To diagnose a marketing problem, work backwards from the commercial outcome to the constraint causing it, before choosing any intervention. Most teams do the opposite: they inherit last year's activity, then optimise it harder when growth stalls.

That confusion, between optimisation and diagnosis, is the most expensive one in modern marketing, and it is one of the five substitutions our essay How Digital Marketing Made Us Forget Marketing argues quietly narrowed the discipline through the digital era. Optimisation improves the system you already chose, while diagnosis asks whether you chose the right system, and you can optimise a landing page forever without ever asking whether the proposition is any good while the dashboard cheers you on the whole way.

Here is a practical method for diagnosing before you optimise, and the evidence for why the sequence matters.

Why Optimisation Cannot Find Your Real Problem

Optimisation starts after the intervention has been chosen. It improves a chosen activity against a chosen objective, and it is genuinely valuable within that frame: refining audiences, testing creative, removing friction, shifting budget to what performs.

What it cannot do is tell you whether the frame is right. A team can improve conversion rates while the proposition stays weak, cut cost per lead in the wrong market, and lift engagement on emails nobody values.

Dashboards made optimisation continuous and absorbing. Something always moved since yesterday, the platform always suggests a next test, and improvement is always observable. Diagnosis moves slower, draws on evidence from several sources, and sometimes returns unwelcome answers that cross departmental lines: the offer is undifferentiated, the price is misaligned with perceived value, the product misses the segment, distribution makes you hard to buy from.

It is always easier to adjust the campaign, and it is rarely the campaign that is actually broken.

Step 1: Start From the Commercial Outcome, Not the Channel

Name the business problem in commercial terms first. "Revenue growth has slowed in the mid-market segment" is a diagnosable problem. "Our CPL is up" is a channel symptom.

Then resist the reflex to translate the problem straight into channel action. When marketing is organised around channels, the conversation starts with what to do on search, social or email rather than with the customer and the market. The capability starts determining the strategy, which is precisely backwards.

Step 2: Locate the Constraint With the Growth Constraint Checklist

Growth is usually limited by one or two constraints at a time. Test each candidate against evidence, not opinion:

1. Awareness: do enough of the right buyers know you exist? (Brand tracking, aided and unaided recall, share of search.)
2. Relevance: do they believe you solve their problem? (Win/loss interviews, sales call themes, website search terms.)
3. Distinctiveness: can they tell you apart and remember you? (Brand asset testing, category review.)
4. Availability: are you easy to find and easy to buy from? (Journey audit, distribution and channel coverage.)
5. Price and value: is the price aligned with perceived value? (Win/loss data, discount frequency, competitor positioning.)
6. Product fit: does the offer actually serve the target segment? (Churn reasons, feature usage, NPS verbatims.)
7. Conversion: is demand arriving but leaking? (Funnel analysis, sales conversion by stage.)

The discipline is writing down which constraint you believe is binding, and what evidence supports that belief. If the diagnosis is missing, the campaign is a guess.

Step 3: Check Whether You Are Capturing or Creating Demand

One diagnosis question deserves its own step because the measurement system hides it.

Professor John Dawes of the Ehrenberg-Bass Institute uses the 95:5 heuristic: in many B2B categories, roughly 95 per cent of potential buyers are out of market at any moment. They cannot be converted today, only remembered. Les Binet and Peter Field's IPA analysis shows what happens when budgets ignore this: over-weighting short-term activation steadily erodes the brand building that creates future demand.

So ask: is the real problem that we are failing to convert today's demand, or that we stopped creating tomorrow's? The second problem never shows up in a conversion dashboard. It shows up two years later as a pipeline that mysteriously thinned.

Step 4: Match the Intervention to the Diagnosis

Only now choose the work. If awareness is binding, no landing page test fixes it. If price is binding, no retargeting budget fixes it. If conversion is genuinely the constraint, optimisation is exactly the right tool, and you can deploy it with confidence because you have ruled out the alternatives.

Design measurement around the objective at the same time. Decide what business outcome the intervention must change, then choose measures that can detect it. Activity metrics (clicks, cost per lead) manage execution; they are not evidence of incremental growth, and reporting them as if they were is how the next mis-diagnosis starts. For the full evidence on why attribution needs support from other methods, see our guide to marketing attribution's limitations.

Step 5: Re-Diagnose on a Cycle, Not Under Pressure

Constraints move. A business that fixes awareness may surface a conversion problem underneath it. Run the diagnosis annually, or whenever the growth trend breaks, rather than mid-crisis when the pressure to "do something visible" is strongest.

Diagnosis under pressure defaults to whatever produces the fastest signal, which is how organisations end up optimising their way to a very efficient wrong answer.

The one-line takeaway: name the constraint before you choose the intervention, because optimising the wrong system just gets you to the wrong place faster.

Sources

• Les Binet and Peter Field, The Long and the Short of It (IPA), on activation vs brand building.
• John Dawes, Ehrenberg-Bass Institute, Advertising Effectiveness and the 95:5 Rule, on out-of-market buyers.
• Philip Kotler, The Past, Present, and Future of Marketing, on diagnosis and planning sequence (research, segmentation, targeting, positioning before implementation).