---
title: "Why mid-market funds can't hire an AI operating partner"
description: "Why a full-time hire rarely pays off for a lower-mid-market fund, and how to structure a fractional AI operating partner so the portfolio actually ships."
url: https://anywayventures.com/insights/why-mid-market-funds-cant-hire-an-ai-operating-partner
updated: 2026-09-29
---

[← ALL WRITING](https://anywayventures.com/insights) OPERATING PARTNER

# Why mid-market funds can't hire an AI operating partner

The people who can do the job are scarce, expensive, and mostly don't want a full-time seat at a £300m fund. There's a better structure, and it costs far less per portfolio company.

WRITTENAugust 2026
READING · Nine minutes
FOR · Fund partners, portfolio directors
RELATES TO[Fractional AI operating partner](https://anywayventures.com/ai-operating-partner)

Large-cap funds are building AI operating teams. Lower-mid-market funds, with six to fifteen portfolio companies and a small operating bench, mostly aren't. That isn't because they don't see the need. It's because the economics of the full-time hire don't work, and a fund that tries it can spend a year and half a million pounds before anything runs. A fractional AI operating partner, retained by the fund, deployed by the deal team where the value case is strongest and backed by delivery bought on fixed scopes, is the better structure for a fund that size.

The workaround most funds have landed on is a workshop. Bring in a consultancy, run an AI strategy day with each management team, produce a roadmap. Six months later very little is in production. A roadmap is not a capability, and management teams were already busy.

> A roadmap is not a capability, and management teams were already busy.

I've set out what the role contains [in a separate piece on the AI operating partner in private equity](https://anywayventures.com/insights/ai-operating-partner-private-equity) , so I won't repeat the job description here. This one is about money and structure: what the full-time seat really costs, why the strategy day fails, and how to set up the fractional version so it survives contact with a real portfolio.

## Why the full-time hire doesn't pencil out

[Korn Ferry](https://www.kornferry.com/institute/the-ai-operating-partner-the-latest-pe-portfolio-value-creation-role) notes that large PE firms have created AI operating partner roles, filled either by full-time employees or by part-time advisers. Large firms can spread that seat across dozens of companies.

A £300m fund needs the capability just as much. The [Bain & Company and StepStone Group 2026 Private Equity GP Outlook](https://www.bain.com/insights/private-equitys-reality-check-gp-outlook-2026/) found GPs get their best returns from generative AI in their own deal sourcing and due diligence, while nearly 40% don't expect material financial impact from AI in their portfolio companies in 2026. The fund's own team is using AI well. The companies it owns mostly aren't.

The bottleneck is people. [FTI Consulting's 2026 Private Equity AI Radar](https://www.fticonsulting.com/insights/reports/2026-private-equity-ai-radar) found only 36% of portfolio companies deploy AI in production with real operational impact, and talent was the top constraint to scaling, named by 35%. The person a mid-market fund needs has run a P&L, can sit opposite a sceptical CEO, and has personally shipped AI systems that are still running. That's a short list. Everyone on it has options.

And the work is lumpy. In any quarter, two companies need heavy delivery, five need a monthly conversation and a nudge, and the rest need very little. A full-time salary pays for the quiet months at the same rate as the busy ones.

## What an AI operating partner costs, on illustrative numbers

The figures below are assumptions I've chosen to make the arithmetic visible. They are not market data, and your numbers will differ. Swap in your own and see whether the conclusion moves.

Picture a fund with ten portfolio companies. For the full-time route, assume a base of £200k, a bonus target of 50%, employer on-costs of 15% on salary and bonus, and a search fee of 25% of base. Assume the hire, sensibly, asks for one senior engineer so they can build as well as advise, at £110k all in. Add £25k for tools, travel and expenses.

For the fractional route, assume a retainer of £10k a month for 3 days a month, £10k of expenses, and four fixed-price delivery sprints in year one at £30k each, placed wherever the deal team deploys the partner.

Year one, ten portfolio companies
Full-time hire
Fractional plus bought delivery

Base salary
£200k
none

Bonus (50% of base)
£100k
none

Employer on-costs (15%)
£45k
none

Search fee (25% of base)
£50k
none

First team hire, one engineer
£110k
none

Operating partner retainer (12 at £10k)
none
£120k

Delivery sprints (4 at £30k)
not included
£120k

Tools, travel, expenses
£25k
£10k

Carry or co-invest
expected
none

Total cash cost
£530k
£250k

Per portfolio company
£53k
£25k

Realistic first thing in production
month 9 to 12
month 2 to 3

The arithmetic: £200k plus £100k plus £45k plus £50k plus £110k plus £25k is £530k, or £53k per company. On the other side, £120k plus £120k plus £10k is £250k, or £25k per company.

Two things make the gap wider than the totals suggest. First, the full-time column still hasn't bought any delivery beyond one engineer, and one engineer cannot build across ten companies. In practice the full-time partner buys sprints too, so add most of that £120k to the left-hand column. Second, if the portfolio companies pay for their own sprints (more on that below), the fund's own cost for the fractional route falls to £130k, or £13k per company.

Then there's carry. A credible candidate will expect points from the carry pool or a co-invest allocation, because that's what an operating partner at a fund gets. It isn't a year-one cash cost, but it comes out of the same pool as the deal team's, and it's a long commitment to someone the fund hasn't yet seen work.

## Recruitment time, time to first result, and the leaving problem

Cost is only half of it. Time matters more in a hold of four or five years.

Assume three to four months to run a search, three months of notice, and a quarter spent learning ten businesses before anyone commits to building anything. That puts the first system in production somewhere around month nine to twelve. On a five-year hold, the fund has spent a fifth of it waiting.

A fractional partner who has done the job before can be deployed into the first priority company within weeks and have a scoped sprint running there by the end of the second month.

Now the uncomfortable scenario. The full-time hire leaves at month eighteen, which senior people with options do. The fund has spent the best part of £800k on the illustrative numbers, the relationships with ten CEOs walk out of the door, and the search starts again. Anything undocumented quietly stops.

A fractional arrangement fails more cheaply. If it isn't working at month six, you end it and you've spent a quarter of the money.

## Why the strategy day per portfolio company fails

The common workaround looks sensible on paper. A consultancy runs a day with each management team, interviews a few heads of department, and leaves behind a prioritised roadmap.

Then nothing much happens. The roadmap names the right workflows, but nobody in the business has the time or the skills to build them, and the consultancy has moved on. The CEO is running the 100-day plan, the finance director is on the refinancing, and the one person who understood the idea is back in their day job.

The wider evidence says pilots die in exactly this gap. The [MIT NANDA study reported by Fortune](https://fortune.com/2025/08/18/mit-report-95-percent-generative-ai-pilots-at-companies-failing-cfo/) found 95% of generative AI pilots it examined showed no measurable P&L impact, and put the core cause down to a "learning gap": systems that don't retain feedback or adapt to how work is actually done. Critics fairly point out that many of those pilots had no baseline measured before launch, which is its own lesson. The same study found buying from specialised vendors and partnering succeeded about 67% of the time, while internal builds succeeded about a third as often. Help that stays long enough to finish the job matters.

[S&P Global Market Intelligence](https://www.ciodive.com/news/AI-project-fail-data-SPGlobal/742590/) found 42% of companies abandoned most of their AI initiatives in 2025, up from 17% in 2024, and the average organisation scrapped 46% of its proofs of concept before production. A roadmap handed to a busy management team is a proof of concept waiting to be scrapped.

Ten strategy days with write-ups can cost as much as a year of fractional cover, with nothing running at the end.

## How to structure a fractional AI operating partner arrangement

The fractional model works when it's written down properly. Failures come from vague scope, days spread so thinly that nothing ships, and nobody owning delivery. Agree these terms before anyone starts.

Retained days, deployed by the deal team. The fund reserves a set number of senior days each month, set by portfolio size and how much is in the pipeline. The deal team decides where they go: a deal in exclusivity, a new acquisition's 100-day plan, a company behind plan, or a business 12 to 18 months from exit that needs evidence. One priority at a time, with the operating partner on call to the deal team in between. Agree a simple way to flex the days for a quarter when a deployment needs more.

A brief per deployment. Before going in, the deal partner and the operating partner agree one page: the outcome, the baseline it will be measured against, and the exit criteria that say when the capacity moves on. Then a one-week diagnostic inside the business, build with management (the method is in [the AI value-creation plan](https://anywayventures.com/insights/ai-value-creation-plan) ), handover, and on to the next priority. There is no baseline of every company and no monthly session with every CEO. That's the evenly spread version, and it's how the strategy day comes back in through the side door.

In scope: pipeline questions from the deal team, deployment briefs, hands-on diagnostics and build oversight inside the chosen company, vendor selection, and oversight of every delivery sprint. Out of scope: running IT, cyber security, ERP selection, acting as any company's CTO, and full deal-side diligence, which should be priced separately as [AI due diligence](https://anywayventures.com/ai-due-diligence) when a deal needs it.

Delivery bought on fixed prices. Each piece of build work is a fixed-price sprint of 2 to 8 weeks inside one company, with a named owner in the business, a before-and-after measure and a handover date.

Who pays. Whoever holds the benefit. If a sprint returns hours to one company's finance team, that company pays, against a plan its CEO has signed. If the work serves the fund, such as diligence on a deal or a pattern three companies will reuse, the fund pays. Check your LPA and fee arrangements before charging anything to portfolio companies.

Reporting per deployment. One page to the deal partner at the end of each deployment: baseline, what went live, hours returned to the business per month, run cost, and what is still running six months on. Then a quarterly summary for the IC across every deployment, same format every time, including what was stopped and why and where the capacity goes next.

Term. Twelve months, reviewed at six. Long enough to ship, short enough that nobody is trapped.

Conflicts and vendor independence. The operating partner should disclose every commercial relationship with a software vendor and take no referral fees. If their own firm also delivers sprints, as mine does through [embedded AI delivery](https://anywayventures.com/embedded-ai-delivery) , say so in the engagement letter, show fixed prices to the IC before work starts, and keep the fund's right to put any sprint out to tender. Independence is proven by the recommendations, including the ones that say don't build anything yet.

## Testing it at six and twelve months, and how it ends

At six months, ask four plain questions. Did every deployment start with a signed brief and a baseline? Did the deal team choose where the days went, and would it choose the same again? Is at least one thing running in production and measured? Has anything been stopped? If the answer to the third is no, the arrangement is producing roadmaps, and you've bought the strategy day again at a monthly rate.

At twelve months the question narrows to money and durability. Hours returned to the business per month across every deployment, net of run costs, and what is still running, unattended, six months after it shipped.

Every fractional arrangement should also say how it ends. One route is handover. Each company has named owners for what was built, run books, and a supplier relationship it controls, so the systems keep working when the operating partner stops turning up. The other route is growth. When the portfolio gets large enough, or software-heavy enough, that a full-time seat pays for itself, the fractional partner should help write that job description and run the handover to whoever takes it.

## What the fund gets from one retained AI operating partner

The first thing is accountability. One person answers for AI across the fund, from diligence on a deal to the evidence pack at exit, and goes wherever the deal team needs them. The IC has one name to ask.

The second is transfer. Portfolio companies in a mid-market fund often share more than they think: similar back offices, similar customer service queues, similar quote-to-cash processes. Take a supplier invoice matching workflow built for a £40m distribution business. Rebuilt for a second portfolio company, it's mostly configuration and data work, so the second sprint might cost half the first and ship in half the time. A consultancy running separate strategy days never sees the pattern. An operating partner deployed into one company after another sees it the second time round.

The third is exit evidence. A buyer's diligence team will mark down anything described as AI-enabled that can't be shown working. Systems in production, usage logs, before and after measures and the cost to run them go into the equity story. Assertions don't.

And the portfolio companies get a counterpart who has sat in their chair. I ran Netsells as CEO for thirteen years and later ran YourParkingSpace as CEO. What management teams want from the fund on AI is someone who will say which idea to kill, then help build the one that's left.

For a fund with ten companies, the realistic comparison is £530k for a seat that starts delivering in year two against roughly £250k for something running this quarter. If you want to test the numbers against your own portfolio, the [fractional AI operating partner](https://anywayventures.com/ai-operating-partner) page sets out how we work, and the first call is free.

## Questions

### How much does an AI operating partner cost?

It depends heavily on the person and the fund, so treat any figure as an assumption. On illustrative numbers, a full-time hire with bonus, on-costs, a recruitment fee and one engineer comes to around £530k in year one before carry. A fractional partner at 3 days a month plus four fixed-price delivery sprints comes to around £250k, and the portfolio companies can fund the sprints.

### Can a lower-mid-market PE fund hire a full-time AI operating partner?

It can, but with 6 to 15 portfolio companies the numbers rarely work. The people who can do the job expect partner-level pay and carry, want a team, take months to recruit and can leave with the knowledge. Most funds that size get more for less from a fractional operating partner with delivery bought on fixed scopes, then hire full-time once the portfolio justifies it.

### How should a fractional AI operating partner arrangement be structured?

The fund retains a set number of senior days each month, and the deal team deploys them into whichever target or portfolio company has the strongest AI value case, one priority at a time. Each deployment gets a brief with an outcome, baseline and exit criteria, and ends with a one-page report to the deal partner. Build work is a fixed-price sprint. Set a 12 month term, reviewed at six months, with vendor conflicts disclosed.

### Who should pay for AI delivery in a portfolio company, the fund or the company?

As a rule, whoever holds the benefit pays. Work that returns hours or revenue inside one company's P&L should be paid by that company, against a plan its CEO has signed. The retainer itself, diligence support on the pipeline and patterns several companies will reuse sit more naturally with the fund. Check the LPA and your fee arrangements before charging anything to portfolio companies.

Brannan Coady
Founder of Anyway Ventures. Exited founder, former CEO of YourParkingSpace, investor and board member; now building AI inside PE-backed and mid-market businesses. [More →](https://anywayventures.com/brannan-coady)

## Keep reading

[ALL WRITING →](https://anywayventures.com/insights)
[OPERATING PARTNER · 10 MIN What does an AI operating partner do in private equity? Plenty of funds now say they want one. Far fewer can describe the job. Here is what it contains, how a fund should deploy it, and how to tell in year one whether it's working.](https://anywayventures.com/insights/ai-operating-partner-private-equity) [OPERATING PARTNER · 10 MIN How to build an AI value creation plan for a portfolio company A working method, step by step: where the hours go, what to rank first, how to size it in pounds, and how to report it so the exit story is evidenced rather than asserted.](https://anywayventures.com/insights/ai-value-creation-plan) [TRAINING · 9 MIN How to train staff on AI so the work actually changes Most AI training is a lecture or a prompt session, and on Monday the work looks the same. Here is the programme I would run instead: small cohorts, real tasks, a working build per person and a six-week check.](https://anywayventures.com/insights/how-to-train-staff-on-ai)