AI in the 100-Day Plan: What PE Operating Partners Should Do After Close

AI is showing up in more 100-day plans. It usually arrives as a line item: roll out licenses, clean up the data, run a pilot in one team.

By day 100, the licenses are issued and the pilot has its own slide in the board deck. Ask how many people at the company actually work differently, and the honest answer is often very few. The window after close, when management expects change and the plan still has room for it, has quietly shut.

It can go another way. When the plan puts people first, the first 100 days can end with one workflow running, a team that trusts it, and a number the board can see.

This post lays out a three-phase plan for AI in the first 100 days, what operating partners should own in each phase, and what to hand over when day 100 arrives.

Why the First 100 Days Are the Right Time for AI

Management expects change right after close, and the 100-day plan sets the priorities for those first 100 days. With new owners and new priorities on the table, a new way of working is easier to introduce than at any point later in the hold.

The firms getting results treat it that way. In a Private Equity International interview, KPMG's James Gardner says the technology has to be part of the 100-day plan as soon as a deal completes, and that most of their clients are hands-on from day one.

BCG's survey of 100 senior PE investors found the most successful firms set up a dedicated digital team right after acquisition, which spends the first 100 days laying the groundwork AI will need. It also found this work delivers stronger results when it's part of the value creation plan from the start, instead of being added after close.

Waiting has a real cost. In the same BCG survey, 90% of investors named competing priorities as the top blocker to digital work. Once the value creation plan is running, AI has to fight every other workstream for attention.

Where Most 100-Day Plans Go Wrong With AI

The common mistake is treating AI as an IT project. The plan covers tools, data, and a pilot, then counts licenses issued as progress. It rarely says who will work differently, or how anyone will know.

Three patterns show up again and again:

• Pilots built for the board deck. KPMG's Gardner describes firms that, a year ago, had 20 AI pilots running mainly to show activity to the board. The stronger firms now wire business targets straight into the value creation plan.
• Capability that leaves with the consultants. BCG found that only 45% of successful firms systematically transfer knowledge from external partners to their own teams. Without that, the portfolio company stays dependent on outside help long after the engagement ends.
• People questions left unanswered. Employees want to know whether AI will take their role, how to work with it, and what to do with the time it frees up. A plan that skips those questions gets polite compliance at best.

If you want a quick test for whether this is happening in your portfolio, the 8 questions operating partners should ask every portfolio company separate real change from license counts.

We've seen the slow version of this up close. A large portfolio company built its own internal AI tool and ran a pilot with around 20 users, while testing other add-ins on the side. People found the in-house tool couldn't handle the Excel and PowerPoint work they actually do, so adoption stalled.

The lesson for the 100-day plan is to put a usable set of tools in people's hands first. Save custom builds for the workflows that are worth the investment.

Days 1–30: Align on Where AI Fits

The first month is about getting a clear picture and making a few decisions. Nothing needs to be built yet.

Your job as operating partner: by day 30, make sure the company has an internal AI lead, a day-one baseline, and no more than two agreed use cases. Everything after builds on those three decisions.

• Audit readiness and current use. Map the tools people already work with, including the ones nobody approved. The five stages of AI adoption help you see where the company stands today.
• Set a baseline. Record how many people work with AI each week and how long the target workflows take today. Without a day-one number, day 100 has nothing to compare against.
• Name an internal AI lead. KPMG's Gardner recommends turning someone at the heart of the original process into the product owner and change champion. That person brings their team along far better than a mandate from the new owners.
• Pick one or two use cases. Our guide to finding the highest-value AI opportunities across your portfolio shows how to score them.
• Put a simple AI policy in place. Smaller portfolio companies rarely have the legal and security depth to write one, so give them a ready-made version. Our AI Policy Generator drafts one in minutes.

Want a fast starting point for the audit? Our two-minute AI Audit scores a company's readiness with instant results.

Days 31–60: Develop the Leadership Team First

The second month builds skills, starting at the top. When the CEO and leadership team work with AI themselves, everyone else can see it's expected. They also make better calls on where it belongs, and BCG makes the same point: technical teams can build the infrastructure, but only business leaders can spot the use cases worth continued investment.

Your job as operating partner: put leadership training on the calendar early in the month, and make sure the CEO takes part alongside their team.

Next, train the team behind your first use case, with sessions built around their roles. Everyone should know how AI affects their job, which tools are approved, and how to work with them responsibly.

Build on their own work, because that's where habits stick. At one PE firm we trained, two finance team members spent a few days building a Claude skill to check an outside adviser's tax work, with people reviewing every output. The prototype showed exactly what a production version would need, at a fraction of the cost of outside review.

Results come quickly. In our 12-week programs, people save 3–5 hours a week by weeks 3–4 and start building agentic workflows around weeks 7–8, so training that starts early in this phase pays off inside days 61–100. Teams that learn to build their own workflows also keep going after the advisors step back.

Days 61–100: Operationalize One Workflow

The last stretch puts one workflow into everyday use. One that runs every day beats five that live in a slide.

That also means resisting the urge to launch a generic company chatbot. As one PE operating lead put it to us, the value comes from rebuilding a process from the bottom up, with people and AI each doing the parts they're best at.

Keep the goal realistic. BCG's research suggests data and systems work often takes 6 to 12 months, so aim for a team workflow built on tools the company already has. Data-heavy projects can follow once that groundwork is in place.

Your job as operating partner: agree on the day-100 report with the CEO at the start of this phase, so the numbers are being collected before the board asks for them.

• Make it part of the job. The workflow should sit inside how the team already works, with a clear owner and a way to flag problems.
• Add governance that fits the risk. Decide who can approve new tools and what data stays off limits. Our AI governance framework guide covers the four parts every working framework needs.
• Check cost against value. AI isn't the answer to every problem. Sometimes process redesign or better use of existing systems gives a faster, lower-risk return, and it's fine to say so.

The day-100 report should show two things: how people work now compared with day one, and what that did to the business metric you chose. Usage without results looks like activity. Results without usage won't last.

Here's what that can look like. A PE-owned trading business we worked with built structured knowledge cards so its team could find answers and get work done faster. Within about a quarter, it was hiring 6–7% fewer people than planned, because its existing team was getting more done.

Frequently Asked Questions

Should AI be in the 100-day plan if the company's data isn't ready?

Yes, with a narrower goal. Start with team workflows built on the tools the company already has, and save data-heavy projects for when the groundwork is ready.

Who should own AI in the first 100 days?

The operating partner owns it in the value creation plan, the CEO sponsors it, and an internal AI lead runs it day to day. BCG found the most successful transformations give C-level executives clear ownership and accountability.

How should operating partners report AI progress to the investment committee?

Use the same two numbers as the day-100 report. BCG found only 11% of firms explicitly link digital progress to their exit story, so tying AI results to value creation early puts you ahead.

How is a 100-day AI plan different from a 90-day AI roadmap?

A 90-day roadmap works for any company starting with AI. A 100-day plan is specific to the post-close window in private equity, tied to the value creation plan and reported to the board. If you're building a general roadmap, our AI implementation roadmap guide walks through it.