Agencies underperform for two reasons — and one is yours

When an agency relationship disappoints, the instinct is to blame the agency. Often the fault is shared: agencies underperform as much from weak client management — vague briefs, no clear success metric, no accountability, junior day-to-day contact — as from weak work. The good news is that the client-side factors are the ones you control. Managing agencies well is a skill, and it materially changes the return you get. Having managed significant agency relationships across paid, CRM and creative, [APPROVAL NEEDED] here’s what actually works.

Brief for outcomes, not tasks

The single biggest lever is the brief. Brief an agency on the outcome you need and the commercial metric it maps to — not a list of deliverables. A task brief gets you tasks; an outcome brief gets you thinking and accountability. Give them the context, the constraint and the number, and let their expertise find the route.

Set one success metric and hold to it

Agree, up front, the commercial metric the relationship is judged on — CAC, qualified pipeline, contribution — and review against it. Without one agreed number, reviews become debates about activity, and no one is accountable for results. The metric is what turns an agency from a vendor into a partner.

Insist on seniority where it matters

The classic agency failure is a senior pitch and junior delivery. Insist on the senior involvement you were sold at the moments that matter — strategy, planning, problem-solving — and accept junior delivery only where it’s genuinely fine. You’re paying for expertise; make sure you get it.

Manage the relationship actively

Good agency relationships are managed, not left. Regular reviews against the metric, fast feedback, shared visibility of results, and a genuine partnership tone (not adversarial, not passive) get far more from an agency than an annual check-in. This is exactly the kind of oversight a fractional growth leader provides — someone senior, on your side, holding agencies to the outcome. It often pays for itself in improved agency return alone.

Not getting enough from your agencies? We manage partners to the outcome, on your side. Book a discovery call →

The real question: seniority, ownership, cost, speed

How you resource growth is one of the most consequential decisions a scaling business makes, and it’s usually made by default rather than design. The honest way to decide is to weigh four things for each option: the seniority you get, who owns the outcome, the cost, and the speed to value. Here’s how the options actually compare.

In-house

Building an in-house team gives you dedicated focus, deep context, and long-term capability. The trade-offs are cost, the time and risk of hiring (especially senior leadership), and the danger of a team that executes well but lacks strategic direction. In-house is right when you have the scale to justify permanent headcount and, crucially, senior leadership to direct it.

Agency

Agencies give you capability and capacity quickly, without hiring. The common failure modes: a senior pitch and junior delivery, execution optimised to the agency’s scope rather than your business outcome, and no one truly accountable for your number. Agencies work well for specialist execution under someone (in-house or fractional) who owns the strategy and holds them to it.

Freelancers

Freelancers are flexible and cost-effective for a specific, well-defined need — a channel, a project, a skill. The limits are narrow scope, no strategic ownership, and coordination overhead if you string several together. Good for filling a defined gap, not for leading growth.

Fractional leadership — the missing middle

A fractional growth leader / CMO gives you senior strategy and accountability without a full-time hire — the option most businesses overlook. You get an experienced operator owning the number, setting strategy, and managing agencies and in-house alike, at a fraction of the cost and lead time of a permanent CMO. It’s the natural fit for the common situation: you’ve outgrown freelancers and agencies-without-direction, but you’re not ready for a £150k+ full-time CMO.

Match the structure to your stage

• Early / lean: freelancers or a fractional leader for direction; avoid premature full-time hires.
• Scaling, no senior head: fractional leadership to set strategy and manage delivery (in-house or agency).
• At scale: in-house team, ideally led by a senior head (a fractional leader can bridge until you hire and even help you hire).

Many of the best setups are hybrids — a senior leader (fractional or in-house) directing a mix of in-house specialists and agencies. The structure should serve the outcome, not the org chart.

Not sure how to resource your growth? Let’s talk it through. Book a discovery call → or explore the fractional growth leader model.

Past the hype, into the workflow

The conversation about AI in marketing is mostly noise in both directions — overblown promises and reflexive scepticism. The useful question for a marketing leader is narrower and more practical: which specific workflows does AI genuinely accelerate, and which still need human judgement? Used well, AI-assisted workflows compress turnaround time and free senior capacity for higher-value work; used as a gimmick, they just produce more mediocre output faster. Here’s where it actually helps.

Where AI saves real time

• Creative production and iteration. Generating and varying creative concepts, ad variants and formats — feeding a structured testing pipeline far faster. We’ve cut creative turnaround roughly in half this way, which matters because velocity of testing drives rate of improvement. APPROVAL NEEDED] (See [creative testing.)
• Copy variants. Producing first drafts and multiple variations of ad copy, subject lines and landing-page sections for testing — a human still edits and decides.
• Reporting and summarisation. Turning raw data into first-draft summaries and surfacing patterns, cutting the manual reporting load.
• Experiment analysis. Accelerating the read of results and the generation of next hypotheses — with human sign-off on the decision.

Where human judgement stays

AI is a velocity multiplier on production, not a replacement for judgement. The strategy, the hypothesis, the read of a result, the brand and the commercial decision remain human. The failure mode is using AI to scale output without the discipline behind it — more content, more variants, more noise, none of it better. The value comes from accelerating the safe-to-accelerate parts while keeping the judgement where it belongs.

The operating-model implication

AI-assisted workflows are part of a modern marketing operating model, not a bolt-on. Embedding them well means redesigning the workflow (who does what, where AI slots in, where the human checkpoint sits), not just handing the team a tool. Done properly, it doesn’t shrink the team’s value — it moves it up, from producing to deciding.

Want to embed AI where it actually saves time — not as a gimmick? A Growth Diagnostic can assess your workflows. Request a Growth Diagnostic →

Plenty of businesses have a reasonable strategy and still can’t execute — because the operating model is wrong. Teams are structured around channels, data is fragmented, decisions are slow.

The four components

People and structure (organised around outcomes, not channels); process and cadence; data and measurement (shared standards, a single view); and technology and workflows, including AI where it genuinely saves time.

Structure around outcomes

The most common flaw is organising by channel, each team optimising its own metric while no one owns the outcome. Restructuring around outcomes ends the silo wars.

Rigour and pace together

Shared standards enable speed by removing the need to re-decide everything; a clear cadence creates pace rather than constraining it.