Brand vs. Performance Marketing: The 95-5 Rule and 60/40

Why the brand vs. performance marketing question is framed wrongly

Most budget discussions run as an either/or: brand building against leads, reach against conversions, the long game against the quarter. Marketing effectiveness research answers the question differently. Brand building and activation are not competitors for the same budget, they are two mechanisms with different time horizons that need each other. Activation collects demand that already exists. Brand building creates the demand that activation can collect in six, twelve or 24 months.

That sounds like a textbook, but it carries a hard economic consequence: anyone who only activates buys the same small pool of ready to buy customers at a higher price every quarter. Anyone who only builds brand has no mechanism to turn the demand created into revenue. The budget question is therefore not a matter of taste but a question of the buying cycles in your category. Three strands of research provide the basis: Ehrenberg-Bass with the theory of mental availability, the 95-5 rule from John Dawes for B2B, and the budget ratios from Les Binet and Peter Field. On top of that comes the counter-position from Byron Sharp, which you need to know before you write a ratio into a budget sheet.

This article treats the theory as budget logic. Which planning frameworks you use for the operational side is covered in the article on the social media concept with frameworks and a KPI cascade. How you measure the effect is covered by the pillar on social media analytics, KPIs and measurement.

Ehrenberg-Bass: growth comes from mental availability

The Ehrenberg-Bass Institute around Byron Sharp laid the empirical foundation with "How Brands Grow". The core thesis: brands grow through penetration, meaning more buyers, not through more loyal buyers. Loyalty programmes and a heavy user focus achieve little, because buying behaviour in almost every category follows statistical patterns that marketing can barely shift. What can be shifted is the probability that a brand comes to mind at all in a buying situation.

Two terms carry this. Mental availability: the probability that a buyer thinks of your brand in a concrete buying situation. It grows out of memory structures built long before the purchase decision. Physical availability: the brand can be bought where the buyer looks, without friction. In B2B that means findable, comparable, contactable.

You build mental availability with two tools. Distinctive brand assets (colours, logo, sound, faces, tone of voice) make the brand recognisable even when nobody is paying active attention. Category entry points are the situations in which a customer enters the category: "our CRM no longer scales", "the bank has cut our credit line", "we need a new tax adviser by Q3". Content that connects to these entry points gets retrieved when the situation occurs. Content that does not is forgotten.

For social media this has a direct consequence: reach among people who buy nothing today is not wasted budget. It is the mechanism through which brand building works. How heavily platforms now charge for that reach is described in the article on organic reach and pay to play.

The 95-5 rule: who is actually in-market?

The 95-5 rule translates the Ehrenberg-Bass logic into B2B numbers. In 2021 Professor John Dawes examined for the LinkedIn B2B Institute how many companies in a category are ready to buy at a given point in time. The answer, gathered on an international data basis: up to 95% of companies are not in-market at any given time. The derivation is simple. In categories such as banking, legal, software or telecommunications, companies change supplier roughly every five years. That produces around 20% willingness to switch per year and around 5% per quarter.

Dawes himself puts it like this: "If I'm chasing clients in commercial banking then it's useful to realise that in any given year only one in 10 of them will be looking to appoint a new bank or switch their lead bank. In a quarter or a month, it's a tiny proportion."

The rule is often quoted in shortened form. Two clarifications matter. First, the 5% applies per quarter, over a year it is closer to 20%. Second, it is a rule of thumb across categories, not a constant: in consumables with short cycles the in-market share is higher, in infrastructure software with ten year contracts it is lower. You should derive the value for your own category from contract terms and churn data instead of taking 5% as given.

The budget consequence is unambiguous nonetheless. Lead gen campaigns, retargeting and intent based targeting address the 5% exclusively. That pool is small, heavily contested and worked by every competitor at the same time. The 95% will go in-market over the coming quarters, and they will then shortlist the supplier they already know. Anyone optimising purely for performance today leaves those memory structures to the competition.

Binet and Field: the 60/40 rule and its B2B variant

In "The Long and the Short of It" (IPA, 2013) Les Binet and Peter Field analysed the database of the IPA Effectiveness Awards, an international data basis. The basis was 996 campaigns entered between 1980 and 2010, from around 700 brands over more than 30 years. Their analysis separates two types of effect. Activation (rational messages, offers, calls to action) produces short term sales spikes that vanish quickly once the campaign ends. Brand building (emotional, broadly distributed communication) works more slowly but cumulatively and lowers price sensitivity over the long run.

The optimal ratio for sustainable growth in their analysis was 60% brand building to 40% activation; the follow-up study "Effectiveness in Context" from 2018 revised this to 62:38. For B2B the LinkedIn B2B Institute presented its own refinement in 2019: around 46% brand building to 54% activation, because B2B buying processes are argued more rationally and sales cycles run longer.

The column that matters in this table is the data basis. Every ratio comes from campaigns entered for effectiveness awards. That is a pre-selection of successful and well documented cases, not a random sample of the market. This is exactly where the criticism starts.

The counter-position: Byron Sharp on the 60/40 rule

Byron Sharp of the Ehrenberg-Bass Institute attacked the 60/40 rule publicly at the Mi3-LinkedIn B2B Next Summit in 2022. His position: 60:40 is "not a scientific law" and rests on "flawed data" from award entries. In the same talk he warned against the hype around attention metrics: "attention is the new metric" is nonsense, and paying more for more than fleeting attention is a waste of money.

This is a genuine, unresolved contradiction inside effectiveness research, and both sides argue with data. The dispute sorts itself onto three levels.

• Consensus: brands that invest exclusively in activation lose efficiency over the medium term, because the pool of ready to buy customers is small and mental availability does not grow back on its own. Sharp, Dawes, Binet and Field all support that statement.
• Disputed: the exact ratio. Whether 60:40, 62:38 or 46:54 depends on category, buying cycle, market position and data basis. None of these numbers is a law of nature.
• Practical consequence: the ratios work as a diagnostic tool for the question "are we roughly in balance?", not as a target value down to the decimal.

Anyone walking into a budget negotiation with the 60/40 rule should know Sharp's criticism. Otherwise one informed CFO is enough to topple the entire argument. The stronger case runs through the 95-5 logic plus your own buying cycles, because it can be derived from company data instead of somebody else's award database.

The reality in 2024: the market has inverted

What marketing leaders consider right and what they budget are drifting apart. The CMO Survey from Duke Fuqua, Deloitte and the American Marketing Association from autumn 2024 (predominantly a US sample) shows: the CMOs surveyed name 50% brand building to 50% performance as the ideal, while the actual split sits at 31.2% brand to 68.8% performance. According to a WARC analysis of the same data, the brand share still stood at 40.1% in 2023 and the performance share at 59.9%. Within a single year the split has tipped further towards the short term.

The reasons are structural, not individual. Performance spend delivers dashboards with clicks, leads and ROAS per campaign. Brand spend delivers brand lift studies and share of voice curves at best, and those become visible a quarter later. Under budget pressure, whatever can be proven in the monthly report wins. That the performance curve itself suffers from this imbalance only shows up with a delay: through rising CPMs, falling conversion rates in retargeting and a growing dependence on ever more expensive auctions.

There are no comparably hard figures for the DACH region. The mechanics are the same, though, and experience from budget conversations in the B2B mid market points towards an even stronger performance bias, because social media budgets there often had to be justified as lead gen budgets to get approved at all.

What this means for your social media budget

The theory translates into five decisions. None of them needs an exact ratio, all of them need a deliberate allocation.

Determine the buying cycle: derive from contract terms, churn and sales cycle what share of your target group is realistically in-market per quarter. With switching cycles of around five years the benchmark sits at roughly 5% per quarter and around 20% per year, with shorter decision cycles correspondingly higher. That number determines how large the performance pool is in the first place.

Audit the actual ratio: assign every social media expense of the past twelve months to a type of effect. Retargeting, lead forms, conversion campaigns and intent targeting are activation. Reach campaigns, video views in the target group, thought leadership distribution and founder content are brand building. In our experience this audit almost always ends with a clear overweight on activation, often without anyone having decided it that way.

Correction with a safety net: if the performance share sits well above 70%, a correction towards 50:50 makes sense as a benchmark in B2B. Not as a jump, but in steps that you back with incrementality tests. Anyone shifting from 80:20 to 50:50 without running geo lift or holdout groups alongside can neither prove nor defend the effect.

CPM as an early warning signal: if your CPMs rise by more than around 30% year on year while the conversion rate stays stable, you are buying the same small in-market pool at an ever higher price. In our view that is the clearest signal to raise the brand share: activation becomes expensive because the mental availability that would make it cheaper is missing.

Share of voice as a control variable: Binet describes the "physics of growth" through excess share of voice. Brands whose share of category communication exceeds their market share tend to grow. How you operationalise that variable for search and social is described in the article Share of Voice: measuring market share in search.

Brand building on social media in concrete terms

In a social context brand building is often confused with image campaigns. In B2B it looks different. It means being present over months among the 95% who are not buying today, with content that connects to category entry points and makes the brand recognisable. On LinkedIn that typically means personal profiles of founders and specialists who publish consistently on a small number of topics, supported by paid reach in the defined target group. Lead forms and demo CTAs do not belong in this phase.

Activation starts where a signal points to in-market status: website visits on pricing pages, engagement with comparison content, search queries with commercial intent. This is where retargeting, lead gen formats and direct offers belong. The craft lies in the transition: activation budget should collect the demand that brand content has created, not bombard cold contacts with demo offers independently of it.

Assigning spend to paid, owned and earned media helps with the budget structure, because brand building is usually more paid heavy (reach costs money) and activation runs more through owned media (website, newsletter, CRM). The PESO model provides the grid for that.

Common mistakes

Brand building as a cost centre without a goal: anyone releasing brand budget without measuring reach in the target group, brand lift or share of voice has no argument when that budget comes up for discussion in the next savings programme. Brand building needs its own KPIs, just different ones from activation.

The ratio as a target instead of a diagnosis: writing 60:40 into the budget template without checking the buying cycle and market position is as blind as 100% performance. The ratio is a test, not a result.

Judging activation with brand metrics or the other way round: assessing a reach campaign by cost per lead inevitably leads to it being switched off. Assessing a retargeting campaign by video views hides its cost.

Last click as the referee: if the attribution model credits every deal to the last click, activation wins every internal budget round, because brand effects stay structurally invisible. The budget question cannot be answered cleanly without incrementality tests or marketing mix modelling.

Ignoring your own category: the 5% rule comes from categories with five year cycles. Anyone selling consumables or monthly subscriptions has a considerably larger in-market pool and can activate accordingly more.

Conclusion

The evidence is unambiguous on one point and debatable on every other. Unambiguous: a budget that goes almost entirely into activation works only the small share of customers who are ready to buy, and gets more expensive quarter by quarter. Debatable: the exact ratio. You therefore do not answer the brand vs. performance marketing question with a number from somebody else's study, but with the buying cycle of your category, an honest audit of the current split, and tests that can prove a reallocation. The ratios from Binet and Field tell you whether you are roughly off target. What is right has to come from your own data.