The Acceleration Framework

What each stage produces, and what it decides.

Five stages, each with a purpose, a deliverable and a decision at the end of it. This page sets out what you receive, what your team has to contribute, and what happens to the work once it is live.

Stage by stage

Every stage ends in a decision, including the decision to stop.

Data protection, security and human accountability run across all five stages. They are not a gate the work passes through once.

Stage 1

Orient

Establish where the opportunity is and whether it is worth pursuing.

You receive

Prioritized use case, with the alternatives and why they were not chosen
Baseline measurement of the current process
Dependencies: systems, data, permissions, people
Business case with the assumptions written down

You contribute

Access to the process owners
Whatever process metrics exist, however partial
A decision-maker who can accept or reject the recommendation

Decision at the end of this stageIs there a case worth building? A recommendation not to proceed is a valid and reasonably common outcome at this stage, and it still leaves you with the baseline and the analysis.

Stage 2

Prove

Test a defined workflow against agreed success criteria.

You receive

A working proof of value on a real workflow, not a slide-deck demonstration
Findings, including what did not work
A scale-or-stop recommendation against the criteria agreed before we started

You contribute

Representative test data
Integration credentials through your own secure process
Named acceptance criteria and someone who can sign them off

Decision at the end of this stageDid it meet the criteria we agreed at the start? The criteria are set before the build, not chosen afterwards to fit the result.

Stage 3

Govern

Set and validate the controls required for the intended level of automation.

You receive

Responsibilities: who is accountable for what the system does
Permissions and the actions the automation may take unattended
Approval routes and escalation paths
Assurance evidence a reviewer can inspect

You contribute

Your control framework and any regulatory obligations that apply
A risk owner who can accept a residual risk on the organization's behalf

Decision at the end of this stageIs the level of autonomy defensible to the people who will be asked about it — your risk function, your auditors, your regulator?

Stage 4

Scale

Integrate the solution into day-to-day operations.

You receive

Rollout plan sequenced by team, site or process
Operating documentation and runbooks
Team enablement so the people using it understand what it does and does not do

You contribute

Change windows and release approval
Time from the teams who will use the system
A named internal owner for the workflow after handover

Decision at the end of this stageCan your teams run this? Scaling something nobody internally can operate creates a dependency, not a capability.

Stage 5

Compound

Use performance evidence to improve the result over time.

You receive

Performance reviews against the original baseline
A prioritized improvement backlog
A view of which further processes the same approach would suit

You contribute

Continued access to the performance data
A cadence you will actually keep

Decision at the end of this stageIs it still delivering what it was built to deliver? Automations decay quietly when the surrounding process changes.

Across every stageData protection, security and human accountability. Controls are designed alongside the workflow, not retrofitted once something is already live.

What we need from you

The engagement depends on a few things we cannot supply.

Most delays in this kind of work are not technical. They are access, credentials and decisions. Being direct about that up front is cheaper for everyone than discovering it in week three.

Access to the people who run the process today. Nothing substitutes for this, and it is the most common reason a stage slips.
Integration credentials issued through your own secure process. We do not ask for shared accounts or credentials over email.
A named decision-maker per stage who can accept or reject the recommendation.
Agreement on the measurement period before anything is built, so the before-and-after comparison is like for like.
Honesty about constraints — budget cycles, in-flight change, a process already being redesigned. These change what is sensible to build, and finding out late is expensive.

After launch

Handover, and what support actually covers.

Handover

Runbooks, monitoring, the operating documentation and a named internal owner agreed during Scale. If nobody on your side owns the workflow, it will drift.

Ongoing support

Available under a separate agreement covering monitoring, incident handling and optimization. It is a commercial arrangement with defined boundaries, not an open-ended commitment bundled into delivery.

Review

Performance against the original baseline, at a cadence you agree. Where an automation has stopped earning its keep, we would rather tell you than quietly keep invoicing for it.

FAQs

Common questions

Do we have to start at Orient?

No. If you already know the workflow and have a baseline, we can start at Prove. What we will not do is skip the baseline, because without it there is no honest way to report whether the work made a difference.

Does governance only start at stage three?

No, and presenting it that way would be misleading. Stage three is where controls are formally set and validated, but data protection, security and human accountability are considered from the first conversation. A workflow designed without them cannot be retrofitted with them cheaply.

What happens when a stage says stop?

You keep everything produced up to that point — the baseline, the analysis, the findings — and you do not pay for a stage you did not run. A recommendation to stop is a result, not a failure of the engagement.

Who owns what we build?

Ownership of the solution, the configuration and the documentation is set out in the engagement contract before work starts. Where a third-party platform is involved, its licensing terms apply to that platform and we will say so explicitly rather than leaving it ambiguous.

What happens after go-live?

Handover includes runbooks, monitoring and the named internal owner agreed during Scale. Ongoing support from ETT is a separate commercial arrangement, not something included by default — we would rather be clear about that than have you discover it later.

Which process would you improve first?

Bring an operational challenge, an AI idea or a stalled initiative. We’ll explore the workflow, the potential value and the conditions needed to move forward.

Request an AI session Explore solutions