What an AI Readiness Audit Actually Covers

"Audit" is a word consultants use to make a sales meeting sound like a service. So here is the whole thing in the open: what gets asked, what gets delivered, what it deliberately excludes, and when you should not bother.

The shape of it

One session of 45 minutes. One report of five pages, delivered within 72 hours. That is the entire engagement. There is no phase two hidden in the small print, no discovery workshop, and no follow-up presentation you have to sit through.

Two people should be in the room from your side: whoever makes spending decisions, and whoever actually runs operations day to day. The second person knows where the time goes. The first person can act on it. Without both, the session produces a document rather than a decision.

The four blocks of questioning

Block 1 — How the business makes money (10 minutes)

Before anything about technology, we need to know how revenue actually arrives. What you sell, to whom, and through how many steps between first contact and payment. Automating a process that sits far away from revenue produces a saving nobody notices.

The most useful question in this block, and often in the whole session, is this one: if double the work arrived tomorrow, what would break first? The answer names your real bottleneck, and the bottleneck is almost always where the first automation belongs.

Block 2 — Where the hours go (14 minutes)

This block hunts for numbers rather than impressions. We walk through a normal day for whoever is busiest, then take each task named and ask the same three follow-ups: how many hours a week, who does it, and what happens when it goes wrong.

Cost of error matters more than people expect. A task taking three hours a week is minor. A task taking three hours a week that causes a shipping error twice a quarter is not.

Block 3 — Tools and data (10 minutes)

What already exists determines what is realistic. A business running on WhatsApp and Excel has a different shortest path than one already on an ERP. We also ask what has been tried before — and specifically, what was tried and failed.

That question is the most valuable one in the block. A previous failed attempt tells you where the internal resistance sits, who lost confidence, and what a recommendation has to work around. Any proposal that ignores the reason the last one failed will fail the same way.

Block 4 — Close (7 minutes)

We read back what we heard, in one paragraph, and check it is right. Then one final question: if one thing could be fixed next month, what would you choose? The answer often differs from the highest-value opportunity — and knowing the gap between what pays most and what the owner wants most is essential to writing a report that actually gets acted on.

What arrives 72 hours later

Page	Contents
1	Executive summary — the three opportunities in a table with estimated annual saving, effort level, and start date for each
2	Where your hours go — the process map with weekly hours, and what that adds up to as a share of a full-time salary
3	Opportunity #1 in full — current state, future state, named tools with real monthly costs, the ROI calculation shown, and a first step
4	Opportunities #2 and #3 — same structure, half a page each
5	90-day roadmap, the risks that could derail it, and an honest note on what we did not cover

Page five carries the section most reports leave out. Forty-five minutes is enough to find the large, obvious losses. It is not enough to assess data quality properly, or to judge whether your systems can integrate. Saying so is more useful than pretending otherwise.

Three rules we hold ourselves to

Estimate conservatively. Where a model suggests 40% time saved, the report says 25%. An estimate that survives a sceptical finance director is worth more than one that impresses in the room and collapses in month three.

Never promise revenue. Cost saved is provable — you can count the hours before and after. Revenue depends on demand, pricing, and competition. We only put numbers on the part we can defend.

Show the arithmetic. Every figure in the report is accompanied by the calculation that produced it. If you cannot check a number, you should not trust it, and we would rather be checked than believed.

When you should not get one

An audit is the wrong starting point in three situations, and it costs nothing to say so:

• Fewer than about five people. Below that there is rarely a process repeatable enough to automate. You are better served by picking one AI tool and learning it properly.
• You already know exactly what you want built. If you have identified the workflow and only need execution, skip the diagnosis and go straight to a scoped implementation quote.
• The real problem is not operational. If the business is struggling on demand, pricing, or cash collection, no amount of automation will fix it, and finding hours to save would be a distraction from the actual issue.

Why ours is currently free

Hafgufa is building its first set of documented results. Ten real diagnoses across UAE businesses teach us more about where the money actually leaks in this market than any amount of desk research — and each one becomes a case study that makes the next engagement easier to win.

You get genuine work at no cost. We get evidence. Both sides understand the trade, which is a better foundation than pretending it is generosity.