Senior leadership, made concrete · Article 4 of 4
Built so you can stop
Read a consultancy’s exits before its entrances. Our engagements are staged with real decision gates, deliverables you keep, and a credit mechanic that puts our money where the plan is.
1 · Read the exits before the entrances
There is a test for consultancies that almost nobody runs at the start of a relationship, and that everybody wishes they had run by the end of one.
Ask how it stops. What do we hold when a phase concludes? What happens if we want to pause after the first deliverable? Where does the knowledge live, in our systems or in yours? What, precisely, would leaving cost us?
The industry’s quiet secret is that a great many engagement models are architected so that those questions have uncomfortable answers. Discovery that only makes sense if it becomes implementation. Deliverables in proprietary formats, or parked in the consultant’s tenant. Documentation promised for the end of the project and somehow never current. Knowledge held in the heads of a rotating bench, so that continuity itself becomes the thing you are paying to keep.
None of that requires bad faith. It is simply what grows when nobody designs the exits, and retention through friction remains the industry’s default architecture.
We built Many Branches on the opposite bet: for firms whose reputation is the product, the provider confident enough to make leaving easy is the one worth staying with. This article is the architecture of that bet, and it closes the series because it is the frame the rest sits inside.
2 · The ladder: start small, earn the next phase
Our engagement model is a ladder with three rungs, and its first property is the size of the first rung.
Assess is a fixed-fee diagnostic: stakeholder interviews, a current-state review, a risk and opportunity map, and a findings report. Priced to be an easy yes. The first decision a client makes about us should be a small one. Plan is scoped per phase, so the roadmap, the AI policy, and the architecture are each priced as their own decision with their own deliverables. Execute is delivery: phased builds, then the steady state, whether that is roadmap execution on subscription or a fractional leadership retainer.
Our site says you can start with any one of our services and that most clients end up using all four. Both halves are true, and the second half is the one worth examining, because clients climb for a reason that has nothing to do with lock-in. Each stage generates the evidence for the next: the assessment proves we understand the firm, the plan proves the thinking holds, and execution proves the follow-through exists. Climbing is what happens when the work goes well.
The rest of this article is about what happens when it does not, since that is where engagement architecture is actually tested.
3 · Stage independence
Every stage of a Many Branches engagement is written to stand alone, and this is contract language rather than brochure language. The statement of work says, in terms a non-lawyer can read, that stage one stands alone, that you may conclude the engagement at its end, and that you retain all of its deliverables. Every subsequent gate reads the same.
Standing alone has a physical meaning too. Deliverables land in standard formats, in your systems, under your accounts: documents you can open in Word, diagrams you can edit, policies your staff can amend, credentials in your password manager rather than ours. Where we build, we build in your tenant. Documentation is part of the definition of done rather than a farewell gift. Our site promises a findings report you own either way, and that phrase generalizes to every gate. What you hold is complete, current, and usable by whoever comes after us, including nobody.
We structure it this way for a reason worth stating baldly. Dependence and trust are opposites. A client who stays because leaving is painful is a hostage with a renewal date, and hostages write bad references. A client who stays because every stage keeps proving out is a reference for life. In a business where reputation is the product, ours as much as yours, only one of those is an asset.
4 · The credit mechanic
One more piece of architecture deserves daylight, because it changes the incentives of the planning stage itself.
In our roadmap and planning engagements, a defined share of the stage’s fee is creditable toward follow-on execution if you proceed within a set window. The exact share and window are engagement terms, set out in each statement of work rather than in a blog post. The mechanic itself is standing policy. What it does is the point, which is that it converts the roadmap from a cost into a down payment on execution.
Consider what that does to the document. A planning consultant with no stake in execution is rewarded for ambition, because the grander the plan, the larger the plan looks. A planner whose fee partially converts only if the client chooses to execute is rewarded for plans that actually get executed, which is to say plans sequenced sensibly, costed candidly, and sized to the firm that has to live with them. The credit mechanic is us accepting a haircut for the privilege of being held to our own roadmap. We think that is the right side of the incentive to be standing on, and we notice how few planning engagements elsewhere are structured to stand there.
Two boundaries. The credit applies when execution proceeds with us, within the window, so it is an alignment device rather than a discount in costume. And if the right executor for your plan is somebody else, a specialist integrator or your own team, the roadmap is built to hand over. That is what stage independence means, and a plan only we could execute would be a lock-in wearing a credit.
5 · The gate questions, portable edition
The decision gates between stages are simple in operation. At each one you continue, pause, or stop, with everything already in hand. The default is open. What gives the gates teeth is that both sides prepare for them. We arrive with the stage’s results measured against what was promised. You arrive with the standing questions, which are portable to any provider you are evaluating, us included.
- What do we hold, today, if this ends on Friday? Ask to see the deliverables list and where each one lives.
- Which formats, and whose tenant? Editable, standard, yours is the passing grade.
- What did the last stage predict, and what actually happened? Plans should accumulate a track record against themselves.
- What would you do next if we did nothing for six months? A good provider has a real answer, because pause is a legitimate gate outcome and the estate should be safe to leave idle.
- Who else could execute this plan? If the answer is only us, the plan is a leash.
A provider who welcomes those questions, and who has the deliverables list ready before you finish asking for it, has architecture worth buying. A provider who bristles has answered them anyway.
6 · Where this closes
One series, one shape. A roadmap that is a set of decisions rather than a wish list. Judgment rented at the fraction that fits. Execution run as a subscription whose best month is your quiet one. And around all of it, gates, independence, and a credit mechanic that keep every phase voluntary.
Clear deliverables and defined decision gates is the promise on our site. This series has been the long version of it, and the short version is that an engagement built so you can stop is the only kind built to deserve continuing.