Every audit firm I know budgets technology the same way. It's a small percentage of revenue, benchmarked against peers, and capped before anyone has looked at what the tool actually does.

Salaries are the majority of an audit firm's cost base. So when a platform that takes work off the team gets judged against the technology line, it looks expensive. Judged against the payroll line, the same platform is a different conversation.

This piece is for whoever owns the firm's P&L. It's about which budget line automation should be measured against, and why that choice decides the answer before the evaluation starts.

Capacity, measured against the right line

In the last 2 pieces I argued that the profession's capacity problem is about what each auditor-hour produces, and that hiring into a manual process buys hours without adding capacity. If that's right, the comparison that matters is with the other ways a firm buys capacity: another junior, overtime through the season, or a delivery center offshore.

Measured that way, it becomes a capacity decision, and the questions change. How many more engagements can the same team carry? What does a senior's season look like when the tie-outs are already done? What would you say yes to that you're turning away today?

[Figure 3: the same 5 people across a busy season, before and after the manual layer moves to RobotX Flow. Illustrative. Team size, engagement count and proportions are not a specific client's.]

What the hours buy back

The same team carries more engagements. The juniors spend the season on exceptions and judgment instead of tie-outs, which is the work that makes them good auditors faster, and the work that makes them want to stay. Work that would otherwise go offshore stays with the team that knows the client. The season ends with people who are tired rather than broken. And the firm gets to take on the engagements it has been turning away.

That's what I mean by ambition. More audits, done better, by the people you already have, who are still with you next year. The juniors of today are the partners of tomorrow, and I'd rather they spent their training years learning judgment.

Giving the time back

For most of the last decade, audit software has meant making the auditor faster. Faster matching, faster extraction, a few minutes shaved off a tie-out. Useful, and the auditor was still the one doing the tie-out.

We started RobotX in 2017 on a different idea: give the auditor the time back. RobotX Flow runs tests of details and walkthrough procedures in the background, without the auditor in the workbook. It works through the population and the evidence, fills in the working paper, and returns it ready for review. You can close your laptop, and the work is done when you open it again.

What comes back is a working paper documented to ISA 230: procedure, evidence, results and conclusion, with every output traceable to its source documents. Every conclusion is marked for review before it goes into the file. The opinion stays with the auditor. They form it from a completed working paper instead of a pile of PDFs at 11 at night.

Flow also comes with the procedures already built. A firm doesn't need an innovation team to translate its methodology into automation before it sees an hour back. The procedures are the ones auditors run, built by auditors who ran them at firms including Deloitte and KPMG.

Where it doesn't reach yet

One honest condition. Flow runs standardized procedures. Today an auditor still decides when a procedure is ready to run and presses the button, so the auditor is still the trigger. The judgment areas, and the decision to start the work, stay with the team.

Start with the hours

If your busy season keeps costing more hours than the plan accounted for, start with where the hours on a typical test of details are going. Then decide what you'd do with them back. In my experience, nobody's answer is "less".

See how Flow turns manual hours into capacity →

This is the final part of a 3-part discussion about the capacity issue in the audit industry written by our CFO, Will Kuo. Read the previous parts:

Part 1: Too many auditors and not enough →

Part 2: Deadlines don't move, people do →