How we work

Choose the improvement, prove it, then build on it.

An engagement starts with a manageable first phase that ends in a decision. From there it moves stage by stage — scope, build, adoption, measurement — and you decide at each step whether to continue.

Talk to us Our method

Our method

We follow Rewired. Our version is FAST.

Rewired (Lamarre, Smaje and Levin; Wiley, 2nd edition 2026) sets out how established companies turn technology into lasting performance. We work by it, in ten rules grouped into four parts: Focus, Accountability, Systems and Traction.

The ten rules →

Routario is not affiliated with McKinsey & Company or the authors of Rewired. We apply the method in our own way and adapt it where our clients need something different.

Two starting points

You may know the process — or only that AI should be doing more for you.

"We know the process we want to improve."

We check that it's worth doing: the baseline, the likely value, the data and the people it depends on. If it holds up, we go straight to scoping it. If it doesn't, we'll say so.

Start with the process →

"Help us decide where AI can create value."

We look across the business with your leadership and the people who do the work, and come back with a ranked shortlist and a recommendation for the first improvement.

Start with an assessment →

The stages

What happens, what you contribute, and what you decide.

Stage 1

Choose where AI matters

We interview leadership and the people who run the work, look at how processes, systems and data actually behave, and size each opportunity. Every estimate carries its assumptions, so you can challenge it.

You contribute: A sponsor, access to the people who do the work, sample data
Output: A ranked shortlist: value range, feasibility, data readiness, owner, prerequisites and the investment each would need
Decision: Which improvement to build first — or that none is worth it yet

Stage 2

Scope the first improvement

We measure how the work runs today, agree what success means and write it down on one page — including what it will cost to run and what has to be true on your side.

You contribute: A process owner, baseline data, agreement on responsibilities
Output: A scope: opportunity, baseline, expected value, prerequisites, implementation, adoption, operating costs, measures, decision points
Decision: Go or no-go

Stage 3

Build it with your people

We build on your real data and connect to the systems you already use — through e-mail, exports, file drops or APIs. The first weeks run under supervision: a person reviews every decision until the results hold up. Testing gets more of our time than building, and nothing runs on its own until it matches your people's decisions.

You contribute: The people who do the work, reviewing and correcting; IT and security review
Output: A working improvement in daily use, typically live on real data within one to two weeks for a well-scoped process
Decision: When it moves from supervised to routine

Stage 4

Adopt and measure

New tools don't change results on their own — people's work has to change, including the steps before and after the tool. We agree who reviews what, train each role once, adjust the rules with the team, and measure against the baseline.

You contribute: Managers who hold the new way of working in place
Output: Value found, value acted on and value realised, reported separately
Decision: Keep, change or stop

Stage 5

Expand and keep improving

The next opportunity on the shortlist reuses the data connections, rules and records already in place. Improvements already running are monitored and tuned as your business changes.

You contribute: The next owner and the next baseline
Output: A growing set of improvements that share one foundation
Decision: What comes next, and at what pace

How an engagement is scoped

Everything that decides the value, on one page.

Before anything is built, we write down what we expect it to be worth, what it depends on and what it costs to run. That scope is what you approve, and it's what we're measured against.

Each stage is scoped and priced before it starts. Running costs — hosting and AI usage on our platform, plus continuing support — are part of the scope, so the value case includes them.

Scope · Supplier invoice exceptions Example

Opportunity: Invoices paid above contract price or twice, found only at month-end
Baseline: Invoices per month, share with an error, hours spent checking — measured on the last quarter
Expected value: A range, with each assumption stated: error rate, share caught before payment, recovery rate
Prerequisites: Copy of the AP mailbox; contract price export; an approver matrix; one AP owner
Implementation: Intake and checks run alongside today's process first; routing switched on supplier by supplier
Adoption: Approvers see only exceptions, with the reason; AP reviews a weekly sample
Operating costs: Platform and AI usage at current volumes; your team's review time
Measurement: Exceptions confirmed, money not paid or recovered, time to approval
Decision points: After the shadow run; after the first full month

How we handle value

A value case you can audit.

Every estimate carries its assumptions, every result is measured against a baseline, and each stage ends with a decision.

Sized before we build

Every estimate comes with its assumptions written down, so your finance team can test them.

A baseline first

We measure how the work runs today before anything changes, so the difference is a measurement, not an impression.

Staged investment

Each stage ends with a decision. You commit to the next step, not to a programme.

Responsibilities agreed

What we deliver and what your side provides — people's time, data, decisions — is written into the scope.

Running costs included

The value case counts what the improvement costs to operate, not just what it costs to build.

Realised value reported

We report what showed up in results, separately from what was found or acted on.

FoundPotential value

What the exceptions and opportunities the improvement surfaces would be worth if all were acted on.

Acted onValue acted on

The part your people confirmed and did something about — a correction, a held payment, a visit.

RealisedValue realised

What actually showed up: money not paid or recovered, sales restored, hours given back and redeployed.

Six capabilities, built together

A platform alone doesn't change how a company works.

Rewired names six capabilities a company needs to turn technology into results. Strength in one doesn't make up for weakness in another. Every engagement builds all six, in proportion to the problem.

A roadmap the business leads

One area at a time, chosen for what it's worth. Its roadmap belongs to the leader who runs it. It goes from seeing every case, to recommending the action, to acting within the limits you set.

Owners and experts

A process owner decides the rules. Your experts correct the system in the first weeks. Each correction becomes a rule they can change themselves, without us.

Decisions that move fast

A weekly session on real outputs, a monthly value review and a quarterly decision on what continues. What the system may do alone, and what needs a person, is written down.

Technology that's already built

Our platform, connected to the systems you keep. Each new improvement reuses what the previous one built, so the second costs less than the first.

Data ready for use

Only the data the improvement needs, such as suppliers and customers, locations, price lists and documents. Each has a clear quality bar, and every correction makes it better.

Adoption, measured

Training by role, a baseline, and value reported as found, acted on and realised. The change holds after we step back.

Who you work with

A small team on our side, a clear set of roles on yours.

Engagements are sized to the value at stake: a few experienced people who stay from the first conversation through adoption, not a large programme team.

Routario

Engagement lead: Owns the value case and the decisions with your sponsor, from the shortlist to the measured result.
Operations specialist: Works with the people who run the process: how the work actually happens, what changes, and what "done" looks like.
Implementation engineer: Builds the improvement on our platform and connects it to your systems and data.

Your side

Executive sponsor: Owns the outcome and the go / no-go decisions.
Process owner: Decides the rules and the exceptions, and holds the new way of working in place.
The people who do the work: Review, correct and shape the improvement in the first weeks.
IT and security contact: Access, data and deployment review.

Questions leaders ask

Commitment, cost and fit.

Who is this for?

Companies with real operational complexity and money at stake, but without a large in-house AI team. Engagements are sized to the value in question, not to a transformation programme.

Is this McKinsey's method?

It's the method set out in Rewired, which we have adopted and adapted for mid-sized companies. We are not affiliated with McKinsey. We use the book as our playbook because it matches what works in practice: a few areas chosen for value, business ownership, investment in stages, and as much attention to adoption as to the build.

How is an engagement priced?

Each stage is scoped and priced before it starts. Running costs — hosting and AI usage on our platform, plus continuing support — are part of the scope you approve, so the value case includes them.

Do we have to replace any systems?

No. We work with the systems you have. Where there's no ready connection, we use e-mail, exports, file drops or your system's API. Your ERP, CRM and accounting stay the systems of record.

Who stays in control?

Your people. Rules and thresholds are yours; anything that moves money or reaches a customer waits for a named person unless you decide otherwise; every decision is logged. See Trust & Security.

What happens after the first improvement?

Either you stop, with a working improvement and a measured result, or we take the next opportunity from the shortlist. Improvements already running are monitored and adjusted as your business changes.

See our work →Where we help →The platform →

Discuss your priorities

Start with a conversation about your priorities.

Tell us what you want to improve — a specific process, or a broader ambition and the question of where to begin. We'll suggest a first step and what it would take.

We reply within one business day, usually with two or three questions and a suggestion for the first step.

Thanks — we have it.

We'll get back to you within one business day with how we'd approach it and what we'd need to start.