The model
How € estimates work
Every leak is a monthly range on one revenue baseline. The number is a hypothesis the buying committee can argue with, not a forecast you guarantee.
Why recovered is not 100% of the leak
Modeled leak is the gap we can see. Year-1 recovered is that gap times twelve, times a 65% recovery assumption. Nobody closes a modeled leak in full: traffic mix changes, the client ships part of the spec, and the rest stays on the table.
65% is a conservative default you can defend in a CFO review. On the home example: −€920/mo leak → €7,176/yr recovered, not €11,040.
Formula: recovered_per_year = leak_monthly × 12 × 0.65. The leak model itself is unchanged.
One monthly revenue figure
There is no default monthly revenue on a client report. You type last month when you add the site. Changing that figure rescales the leak. Agency retainer defaults stay off the live share. If monthly revenue is missing, the report says so instead of pretending €10,000 is theirs.
What gets priced
Funnel, tracking, paid, CRO, and email. Each critical or warning check carries an estimated €/mo and an action item (one change, one metric) inside the AuditFlow report. Email sequences then execute in ImpactFlow. SEO visibility is a FixFlow handoff, not an AuditFlow leak line. AuditFlow specifies. It does not operate paid or the site.
Conservative ranges
- Performance uses a Core Web Vitals conversion heuristic (about 1% conversion per 100ms), then a 0.7x to 1.3x band.
- Checkout and tracking gaps use sector friction vs the steps and pixels we detect, not last-click attribution.
- The live report CFO view shows the range, the baseline, and the assumption list. Print that view. Do not hide it.
Checkout steps (model target: 2)
Value used: 2-step checkout as the priced target, not a claimed industry mean.
Source: Baymard Institute checkout-usability research (public summaries, 2023–2024). Best-in-class stores keep checkout as short as the payment actually requires. AuditFlow prices extra steps against that target.
Applies to: ecommerce and checkout-led funnels. Lead-gen sites use form-field friction instead.
Excess abandonment
Value used: about 9% incremental abandonment per extra checkout step, so two extra steps ≈ 18% above the 2-step target.
Source: AuditFlow model, Baymard-informed step-drop research (simplified). Dated with this model version, 2026. Ecommerce checkout only.
Modeled CAC (€25)
Value used: €25 per lead when the consultant has not entered a client CAC.
Source: conservative SMB paid-social lead-cost band (WordStream / industry CPC-to-lead summaries, 2024), rounded down so the leak is harder to inflate. Replace it with the client's real CAC when you have it.
Applies to: mixed SMB ecommerce and lead-gen until a client figure is typed.
Conversion levers
Tracking, CTA, and proof gaps are priced as a projected % of the monthly revenue you enter — not as last-click revenue. Missing pixels make retargeting spend unattributable; the leak is the unmeasured conversion path, not a claimed ROAS.
Example: client-shop.io
Same figures as the home teaser and the demo dashboard. Hypotheses, not a client result.
- Baseline monthly site revenue: €10,000 entered by the consultant. Leak € is this monthly revenue times the projected impact.
- Tracking gaps modeled from missing Meta conversion events on key funnel steps.
- Checkout friction estimated vs a 2-step model target (4 steps detected). See /how-estimates#checkout-steps.
- Email capture gap: blog traffic × typical subscribe rate × modeled €25 CAC. See /how-estimates#cac.
- Year-1 recovered is leak × 12 × 65%. Not 100% of the leak. See /how-estimates#recovery-rate.
- Ranges are conservative. Actual impact varies by traffic mix and offer value.
What this is not
Not last month's actual lost revenue. Not a guarantee. Not a timesheet. The issue checklist is before → after € on later audits. Three close options (sprint, 90-day, watch) are priced on the range, on Pro. Starter quotes the sprint only.
See the full motion on the demo dashboard.