conversion Path

  • Contact and booking links
  • Form-field counts
  • Required fields
  • Form-label structure
05

Technical Experience

  • Reachable public pages
  • Response status and timing
  • Basic image and form labels
  • Viewport metadata, not rendered layout
06

Competitive Position

  • User-selected public sites
  • Captured action labels
  • Proof-reference comparison
  • No inferred traffic or performance

Checks use the captured public HTML. Browser-rendered behaviour, private analytics and unavailable pages remain outside the study.

04Why it's different

Evidence first. AI second.

A language model without sources produces opinions. GapStudy is built around the standard, not the model.

Generic AI auditno sources
  • ✕Recommendations without page references
  • ✕Unclear separation of fact and interpretation
  • ✕Unexplained confidence
  • ✕One-size-fits-all next steps
GapStudyevery claim cited
  • ✓Source-linked observations
  • ✓Observed evidence and modelled implications
  • ✓Confidence calculated from coverage
  • ✓Fixes and proposal angles tied to findings

“If the evidence does not support a claim, GapStudy should not present it as fact.”

The operating rule
05Example report

The report you take into the room.

A complete, structured document — not a chat transcript. This preview shows the real structure, with fictional demonstration data.

Fictional demonstration data
Report contentsEvidence report
01Executive summary01
02Prioritised findings02
03Evidence references03
04Captured HTML structure04
05Confidence and limitations05
06Selected competitors06
07Recommended actions07
08Proposal angles08
09Suggested scope09
107-day action plan10
Example page structurefictional
/pricing

4 tiers · 0 visible prices · 1 action: “Book a call”

Studies store extracted text and HTML structure with source references. Screenshots and rendered viewport captures are not included.

Confidence & limitations
Observed6 of 8 refs
Modelledlabelled in-place
Unavailablenamed, not hidden

The limitations section states exactly what the study could not measure — private analytics, actual conversion rates, customer behaviour.

Run a Study on a Prospect→
06Use cases

Built for people who have to justify the fee.

The report is the opener in a pitch, the diagnostic in a retainer, and the business case internally.

Agencies01

Start the client conversation with a specific observation from the prospect's own website.

Freelancers02

Create a focused starting audit before committing hours to a full proposal.

Consultants03

Support recommendations with traceable evidence and a clear boundary around what is inferred.

Existing client work04

Review a business journey and agree the next practical change with your client.

No success rates, no invented outcomes. The claims on this page are about what the tool does, and each one is demonstrable in the example report.

07Pricing

Start with one real prospect.

One free study today. Join the waitlist for the paid workspace that fits your practice.

Free

One real prospect

£0

1 Gap Study

  • Core journey analysis
  • Evidence references
  • Limited proposal angles
  • Example report format
  • No card required
Run One Free

Practice

Freelancers and independent consultants

£79 / month

15 studies / month

Planned paid access. Not available to purchase.

  • Full findings and evidence
  • Proposal angle library
  • PDF export
  • Competitor comparison
  • 1 workspace

Agency

Agencies and studios with an active pitch pipeline

£199 / month

60 studies / month

Planned paid access. Not available to purchase.

  • White-label reports
  • Shareable client links
  • Bulk prospect queue
  • Team collaboration
  • Re-scan and change tracking
  • CSV export
  • Up to 5 team members

No payment is taken. Paid-plan exports, bulk queues, team features and recurring allowances are planned and will be confirmed before billing opens.

08Methodology & limitations

Clear about what is observed, modelled and unavailable.

Three boxes, no ambiguity. Every output from GapStudy fits in exactly one of them — and says which.

Observed

Captured public HTML: page copy, links, forms, headings and accessibility attributes. A timestamp and source identify every evidence record.

Modelled

Qualitative commercial implications and recommended changes. These are starting points for a discussion, not measured customer behaviour.

Unavailable

Private analytics, actual conversion rates, revenue, customer behaviour, rendered mobile tests and browser performance measurements.

The rule

“Never present a modelled conclusion as measured customer behaviour.”

The evidence standard

Six rules, applied to every study, including the free one. When a rule cannot be satisfied, the affected claim is withheld and the limitation is named in the report.

  1. 01Every finding cites stored page evidence.
  2. 02Capture times and source URLs remain visible.
  3. 03Confidence is calculated from evidence coverage and source agreement.
  4. 04Unsupported claims are withheld.
  5. 05Competitor observations name the selected source.
  6. 06Each recommendation includes its important limitation.
09Trust

Trust built into every study.

Clear evidence references, visible limitations and control over the studies stored in your workspace.

Private until you create a share link.

Evidence and limitations alongside findings.

Delete a study and revoke its shared report.

10Questions

Asked before anyone trusts it. Answered directly.

Mostly about the evidence standard. Which is the right thing to ask.

A study first captures public HTML, then applies bounded evidence rules. Each finding points to a stored page reference. Commercial implications are kept separate from observations. The current scanner does not use a language model to invent findings.

Start a study

Your next proposal should begin with evidence.

Run a Gap Study on a prospect you genuinely want to win. See the gaps, inspect the proof and turn the strongest finding into a credible reason to start the conversation.

Run a Free Gap Study→

No card required · One free study · Evidence and limitations included