From analysis to verified savings in 90 days

Ninety days is one quarter. It is long enough to build two automations, run a clean measurement window, and put a verified number in front of a finance team. It is not long enough to transform a company, and a firm promising otherwise is selling a calendar it cannot keep.

Here is what the first quarter of a performance-based engagement actually looks like, week by week, including the parts that go sideways.

Weeks 1-2: analysis and strategic review

The engagement opens with diagnosis, not a kickoff deck. The consultant maps how work actually moves through the business: who touches an order, an invoice, a support ticket, and where hours pool. Sample data comes out of the real systems, and the people doing the work get interviewed, because the org chart never matches the workflow.

The output is a ranked list, typically eight to twelve candidate automations, each with an estimated yield in hours and dollars. The distribution is always lopsided. In most businesses we have analyzed, the top two candidates carry more than half of the total projected savings. That concentration is what makes a 90-day plan possible at all.

The two weeks close with a strategic review: leadership sees the ranked list, the assumptions behind each estimate, and a recommendation on where to start. At Qonnex, the analysis and the strategic review are free; beyond that, you don't pay unless verified savings exist. The mechanics are on the how it works page, and the starting point is the analysis itself.

Weeks 3-4: agreement and baseline

This is the unglamorous fortnight, and the most important one. Both sides sign the agreement and lock the baseline: current costs, current cycle times, current volumes, written down from the client's own numbers.

Expect the first surprise here. Every baseline exercise we have run has surfaced at least one number the company believed but its own systems could not support. The order count in the operations tool disagrees with the finance spreadsheet. Two exports of the same report differ by a few hundred rows. A cost everyone quotes turns out to be a three-year-old estimate.

This is normal, and it is why baseline work gets two full weeks instead of an afternoon. Reconciling those numbers is slow, but a baseline both sides trust is the foundation for everything that follows. A verified saving is only as credible as the "before" picture it is measured against.

Weeks 5-8: the first two automations

Implementation starts with two automations, not five. Two is enough to matter and few enough to do properly. The first typically ships around week six, the second by week eight, each starting with a small pilot group before wider rollout.

Two things go wrong here often enough to plan for.

First, one of the two automations always needs rework. Not because the build was sloppy, but because the first contact with live volume surfaces exceptions nobody documented: the supplier who formats invoices differently, the approval step that exists only in one veteran employee's head. Budget roughly a week of rework into the schedule. When it isn't needed, the schedule gains a week; when it is, nobody panics.

Second, adoption dips before it climbs. Around week seven, usage often drops as the team quietly runs the old process in parallel to check the new one against their own judgment. This looks like failure on a dashboard and is actually the healthiest thing that can happen: people are auditing the system before trusting it. Teams that skip this dip usually haven't engaged; they've complied.

1. 1Weeks 1-2: analysis and strategic review
2. 2Weeks 3-4: agreement signed, baseline locked
3. 3Weeks 5-8: first two automations built and shipped
4. 4Weeks 9-12: measurement window, first verified numbers

Weeks 9-12: the measurement window opens

With both automations stable, the measurement window opens: a fixed period, agreed back in week three, during which actual costs are compared to the locked baseline, line by line.

The first verified number is usually smaller than the projection and larger than the skeptics expected. In our experience a first window verifies roughly two-thirds of the projected figure, with the gap explained by ramp-up: the automations only reached full volume partway through the window. The second window, run after the quarter ends, typically closes most of that gap.

What matters more than the size of the first number is its quality. The client reviews the calculation before anything is settled. Every line traces to a system of record. When a finance chief can follow the arithmetic from baseline to outcome without taking anyone's word for anything, the engagement has produced its most valuable asset: a measurement machine both sides trust.

What 90 days proves, and what it doesn't

At the end of the quarter, two automations are live, a baseline is locked, and a first verified number exists. Perhaps two of ten worthwhile processes have been touched. The company is not transformed, and it was never going to be.

Ninety days is enough to verify direction. It is not enough to finish transforming a company, and it was never meant to be.

What the quarter buys is something more durable than speed: proof that the method works on your numbers, in your systems, with your people. The remaining automations on the ranked list now run through machinery that has already survived contact with reality: messy data, one rework, an adoption dip, and a verification the finance team signed off on.

If you are evaluating a firm that promises full transformation in a quarter, ask to see the measurement plan for week nine. If there isn't one, the promise is a calendar, not a commitment.