AI automation assessment across the whole operation
Four to five weeks. Six to eight processes across teams, volumes measured rather than estimated, and a build-ready specification for the highest-value opportunity.
What the Deep Audit covers
Six to eight connected processes, spanning more than one team, department, site, or trading entity.
Where the AI Readiness Audit maps up to three processes in one part of the business, this covers the operating picture end to end — including the points where the same information is handled separately by two teams who do not know the other is doing it.
Volume is measured, not estimated.
We work from system exports — ticket counts, invoice volumes, time records, message logs — covering twelve months where the data allows. Estimates are enough to rank three opportunities against each other. They are not enough to justify a decision of this size.
How the assessment runs
We start with the data, then interview.
System exports are analyzed before interviews begin, so conversations start from what the numbers already show. It changes the quality of the interviews considerably — people respond differently to “this runs four hundred times a month, walk me through it” than to “tell me about your process”.
We look across processes, not only along them.
The same information is often handled by two teams who do not know the other is doing it. That overlap is invisible from inside either process, and is consistently among the largest single opportunities.
We score readiness per team.
An averaged business-wide figure conceals both the team ready to move now and the one that is not.
We sequence, not just rank.
With six or more candidates, order determines whether a program compounds or stalls.
We specify the top recommendation in full.
So the decision afterward is whether to build, not what to build.
What arrives at the end
Everything in the AI Readiness Audit, plus:
- Measured volume analysis Actual counts and durations from your systems, twelve months where available, with the source of each figure named.
- Cross-process view Where the same information is handled by more than one team, and what it costs in aggregate.
- Departmental readiness scorecards Six dimensions scored per team, with reasoning.
- Dependency and sequencing plan What depends on what, what can run in parallel, and the critical path across twelve months.
- Build-ready specification for the top recommendation The sequence as it will run, every system involved, what good output looks like at each step, review points and who performs them, cases requiring separate handling, and what testing would cover. Detailed enough to build from, by us or anyone.
- Twelve-month roadmap Quarter by quarter, with the sequencing logic set out.
- Leadership session 60 minutes, separate from the walkthrough with the people interviewed. Different audiences need different conversations.
Timeline
| Stage | What happens | |
|---|---|---|
| 01 | Kick-off | 90 minutes. Scope, teams in question, and the system exports we will need. |
| 02 | Data analysis | System exports analyzed before any interview takes place. |
| 03 | Discovery | Eight to twelve interviews, starting from what the numbers already show. |
| 04 | Cross-process review | Where the same information is handled by more than one team. |
| 05 | Analysis and specification | Scoring, sequencing, and the build-ready specification. |
| 06 | Walkthrough | 90 minutes with the people interviewed. |
| 07 | Leadership session | 60 minutes, separate. Different audiences need different conversations. |
| 08 | Handover | Full written report, twelve-month roadmap, and the specification. |
| 09 | Follow-up | A call once the report has been read. |
What we need from you
- Eight to twelve people for 45 minutes each.
- System exports covering twelve months where available.
- A named contact per system who can obtain that data.
- A list of the software each team pays for.
- Fifteen to twenty real examples of the work.
Around twelve hours in total, spread across teams and five weeks. The system exports are worth arranging early — they usually have the longest lead time.
Outside this engagement
- Building anything. This is a review, though it produces a specification that can be built from.
- Software selection or procurement.
- Financial modeling beyond the time and volume figures in the report.
- Organizational or process redesign unrelated to automation.
- Specifications for every recommendation — one is included, further ones are quoted separately.
- Ongoing advisory after the follow-up call.
When the AI Readiness Audit is enough
Where up to three processes within one part of the business are in question, the AI Readiness Audit covers them in two to three weeks for half the fee. Where interview estimates are enough to decide, it is the cheaper and faster route, and we will say so. Compare the two →