What a VP of Marketing Should Demand from Their Agency (And Usually Doesn't)

Most B2B agency relationships produce activity without pipeline. Weekly reports showing campaigns running, creative produced, posts published. Quarterly business reviews showing engagement metrics, CPLs, content performance. What they don't reliably produce is the thing the client actually needs: revenue that wouldn't otherwise have been there. The gap between what agencies deliver and what clients need isn't primarily about agency quality. It's about how the relationship was set up in the first place, what the client demanded, what the agency agreed to be accountable for, and what got let slide when it shouldn't have.

This article is written from the uncomfortable position of being an agency, writing about what clients should demand from agencies. Some of what follows will sound self-serving, some will sound self-critical. Both are intentional. The agencies worth working with are the ones that welcome the demands below; the ones that dodge them are the ones producing activity-without-pipeline.

Why Most Agency Relationships Underperform

Three specific setup failures that produce the activity-without-pipeline pattern.

The agency is paid for activity, not for outcomes. A fixed monthly retainer that doesn't vary with results produces incentive to maintain activity, hit the deliverable count, produce the reports, run the meetings, regardless of whether any of it is moving the business. The agency is economically rewarded for continuing the relationship, which is subtly different from being rewarded for making the business grow.

Scope is defined by deliverables, not commercial outcomes. The contract specifies "four blog posts per month, two paid media campaigns, weekly reporting." It doesn't specify "pipeline produced," "MQLs delivered at target cost," or "revenue influenced." Agencies deliver what's in the contract and clients pay for it. The deliverable gets produced. The outcome may or may not follow.

Accountability is reviewed quarterly at most. QBRs happen every 90 days. Between them, the relationship operates on trust that the work is producing returns. Three months is long enough for meaningful problems to develop without being surfaced. By the time the QBR arrives, the problems have accumulated beyond what a single meeting can address.

Most agency relationships fail on one or two of these; the worst fail on all three. Fixing any of them improves the relationship; fixing all three produces partnerships that actually move businesses forward.

What Clients Should Actually Demand

Five demands that change the economics and outcomes of agency relationships. Each one is uncomfortable for agencies to accept. Agencies that accept them are the ones capable of producing what clients need.

1. A specific commercial outcome the agency is accountable for

Not a deliverable count. A specific commercial outcome, pipeline generated, MQLs produced at a target cost, SQLs generated for sales, revenue influenced, retention improved. Something measurable, attributable, and relevant to the client's actual commercial situation.

What this looks like in a contract: "The agency will produce 40-60 MQLs per month within 6 months, at a blended cost below £300 per MQL, with lead-to-opportunity conversion above 15%." Specific numbers. Specific timeframe. Specific performance bar.

What agencies commonly do instead: "The agency will run LinkedIn Ads, produce content, and manage CRM workflows." These are inputs. They describe what the agency will do, not what it will produce.

Clients should demand outcome accountability. If an agency refuses to commit to a specific outcome, the client should take that as the diagnostic signal it is, the agency isn't confident they can produce outcomes for this business, or isn't willing to be held to them if they try.

2. A diagnostic before any commitment to execution

Before committing to 6-12 months of retained work, the agency should produce a diagnostic that demonstrates they understand the specific constraints this business faces. Not a discovery call, not a deck, a written document that shows where the funnel is breaking, why, and what would need to change for growth to compound.

Agencies that can produce this diagnostic well are the ones likely to produce good execution. Agencies that move straight to tactics without diagnostic capability usually produce tactical activity that isn't solving the right problem.

The practical version: a 30-day paid diagnostic engagement as the entry point to any longer retainer. The cost is modest compared to a year of misdirected retainer work. The output is either a diagnostic that shapes productive subsequent work, or evidence that this agency isn't the right partner, both useful outcomes.

3. Transparent access to the data and the work

The client should have direct access to the ad platforms, the CRM, the analytics, and the ongoing work product. Not agency-mediated reporting that presents polished versions. Raw access.

This demand is uncomfortable because it means the client can see the work in real time, including the messy parts. Agencies that resist it usually have something about the work they don't want visible, underperformance, quick fixes, late deliveries, team changes they haven't announced.

Agencies that welcome transparency produce better outcomes because there's no gap between what's happening and what the client sees. Problems get surfaced early. Successes get recognised immediately. The relationship operates on real information rather than managed information.

4. A monthly review with real data, not quarterly with polished narrative

Quarterly business reviews are theatre. By the time three months have passed, the narrative has been constructed and polished. Data that doesn't fit the narrative has been explained. The client is reviewing a performance rather than reviewing the underlying reality.

Monthly reviews with unfiltered data produce a different relationship. Issues surface within 30 days rather than 90. Course corrections happen while they still matter. The agency has less space to let things drift without consequence.

The demand: monthly 60-90 minute reviews looking at raw data, with the agency explaining what they see, what concerns them, and what they're changing this month. Less polish, more candour. Quarterly reviews still happen but become the strategic conversations they should be rather than the performance evaluations they've become.

5. A named accountable partner, not an account team

Agencies typically staff accounts with multiple people: strategist, specialist, account manager, analyst. This produces institutional memory resilience but also diffuses accountability. When something goes wrong, it's not clear who owns the fix.

The demand: one named partner who owns the commercial outcome of the engagement. Other team members work the tactics; the named partner is accountable for whether the work is producing the outcome the client is paying for. When the outcome slips, the named partner is the one who shows up to explain what's happening and what's changing.

This demand tends to upset agency account structures. It also produces relationships that actually work. The client knows who to escalate to. The agency knows who's on the hook. Diffuse accountability gets replaced with named accountability.

What Agencies Often Do In Response

When clients start making these demands, agencies respond in one of three ways.

Some agencies refuse or deflect. They explain why outcome accountability isn't possible in their model, why access to data is restricted by their tooling, why monthly reviews would be operationally difficult. These agencies are signalling that their business model isn't compatible with outcome accountability, which is diagnostic.

Some agencies partially accept. They agree to some demands and negotiate others. "We can commit to MQL volume but not lead-to-opportunity conversion because that's influenced by sales." "We can provide access to the ad platforms but not the CRM." Partial acceptance usually produces partial alignment, better than none but leaves the underperformance patterns partly intact.

Some agencies welcome the demands. They're already operating this way with their best clients and recognise that the demands match how good partnerships actually work. These agencies tend to have longer average client tenure, more direct commercial outcomes to cite, and fewer clients-who-technically-renew-but-aren't-really-growing.

What Clients Often Don't Demand

Worth naming the things clients should demand but usually don't, because their absence is what produces the low-accountability agency relationship pattern.

Specific numeric targets at six and twelve months. Clients often accept agency commitments to "grow pipeline" without specifying by how much. The specificity makes the commitment real.

Pre-agreed escalation points. What happens if CPL is 40% over target at month three? What happens if MQL volume is 50% below target at month six? Pre-agreed responses prevent the situation where underperformance drifts without anyone forcing a decision.

Exit clauses tied to performance. Most agency contracts have a 30-60 day notice period but no explicit link between performance and exit rights. Adding "if agreed KPIs are missed by more than 30% for two consecutive months, the client may exit without notice" changes the urgency with which the agency responds to early underperformance.

Including these terms up front is uncomfortable. It's also what distinguishes serious commercial relationships from vendor-client ones where the vendor is essentially running a subscription business.

Where This Sits

This piece is written for the buyer, VP of Marketing, Head of Growth, CMO, or founder who's about to hire an agency or evaluate the one currently engaged. The demands above reshape the economics of the relationship in favour of outcomes rather than activity.

From the agency side (where NBT sits), accepting these demands is how we make sure our clients actually get what they're paying for. A client who demands outcome accountability, diagnostic clarity, transparent access, monthly candour, and a named accountable partner is a client we can produce real results for. The relationships that go wrong are the ones where these demands weren't made and the work drifted into activity-without-pipeline by default.

If you're selecting an agency or evaluating one currently engaged, and you're uncertain what to demand or how to tell whether the relationship is actually working, get in touch for a strategy call. We'll walk through the specific evaluation framework for your situation and help you ask the questions that matter.