Sample report

Sample Reality Check, run on GarageHub

We cannot show a client's report, so we ran the same process on GarageHub, the car maintenance app we build and run. Below is page one as a client receives it, in shortened form, plus two findings in full. Security details stay private, as in any client report.

Reality CheckGarageHub

Keep.

GarageHub is a sound system. The data model fits the business, deploys are frequent and rarely fail, and no area scored Critical. The work ahead is about people, access and upkeep, not code quality.

  • 56 findings, 15 of them strengths 23 Sep 2026
  • 31% day-2 return. We had been reading 17%. 23 Sep 2026
  • 35 to 76 senior days in the 90-day plan 23 Sep 2026
  • €50 to 60 a month for all infrastructure 23 Sep 2026

Reality Check · GarageHubPage one

Verdict: Keep.

Six areas

AreaWhat we looked at
Code and architectureStructure, tests, dependencies, the places a change is likely to break
Running costsEvery recurring charge, matched to the bank, and what grows with usage Infrastructure costs about €50 to 60 a month.
Keys and ownershipEvery account the business depends on, who controls it, how it transfers
Security and dataWhere personal data lives, access, secrets, backups and restores A 14-day grace period before an account’s data is deleted.
DeliveryHow code reaches production, how often, and how it rolls back 4,400+ API unit tests, 110 integration tests, about 60 CI checks.
Product and usageCore journeys, errors, usage numbers and what they really measure Day-2 return is 31%, not the 17% we had been reading.

How each area is scored

  1. Healthy A new developer could work here safely next week.
  2. Watch Known issues, nothing urgent, fix as you go.
  3. Risk Will cause real cost or an incident within months if left.
  4. Critical Can lose data, money, access or users soon. Fix first.

First 90 daysPlan

First 90 days: 35 to 76 senior days.

The 90-day plan in three phases, with the effort for each in senior days
WhenWhat goes thereSenior days
Now weeks 1 to 2Anything time-bound or high severity: dependency deadlines, access and account hygiene, release safety7 to 15
Next weeks 3 to 6Measurement gaps, a parked feature to retire, automated quality checks, policy and storage housekeeping10 to 25
Later weeks 7 to 12Onboarding documentation, retiring duplicated data, planned upgrades one per release, splitting the largest files18 to 36
Total35 to 76

Each bar runs from the low to the high estimate. A client report prices every action in euros and ranks each finding with its evidence. The full GarageHub report has 56 findings, 15 of them strengths.

Two findings in full

Each finding says what we saw, why it matters, what to do and what it takes.

F-01

Severity: Medium

The retention number was wrong, and the truth was better

What we saw
The product tracked day-2 return at 17%. Re-running the query gave 31%. A “last seen” field added in late July was empty for older users, so every earlier signup counted as never returning.
Why it matters
Roadmap decisions were made against a number half the real size. The fault was in the measurement, not the product.
What to do
Report only cohorts after the field was added, record the correction next to the metric, then work on the two thirds who do not return.
Effort
Half a day

B-02

Severity: Medium

€50 to 60

a month for all infrastructure, and falling

AI spend is the fastest-growing cost and needs a hard cap

What we saw
Infrastructure costs about €50 to 60 a month and is falling. The AI mechanic is usage-based. A daily budget per user exists in code, but there is no monthly limit or alert at the model providers.
Why it matters
A bug, runaway loop or abused account shows up on the invoice, not on a dashboard. Unit economics stay invisible until the month closes.
What to do
Set monthly spend limits and alerts at each model provider, and report AI cost per active user every month next to revenue per user.
Effort
Half a day to one day

What to leave alone

  • More than 4,400 API unit tests run in under a minute, plus 110 integration tests against a real database, covering payments and login.
  • About 60 automated CI checks enforce authorization, data ownership and compatibility with app versions already in the stores.
  • Encrypted off-site backups, restored automatically every week to prove they work.
  • Account deletion removes photos, payment customer data and analytics profiles after a 14-day grace period.
  • The data model fits what the business does now. No rebuild trigger.

What we did not check

  • No penetration test; secrets were searched with targeted patterns.
  • Journeys were read from the code, not clicked through on devices.
  • App Store and Play Console vitals were not pulled.
  • Usage figures were re-run once, on the day of the check.

Questions a buyer can ask a seller before any of this: the SaaS due diligence checklist.

Your own Reality Check.

3 to 5 working days, from €1,800, fixed before we start. Credited in full if you continue within 60 days.