Holding an Agency Accountable for Results

The commercial relationship between a business and its marketing agency has an awkward asymmetry. The agency knows considerably more about the channels, the tools, and the reporting than the client does, and the client is paying for outcomes they cannot easily verify. That gap is where dissatisfaction develops, usually not because anyone is acting badly but because nobody defined what success looked like.

The remedy is not closer supervision of activity. It is agreeing in advance what will be measured, how, and what the numbers need to reach, and then reviewing against that rather than against impressions.

A performance-driven marketing firm should welcome that framework rather than resist it, since accountability works in both directions. A defined measure protects an agency from being judged on someone’s impression of how things feel as much as it protects a client from paying for activity that produces nothing.

Defining the Outcome Before the Work

The conversation that prevents most problems happens before any campaign runs.

Establish the business objective in terms the business actually uses, meaning revenue, qualified leads, customer acquisition, or retention, rather than in marketing metrics.

Set the target specifically, including the figure and the timeframe, and be prepared for the answer that the budget cannot produce it.

Agree the measurement method, including which system is the source of truth, since discrepancies between platforms and internal systems are normal and cause enormous confusion when they surface later.

Baseline the current position, because improvement cannot be demonstrated without knowing the starting point and this step is skipped constantly.

Decide the review cadence and what happens at each review, including what would trigger a change of approach.

Identify the constraints honestly, including seasonality, capacity to handle increased demand, and anything in the business that would limit results regardless of marketing performance.

Metrics That Connect to the Business

The reporting should answer whether the investment is working, which many marketing reports do not.

Cost per acquisition, meaning what it costs to obtain a customer through a given channel, is the figure most directly comparable against value.

Customer lifetime value gives the context that makes acquisition cost interpretable, since a high cost per customer is fine if the customer is worth considerably more.

Return on advertising spend measures revenue against media investment, and it should be understood as excluding agency fees and other costs unless specified.

Conversion rates through the funnel show where losses occur, which directs where to work next.

Lead quality matters as much as volume, and an agency delivering many unqualified leads is producing activity rather than results. This requires feedback from sales, which the client has to provide.

Incrementality, meaning whether the activity produced results that would not have occurred anyway, is the hardest and most important question, particularly for branded search and retargeting.

The Attribution Problem

Measurement in marketing is genuinely difficult and pretending otherwise causes disputes.

Customer journeys involve multiple touchpoints, and assigning credit among them is a modelling choice rather than a fact.

Last-click attribution, the default in many systems, systematically overvalues whatever came last and undervalues everything that created awareness.

Platform-reported conversions are typically self-reported and generous, and summing them across platforms usually exceeds actual totals substantially.

Privacy changes have reduced tracking capability considerably, which means measurement is less precise than it was and any agency claiming otherwise is overselling.

Practical approaches include agreeing one measurement source, using holdout tests where volume allows, and asking customers directly how they found you, which is imperfect and useful.

The reasonable position is that measurement is directional rather than precise, and both parties should work from the same imperfect numbers rather than each citing the ones that favour them.

Reviewing the Relationship Properly

Regular reviews prevent the accumulation of unspoken dissatisfaction.

Review against agreed targets rather than against a general impression of activity.

Expect explanation when results fall short, and distinguish between an explanation and an excuse. A good agency identifies what is not working, says so, and proposes a change.

Watch for reporting that emphasizes favourable metrics while omitting the ones that were agreed, which is the most common warning sign.

Ask what has been learned, since a year of campaigns should have produced knowledge about the audience, the messaging, and the channels that is worth more than any single campaign result.

Provide the feedback the agency needs, particularly on lead quality and sales outcomes, since an agency optimizing without that information is working blind.

Discuss what is not working from your side too, since relationship problems are frequently mutual.

Structuring the Commercial Terms

The arrangement should align incentives without creating perverse ones.

Retainer arrangements provide continuity and can disconnect payment from outcome if no targets are attached.

Performance elements align interests and require carefully defined measurement, since disputes about whether a target was met are worse than having no target.

Media spend should be transparent, with clarity about whether the agency receives anything from media purchases beyond its fee.

Scope definition prevents the gradual expansion that strains relationships, and a process for handling additional work keeps it manageable.

Notice periods and exit terms matter, including ownership of accounts, data, and creative assets, which should be yours and frequently is not specified.

See also: Business Class Flights to Bangalore for Business and Leisure Travelers

Knowing When to Change

Some relationships should end and the signals are identifiable.

Results consistently below agreed targets with no credible plan to address them.

Reporting that has become opaque or that changes metrics when results decline.

Loss of the senior people who were involved when the relationship began.

Recommendations that never change, which suggests the account is being maintained rather than managed.

An absence of curiosity about your business, since an agency that does not understand what you sell cannot market it well.

Conversely, the relationship is working when results are measured honestly, problems are raised by the agency before you notice them, and the recommendations change as evidence accumulates. That is what accountability looks like in practice, and it is worth more than any particular campaign result.