When the Brief Is the Problem

There is a persistent weakness in the relationship between advertisers and agencies that rarely receives the attention it deserves. Much of the industry’s focus is placed on the quality of the creative work, the strength of the strategy, the effectiveness of the media and, increasingly, the technology used to produce and distribute campaigns. Yet before any of these things can have an impact, there is a more fundamental question to address: has the agency been given the right problem to solve?

The briefing process is often treated as an administrative step between client and agency. The client has an objective, the agency receives a brief and the creative process begins. In reality, the brief is one of the most consequential parts of the entire relationship. It defines the boundaries within which the agency is expected to think and establishes the problem against which the eventual work will be judged.

If that problem is poorly defined, misunderstood or simply the wrong problem, even an excellent agency can produce work that fails commercially.

There is now another problem emerging, and it is one that the industry should be particularly wary of. Briefs are increasingly being "phoned in" by AI.

It is not difficult to imagine how this happens. A CMO, marketing manager or brand team needs a brief. They provide an AI system with the brand name, a few objectives, some existing campaign material and perhaps a brand strategy document. Within seconds, an apparently comprehensive brief appears. It has the correct terminology. It understands the category. It can summarise the brand's history, identify previous campaigns, describe the target audience and produce a respectable collection of strategic objectives.

It looks like a brief.

It may even be a very good summary of everything that can be found about the brand online.

But that is not the same thing as understanding the brand, its lived experience, the human and cultural reasons for decisions being made that never made it to a floppy disc or intranet.

The internet is not institutional memory

A brand's history is not the same as its digital record.

The internet can tell an AI what campaigns a brand has run. It can tell it what the company says about itself, what journalists have written, what advertising awards have been won and what products have been launched. It can reconstruct an impressive chronology of public events.

What it cannot reliably understand is the human experience behind them.

It doesn't know what it was like to sit in the room when the CEO killed the campaign everyone thought was brilliant. It doesn't know why the marketing director quietly abandoned an idea that looked perfect on paper. It doesn't know what the sales team actually thought about the new proposition. It doesn't know the conversation between the client and agency after a campaign spectacularly failed. It doesn't know which customer complaint changed the way a product was developed. It doesn't know which internal disagreement shaped a decision that, years later, looks entirely rational in the official history.

Those things are rarely online.

They exist in people.

They exist in memory, relationships, conversations, experience and institutional knowledge.

That distinction becomes critically important when an agency is being briefed.

The history of a brand is more than its style guide and campaigns

One of the most valuable assets in a long-term agency relationship is institutional knowledge. An agency that has worked with a business for years can accumulate a remarkable understanding of how that business actually operates. It knows the history behind decisions rather than simply the decisions themselves.

It understands why previous campaigns succeeded or failed. It knows which propositions have already been tested. It knows which audiences have proved difficult. It knows which internal stakeholders matter and where the political sensitivities sit. It knows which apparently sensible ideas have been tried before and which problems have been quietly circulating inside the organisation for years.

That knowledge is difficult to document because much of it is experiential.

It is the accumulated intelligence of being there.

Yet that is precisely the knowledge that is most vulnerable when agency relationships change or when senior marketing leadership changes.

A new CMO arrives. A new agency is appointed. The old relationship is discarded. The new agency receives a series of documents and perhaps an AI-generated synthesis of the brand's history. The organisation believes it has transferred the knowledge.

It hasn't.

It has transferred information.

Information is not knowledge, and knowledge is not understanding.

The danger of synthetic familiarity

AI makes this distinction more important, not less.

The danger is not that AI will write bad briefs. The danger is that it can write incredibly convincing ones.

A poor brief is easy to challenge. A polished, apparently intelligent brief creates the impression that the hard thinking has already been done.

An AI system can absorb a brand's website, campaign history, published strategy, social activity, press coverage and publicly available customer information and produce a coherent account of what the brand is, where it has been and what it appears to want to achieve.

But it is constructing a picture from available evidence.

It does not possess the lived experience of the organisation.

It cannot know what isn't documented.

And perhaps most importantly, it cannot know what the organisation itself has forgotten.

This creates a new form of briefing risk: synthetic familiarity. The brief feels informed because it contains a lot of information, but the information may only describe the visible surface of the brand.

The agency is then asked to solve a problem using a history that has been reconstructed from the internet rather than understood through the people who lived it.

That is a very different thing.

The business problem gets lost on the way to the agency

This matters because business problems are already prone to being diluted as they move through an organisation.

A business might be struggling with declining sales, a poorly differentiated product, weak distribution, changing customer behaviour or increasing competitive pressure. These are commercial problems. They may require marketing, but they are not necessarily marketing problems.

By the time they reach the marketing department, they may have become objectives such as increasing awareness, improving consideration, strengthening engagement or refreshing the brand. By the time the agency receives the brief, the problem can have become a request for a campaign.

Now add AI into the process and there is a risk that the sanitisation becomes even more efficient.

The messy human context disappears.

The organisational uncertainty disappears.

The uncomfortable conversations disappear.

The things nobody has written down disappear.

What remains is a beautifully structured representation of what the business believes its problem to be.

That may be useful.

It may also be completely wrong.

What is the actual problem our creative thinking is required to solve?

This is the question I would like to see at the heart of every agency brief.

What is the actual problem our creative thinking is required to solve?

Not simply what campaign is required. Not what content needs to be produced. Not what the brand wants people to think or feel.

What is the actual problem?

Because creative thinking is capable of being applied far beyond advertising execution. It can be applied to products, propositions, customer experiences, technology, distribution and business models. An agency's value can extend well beyond the production of communications if it is given access to the real problem early enough.

A product may be difficult to sell because its proposition is weak rather than because its advertising is ineffective. A service may be losing customers because the experience is poor rather than because awareness is insufficient. A brand may be struggling because its positioning no longer reflects the market rather than because it needs another campaign.

In each case, better advertising might improve the expression of the problem without addressing its cause.

The agency needs enough context to challenge that possibility.

The expertly executed wrong answer

Perhaps the most damaging consequence of poor briefing is that it can remain invisible.

A bad brief does not necessarily produce bad creative work. It can produce excellent work. It can be beautifully strategic, creatively outstanding, perfectly on brand and delivered precisely on time and on budget.

The campaign can even perform well against the objectives established in the brief.

And the business can still fail to move forward.

This is one of the most frustrating situations in the agency-client relationship. The client concludes that the agency isn't delivering. The agency points to the agreed brief and demonstrates that it has delivered exactly what was commissioned. Both sides can be correct.

The failure happened before the work began.

A brief may have asked an agency to increase awareness when awareness was not the limiting factor. It may have asked for a brand campaign when the proposition itself was weak. It may have asked for greater engagement when the real problem was conversion. The agency can execute each of these tasks exceptionally well and still fail to produce the commercial outcome the organisation needed.

Eventually the relationship becomes characterised as unsuccessful.

The agency is replaced.

Another agency arrives.

Another brief is written.

And the cycle begins again.

How many relationships have broken down because the brief was wrong?

This is the question the industry rarely asks.

How many agency relationships have broken down because a half-arsed brief was answered expertly, on brand and on budget, but never managed to move the needle because it was addressing the wrong problem?

How many agencies have been fired for failing to solve problems they were never actually asked to solve?

How many CMOs have changed agencies when what really needed changing was the question?

And now, perhaps, how many briefs are being generated by AI that look impressively informed while lacking the human experience required to understand the organisation behind the brand?

There is a particular irony here.

We are entering an era in which AI can make the production of marketing assets dramatically faster, while potentially making the thinking that precedes those assets more superficial.

That would be a terrible trade.

If AI is used to help interrogate a problem, uncover patterns, research markets and challenge assumptions, it can make the briefing process considerably better.

If it is used to avoid the uncomfortable work of talking to people, understanding history and confronting the real business problem, it may simply allow us to produce bad briefs faster.

A better brief begins with people

The answer isn't to reject AI. The answer is to understand what AI cannot replace.

A genuinely useful brief needs the history of the brand, but not simply the history that can be found online. It needs the history held by the people who experienced it.

It needs the agency to understand what has actually happened, not merely what was published about what happened.

That means conversations. It means talking to the people who have been around long enough to remember. It means listening to sales, product, customer service and leadership, not simply marketing. It means asking what has been tried, what failed and why. It means admitting what the organisation doesn't understand.

It also means allowing the agency to challenge the premise before it is asked to execute the solution.

The best briefing process may therefore be less about writing a better document and more about creating a better conversation.

The document can follow.

The future of briefing shouldn't be automated

There is an understandable temptation to automate anything that looks like administrative work. Briefing can certainly benefit from technology. AI can help organise information, identify gaps, research markets and interrogate assumptions.

But briefing is not fundamentally an information-management exercise.

It is an act of understanding.

The most valuable information about a brand is not necessarily the information that has been published. It is often the information that lives inside the people who have spent years working on it.

That institutional knowledge is one of the great assets of an agency relationship. It is also one of the easiest things for a business to lose.

Perhaps the future agency brief should therefore contain less synthetic certainty and more genuine curiosity. Less recitation of what the brand has done and more understanding of why it did it. Less confidence that the problem has already been defined and more willingness to ask whether it has.

The fundamental question remains remarkably simple.

What is the actual problem our creative thinking is required to solve?

If we can answer that honestly, with the benefit of human experience and whatever intelligence AI can usefully add, then we might finally give agencies the opportunity to do what they are supposed to do.

Not simply make the thing.

Not simply make the thing faster.

But help solve the thing that actually matters.

Because the most expensive brief is not the one that costs the most to produce.

It is the one that gets answered perfectly and solves nothing.