The bottleneck audit we run before quoting any project

Most agencies quote the project you asked for. We quote the project you need — which is usually smaller. The difference is a short diagnostic we run first, every time, before a single line of code. Here’s the whole thing, so you can run it yourself.

Why diagnose before building

A rebuild feels like progress. It rarely is. If you replace a system without knowing where the time actually leaks, you’ll faithfully rebuild the leak. The goal of the audit is to find the one or two steps carrying most of the cost, so the fix can be precise — and cheap.

Step 1 — Pick the unit of work

Don’t try to “map the business.” Pick the single thing that flows through it most often: an order, a ticket, a lead, an invoice, an onboarding. Whatever happens dozens or hundreds of times. Volume × friction = cost, and you can only see friction by following volume.

Step 2 — Follow one all the way through

Take one real example and trace it end to end, naming every hand-off and every tool it touches. Write it as a flat list:

Where does it enter? (form, email, API, phone)
Who touches it, and what do they do?
What tool does each step happen in?
Where does it wait? (waiting is invisible and huge)
Where does it exit “done”?

Keep it to one page. If it doesn’t fit on a page, you’ve zoomed in too far.

Step 3 — Time the boring parts honestly

Now put a number on each step — including the parts everyone dismisses as “quick.” The 90-second copy-paste that happens 200 times a month is 5 hours. The “I just check it real quick” is a context switch that costs more than the check.

Two columns: minutes per occurrence and occurrences per month. Multiply. Sort descending. The top of that list is your bottleneck — not your opinion of it, the math’s opinion of it.

Step 4 — Mark the failure points

Next to each step, note where things go wrong: the transposed number, the dropped email, the status nobody updated. Errors are bottlenecks with a multiplier, because each one triggers rework far longer than the original step.

A 2-minute step that fails 10% of the time, where each failure costs an hour to unwind, is not a 2-minute step. It’s a 6-minute step in disguise.

Step 5 — Prescribe the smallest fix

For the top one or two rows only, ask: what’s the least we can change to remove this cost? In order of preference:

Delete it. Does the step need to exist at all?
Automate the hand-off. The data usually already exists in structured form; a human is just moving it. (See how we cut a 14-minute step to zero.)
Tighten the tool. A small internal tool or a better form beats a platform migration.
Only then, rebuild — and only the part that earns it.

The output

One page: the flow, the numbers, the top two costs, and the smallest fix for each. That page is the quote. It tells you what to do, what it’s worth, and — just as often — what not to spend money on.

If you’d like us to run this on your busiest workflow, show us where it hurts. The audit is the part we’re happy to do out loud.