Sooner or later, most marketing leaders consider bringing paid media in-house. The economics make sense on a spreadsheet. The reasons are usually some combination of: agency fees feel high relative to output, response times have stretched, the senior person you originally signed with no longer touches the work, and the in-house team feels like they could do better if they just had access.

Sometimes that's true. Sometimes the move actually does work. But more often than not, the transition costs more than people expected and the performance never quite recovers. This piece is about why - and how to do it properly when it is the right call.

Why most in-house moves go badly

The pattern we've seen, almost always:

1. The decision gets made on cost, not capability. "We'll save £X by going in-house" is a starting calculation that almost never includes the real cost of building capability from scratch.
2. One senior hire is treated as the solution. Senior practitioners are scarce and expensive. The role gets filled with a strong mid-level operator who's never had to build measurement infrastructure or restructure a complex account from scratch.
3. The agency hands over the keys but not the capability. Account access and a recorded handover call. No documentation of why the structure is what it is. Six months later, the in-house team can't explain decisions made before they arrived.
4. Performance degrades slowly. Not in week one - usually month three or four, when the routine work is fine but the strategic decisions stop being made. By month nine, year-on-year is noticeably worse and no one can quite explain why.

What an in-house transition that actually works looks like

The version that holds up - the version we've now run as the "Educate" arm of several engagements - looks fundamentally different. The shape:

1. Capability mapped before hire

Before any hiring, document what the in-house team will need to be able to do. Not a job description - a capability matrix. Across paid search, paid social, and analytics, what's the level of skill required? Who'll do feed work? Who owns measurement? Who'll handle creative testing programmes?

Most "head of paid media" hires can't do all of these well. The matrix tells you whether one hire is enough, or whether you need two, or whether the model is one in-house lead with an external partner for analytics or creative.

2. Structured handover, not a key transfer

The handover documents the thinking, not just the access. For each major account decision, why was the campaign structure chosen? What was tested and rejected? What's known to underperform? What are the bidding strategies tuned for, and why?

Done well, this is a 30-50 page document plus a series of working sessions. Done badly, it's a recorded call and a Loom walkthrough.

3. Six-month parallel period

The first six months after the hire, the agency stays involved - on a reduced scope, in a clearly-defined oversight role. Weekly review calls, monthly strategy sessions, and a clear escalation path when the in-house team hits something they haven't seen before.

This is the single most important phase. Six months is roughly how long it takes for the in-house team to encounter every category of decision the agency was previously making. Without an oversight relationship during that period, mistakes happen alone and don't get corrected.

4. Capability lift, not just delivery

The "Educate" engagement model only works if capability genuinely transfers. That means structured sessions on the topics most in-house teams haven't seen up close: measurement architecture, advanced bidding, account restructuring, audience signal strategy, creative testing programmes. Not webinars - working sessions on the actual account.

5. Honest exit criteria

The agency engagement should end when specific capabilities are demonstrably in place - not on a fixed date. The criteria should be agreed up front: "in-house team can independently scope and execute X, Y, Z without external escalation." This protects both sides.

When to actually do it

The honest truth is that in-house works really well for some businesses and really badly for others. The factors that make it more likely to succeed:

• Single-product or narrow product range. Less complexity to manage, fewer specialist disciplines needed.
• Spend large enough to justify multiple senior hires. Below £500k/year in agency fees, the cost-benefit is rarely favourable.
• Marketing leadership with paid-media depth. Someone above the in-house team who can actually evaluate the work, not just receive reports.
• Long-term commitment to the team. If churn is high, the capability investment evaporates with each leaver.

The factors that make it likely to go badly:

• Multi-product or multi-region complexity. The breadth of expertise required usually exceeds what one or two hires can provide.
• "We'll save money" as the primary driver. The savings are real but smaller than expected; the performance risk is bigger than expected.
• Marketing leadership without paid-media depth. Without someone who can spot drift, drift goes uncorrected for months.

What to do this week

• Map the capability matrix - what skills does your in-house team need, what level, across which disciplines? Be specific.
• Be honest about your current marketing leadership's depth in paid media. Can they spot drift? If not, the in-house move is higher-risk than it looks.
• Cost the transition properly - hiring, infrastructure, the parallel-running period, the inevitable performance dip during ramp-up. Compare against current agency cost over a 24-month horizon.
• Talk to your agency about a re-scoped engagement before deciding. The conversation usually reveals options that weren't visible.

The Educate arm of our work is built specifically for clients who want their in-house team running paid media confidently - either as the destination state or as a phased transition over time. If that's the conversation you're having internally, the strategic review is a sensible starting point.