Before you commission a workflow audit, hire a transformation consultant, or recruit a new operations manager, work through this list.
It is structured the same way the bureau structures the first ninety minutes of a real audit: top-down, starting with what the business is trying to do, ending with where the money is leaking. If you can answer three quarters of these in writing, you are far enough into the diagnosis that the next step is action, not more discovery.
Twenty questions. Save the page, write the answers, share with whoever else needs to see them.
Section A: The business shape (5 questions)
1. What is the operation actually selling?
Not the marketing answer. The honest one. A consultancy sells access to specific people. A recruitment firm sells access to a candidate pool and the work of placing them. A care provider sells trustworthy daily care. The shape of the product determines which workflows are core (cannot be automated away) and which are support (can be).
2. Who in the team has the most context on how things actually run?
Often not the founder. Often the person who has been there longest in an operational role. Identify them. They are the single best source of audit input.
3. If you doubled your customer count next year, where would the operation break first?
The honest answer is usually a specific person, a specific spreadsheet, or a specific manual process. That is your bottleneck. Everything downstream of it is a symptom, not a cause.
4. Which 20% of your customers generate 80% of the operational work?
The Pareto split is real. The 20% that generate 80% of the admin work are usually a specific segment or product type. Knowing which makes scoping any fix cheaper.
5. What would have to be true for the operation to run for two weeks without you?
The answer is a list of things you do that nobody else can. Each one is a candidate for documentation, automation, or hiring.
Section B: The systems (5 questions)
6. List every tool the team uses on a normal week.
Be ruthless. Spreadsheets count. Personal Gmail counts. The shared Dropbox folder counts. You will be surprised how many there are. Most UK SMBs the bureau audits have between 12 and 25 active tools.
7. Which two tools do not talk to each other but should?
Every operational pain point eventually traces back to a missing connection between two systems. The salesperson rekeys deals because the CRM does not push to accounts. The operations team builds reports manually because the project tool does not export to the dashboard. List the missing edges.
8. Which tool would cause the most chaos if it disappeared tomorrow?
That is your most load-bearing system. Whether it is the CRM, the spreadsheet that runs scheduling, or the Gmail account where every approval lives. Backup, ownership, and continuity decisions all start there.
9. Which tool do you pay for and barely use?
Every UK SMB has at least one. The licence renewal that lives on autopay because nobody has the time to evaluate alternatives. Cancelling those is the fastest cost win in any audit.
10. Where is the data that matters?
Customer data, financial data, project data, compliance evidence. Often it is in three places at once and disagrees with itself. The audit answer is to pick the one that is going to be the source of truth, then reconcile to that. Before commissioning anyone, write down where you think it is.
Section C: The people (5 questions)
11. Who on the team is burning hours on work that does not match their role?
Salespeople doing admin. Senior consultants generating reports. The owner reconciling the bank statement. Each one is an opportunity cost: their time is worth more spent on what they were hired to do.
12. Who would notice first if a manual workflow stopped happening?
Whoever is downstream of it. Their answer to "what would break for you" is the audit's first useful piece of evidence.
13. Which roles have you been trying to hire and failing to fill?
Admin-plus-technical hybrid roles are notoriously hard to fill in UK SMB markets. If you have been hunting for an operations manager who can also build dashboards for six months, the role might genuinely be a build, not a person.
14. Who has left in the last twelve months, and what did they take with them?
Tribal knowledge walks out the door. If a key process now relies on remembered context that a former employee held, that is a documentation emergency.
15. Which team member do customers ask for by name?
That person is a constraint. Their work cannot be automated, but the admin work that surrounds them probably can be, and freeing them from it pays back fast.
Section D: The money (5 questions)
16. What is the operation costing per month, all in?
Salaries plus tools plus overhead. Most owners can answer this within a week. If you cannot, the audit cannot tell you what saving any of it would mean.
17. Of that monthly cost, what proportion is going to "the work that should not exist"?
The shadow admin from earlier in this post. A conservative estimate based on the bureau's audits is 15 to 30% of operational labour cost. Half a day a week per worker, sometimes more.
18. What is one extra customer a month worth?
If the operations team currently cannot handle one more customer without breaking, you have a constraint. The maths of fixing it (a £4,000 build vs the gross margin of one extra customer per month) is usually decisive.
19. What is the cheapest version of "fixed" you would accept?
Some pain is worth £20,000 to fix. Some pain is worth £2,000 to fix. Knowing where the threshold is upfront prevents an audit recommending a build that is more than the pain justifies.
20. What is the budget ceiling you are willing to commit to a fix in the next six months?
Be honest with yourself. If the answer is "less than £2,000", an automation build is not the right answer; you need a process change, a smaller tool, or a no-code workaround. If the answer is "£2,000 to £25,000", that is the bureau's typical engagement range. Above that, you may be in agency territory.
What to do with the answers
If you can answer 15+ of these in writing, you have a partial audit already. The bureau's job at the formal audit is to validate, observe, and write down the punch list properly, not to discover from scratch.
If you can only answer 5 or 6, the work is not ready to be commissioned yet. The first useful step is a one-week time-log across the operations team to surface the workflows. Then come back to the list.
A few patterns the answers tend to reveal:
- If sections A and C are clear but B is a mess, you have a tooling and integration problem. A custom automation build is likely the right answer.
- If section B is clear but A is fuzzy, you have a positioning problem. A workflow audit will not fix that; the right first step is owner-level strategy work.
- If section D shows the budget ceiling is below £2,000, the right answer is probably no-code (Zapier, Make) or a small process change, not a custom build.
- If section C shows the team is the bottleneck and section D shows the budget supports a hire, the right answer may genuinely be hiring, not building.
What an audit adds beyond the list
A formal operational audit takes the same questions and validates them through observation. The bureau spends half a day to a day with the team, watches the actual workflows run, talks to the people doing the work, reads the systems involved.
Two things change after observation. First, the diagnoses get more specific. "The Monday report takes too long" becomes "the Monday report involves four tools, two manual joins, and one lookup table that goes out of date weekly". Second, the proposed fixes get fixed-cost numbers attached. The output is a punch list with each item priced, not a roadmap.
If you have worked through this checklist and the answers point at a build, the 15-minute triage is the cheapest way to confirm. The audit fee credits in full against any build that follows, so the engagement either delivers the diagnosis or the build cost, not both.
If the answers point at hiring or at no-code, the bureau will tell you that at the triage and point at someone better suited to the work. That is the point of the conversation.