What does a 100-day operations plan look like after acquiring a founder-led company?
A 100-day operations plan has three phases. Days 1 to 30: map where work and decisions sit, and stabilize cash and customers. Days 31 to 60: write down the top processes and set decision limits. Days 61 to 100: run a weekly rhythm and pilot AI on one process that's ready.
Last updated:
The plan on one page
| Days | Focus | Done when |
|---|---|---|
| 1 to 30 | Stabilize and map | Every recurring decision has a name next to it |
| 31 to 60 | Codify and set authority | The top five processes are written, with limits |
| 61 to 100 | Rhythm and one pilot | The weekly meeting runs without the owner |
The days are a frame for the work. How far a company gets depends on where it starts.
Days 1 to 30: stabilize and map
Protect cash and customers first. Find out who collects receivables, who talks to the top accounts, and who signs off on payroll.
Then map where decisions sit. For one week, log every decision that reaches the owner: what was asked, who asked, and whether it has come up before. In a founder-led company, the list is long.
Every company has a Bob. Somebody asks how we do something, and the answer is: "just ask Bob." If it lives in Bob's head, the AI will ask Bob too. The map shows who your Bobs are.
Days 31 to 60: codify and set authority
Write down the five processes the business can't run without. For each one, record the decision as well as the steps: what good looks like, who decides, and up to what amount. We start from a library of pre-built SOPs and fit them to how the company actually works.
Then set authority limits. Every recurring decision gets one owner and one number, written into a decision authority matrix. For example, refunds up to $500 with the office manager, discounts up to 10% with the sales lead. Your numbers will be different.
Days 61 to 100: rhythm and one pilot
Stop approving one request at a time. Review once a week. One weekly operations meeting: the numbers, what went over a limit, and who decided.
Then pick one written process with a clear owner and clean data, and pilot AI on it. Every agent runs until it hits a gate a person still owns. If no process is ready yet, the pilot waits. AI scales whatever's already there.
What the owner does in each phase
- Days 1 to 30: explains the decisions. The owner is the main source for the map.
- Days 31 to 60: reviews the written processes and agrees to the limits.
- Days 61 to 100: steps out of approvals inside the limits and joins the weekly review.
The owner's logic and judgment need to be written down too, because that's what made the company great in the first place.
Metrics to watch
Track a short list of weekly KPIs, each with one owner. Watch two more signals during the plan: how many decisions still reach the owner each week, and how many processes are written down and in use. When the first number falls and the second rises, the system is taking hold.
Where this sits in SCOPE™ by Efficiency Sprint
The plan follows the first states of SCOPE™ by Efficiency Sprint: Stabilize repairs the basics, Codify makes the recipe clear, and Ownership means people run it reliably. AI starts at Pulse, where it watches and reports, and acts on its own at Engine.
We install systems in companies that make them run without owners and ready for automation. Efficiency Sprint makes companies operationally ready for AI. The ongoing rhythm can run on EfficiOS, our subscription. For a deeper look before the deal closes, see operational due diligence, or how we work with private equity.
The operational AI readiness check takes about two minutes and shows where a company stands.
Questions
Should the founder stay on after the acquisition?
For the first phases, yes, if they can. The founder holds the decisions the map needs. Plan for the role to change from approving everything to reviewing once a week.
What should happen in the first week?
Confirm who controls cash and who talks to the top accounts. Start the decision log. Make no structural changes until you know where decisions sit.
When is it safe to introduce AI?
When one process is written down, has one owner with a limit, and runs on data you trust. If the work can't go to a good operator without explaining it every time, it can't go to AI either.
Who owns the plan?
One person on the new owner's side, usually the operating partner, with the founder as the main source. You always need to know who is responsible. And it can't be AI.