The Five Hours Back Assessment · owner-led service teams of 2 to 20
Find five team-hours your business can get back every week - or get your $999 back.
One paid assessment, $999all-in. I take your sanitized process evidence, put every claimed hour through an overlap check, and hand back a verdict you can defend: five validated team-hours a week trapped in repeated work, and the one workflow worth fixing first. No dashboard, no pilot program, no “AI strategy” deck.
$999 USD. Prices in US dollars.
The clock does not start when you pay. It starts when the inputs are real. Three things happen first, in order:
You pay the $999 and your reserved seat becomes yours. You get the sanitized prep checklist the same day.
You gather evidence of how the work actually runs: steps, frequency, time. Client details stay out of it, always.
We map the week as it really runs, on a call. You only see discovery slots I can literally fulfil, weekends and public holidays included.
The 48-hour clock starts only after discovery is done and your evidence-ready prep is complete. Both, not either.
Consecutive means consecutive. Not two business days, not “within the week”. One document, six things inside it:
Whether five validated hours exist in your week, stated plainly, with the evidence trail behind it.
Every hour counted once. The arithmetic survives someone hostile checking it.
Where the hours hide, ordered by how much they return and how safely they can be addressed.
What could go wrong in each opportunity, before you commit anything to it.
The points where a person must stay in the loop, so nothing runs unsupervised that should not.
The single best place to start, chosen from the ranked list, with the reasoning shown.
Then we close it out together
One combined 45-minute session: findings review plus Blueprint Lock-in. You see the evidence, challenge anything, and we lock the one workflow the Blueprint gets written for.
This rule is the whole product. A generous count would make the promise easy to keep and the verdict worthless. So every counted hour carries all five conditions, and everything on the strike list dies in the overlap check.
Counted: needs every one of these
A representative, recurring baseline: this happens every normal week, not once during a bad month.
Contemporaneous process evidence: something recorded while the work happened, not a memory of it.
Corroboration from you or the process lead: the person who does the work agrees with the number.
A specific intervention inside your control: a named change your business can actually make.
Recurring replacement, admin and review time already deducted: the net hour, not the gross one.
Struck out: never counts
Hours counted twice across two opportunities
The same work shifted onto someone else on the team
Speculative growth
One-off backlog clearing
Hoped-for behaviour change
Unverified estimates
If striking these out means five hours cannot be validated, the Find Five Promise below takes over. That is the point of having it.
One boundary in plain terms: validated hours are addressable capacity, not realised savings. The Assessment proves where the hours are. It does not implement anything, and it does not guarantee five hours actually reclaimed, because nobody can guarantee that honestly before a build exists.
Within two business days of a successful Lock-in session, three working documents arrive. Each one exists to prevent a specific, expensive mistake:
Prevents building from a guess.
Goal, scope, inputs, outputs, roles, tools, steps, exceptions, human gates, access requirements and responsibilities, written down before anyone writes code. Portable enough for me, your own team, or any other implementer.
Prevents an untested automation touching real work.
At least five sanitized normal, exception and failure cases, with expected outputs, prohibited outputs, human-review requirements and a signed acceptance rubric. Nothing goes live until it passes tests you approved.
Prevents committing money on a hunch.
Conservative capacity economics, implementation cost, payback scenarios, a named owner, rollout steps, an adoption plan and a measurement schedule. The build decision gets made on numbers you can defend.
If I cannot validate five non-overlapping team-hours in a representative week, you have one business day to pick a door:
Door one
$999 back within 5 business days, plus an evidence closeout showing exactly what was checked and why it fell short.
Door two
Capped at 5 business days: one extra 45-minute call and 4 more hours of my analysis. If it still fails, the refund is automatic. There is no second extension, ever.
The attendance boundary
The 48-hour promise only works if both of us show up. One missed prep deadline or missed session gets one reset, no questions. A second miss closes the engagement: your $999 is refunded, the seat is released to the next team, and no bonus assets or Sprint credit are created. That is not a punishment, it is how the promise stays literal for everyone else in the queue.
I work with at most 4 active paid businesses across everything I do. A passed Fit Check holds one seat for 48 hours; payment secures it, expiry releases it. When the roster is full, qualified teams join a waitlist and cannot pay, which is annoying and also the only way the 48-hour promise stays true.
If the Assessment succeeds
A successful Assessment creates a $999 USD credit toward the 14-Day Workflow Proof Sprint, valid for 7 calendar days. Use it or not; the Blueprint is portable either way.
On privacy: everything runs on sanitized evidence. I never accept client files, credentials, or confidential documents, and this website never collects workflow evidence at all. Full detail on the privacy page.
Sanitized process evidence: how the work runs, how often, and how long it takes. Screenshots with client details removed, a written run-through of the steps, timesheet or task exports, the templates you reuse. If you can show how a workflow runs without exposing a client, you have enough. The prep checklist you receive after payment spells out exactly what to gather.
I work from sanitized evidence only. I never accept client files, credentials, or confidential documents, and if something arrives un-sanitized I delete it and ask you to resend it clean. The website itself never collects workflow evidence at all. The privacy page covers the full detail.
No. You need to know how your own business runs, which you already do. I do the analysis; you corroborate the baseline. Every deliverable is written for an owner making a decision, not an engineer reading a spec.
Not inside the Assessment. Find Five in 48 identifies and validates the hours and specifies one workflow worth fixing first. Building it is a separate, optional engagement offered privately after a successful Assessment. The Blueprint is deliberately portable, so you can also build with your own team or any other implementer.
After two things are done: your 45-minute discovery call and your completed evidence-ready prep. From that moment the verdict arrives within 48 consecutive hours, weekends included if that is where the clock lands. I only offer discovery slots I can literally fulfil, which is why the calendar is not wide open.
Above 20 people, join the larger-team waitlist and I will contact you when that lane opens. Solo, Codex Club is the honest route: it is free and built for putting AI on one real job. The Fit Check routes both cases automatically, which is one reason it exists.
If I cannot validate five non-overlapping team-hours in a representative week, you have one business day to choose: a full $999 refund paid within 5 business days with an evidence closeout, or one bounded second look. If the second look still fails, the refund is automatic. If you do not respond within the business day, it defaults to the refund.
Because a sales call is where vague promises live. The price is public, the promise is in writing, and the Fit Check does the qualification in two minutes. If this is not a fit, you find out before paying, not an hour into a pitch.
The fine print lives in plain English too: Find Five in 48 terms.
Find out in two minutes.
The Fit Check asks a handful of blunt questions and routes you honestly: to a reserved seat, a waitlist, or somewhere free that fits better. No sales call at any point.