CMO playbook & advanced tactics for budget allocation & forecasting

The cut lands as a number, not a conversation. Finance circulates a revised plan with marketing down fifteen percent, the CEO has already given the board the figure, and you have about ten days to say what dies. Everything in this lesson happens after that email. Allocation at CMO level is an argument conducted in front of people who do not accept "brand takes time" as evidence, under contracts you signed eighteen months ago, inside an organisation with strong opinions about which jobs are safe. The method is assumed here: response curves, marginal returns and the zero-based rebuild belong to the frameworks lesson. What follows is what you do with them when other people control the outcome.

What the board is actually buying when it funds brand

Boards rarely doubt that marketing works. They doubt that *this much* marketing works, and they have noticed that nobody in the room can state what the next ten percent down would cost. Performance spend answers that question badly but quickly. Brand spend answers it well and slowly, which is why it loses. The asymmetry is structural, not intellectual.

Two numbers make the defence legible to a CFO. The first is the excess share of voiceshare of voiceYour brand's share of total advertising or conversation volume in your category, measured against competitors over a defined period.View full definition → relationship documented by Binet and Field across the IPA effectiveness databank: roughly ten points of share of voice above share of market associates with about half a point of annual market share growth. It is an average, not a promise, but it converts a media budget into a share forecast the finance team can model. The second is time. Brand effects accumulate over two to three years and decay slowly, which means a six-month read on a brand cut is always flattering. Sales hold. The cut looks free. The bill arrives in base volume and price elasticityprice elasticityHow sensitive demand is to a price change. High elasticity means customers react strongly to price increases.View full definition → somewhere around month eighteen, by which point the person who made the cut has usually been promoted.

The recovery asymmetry is the part most CMOs fail to price. Rebuilding reach you gave up costs more than holding it: competitors bought the audience you vacated, your negotiated rates were reset, and awareness recovers more slowly than it fell. Put a recovery cost line in the paper. A board that has been shown the cost of getting back in will often cut somewhere else.

In-housing: what you are really trading

The in-housing decision is sold as a fee saving and is actually a change in the shape of your cost base. Agency retainers are variable with notice periods. Salaries are fixed until you pay severance, which in most European markets runs to months, not weeks. In-housing converts the part of the budget you can flex into the part you cannot, and it does so in the years when trading is good, so the constraint only becomes visible in the year you need flexibility.

A workable test has two axes: frequency of the work, and whether peak talent matters more than steady talent. High-frequency, data-adjacent, cheap-per-unit work belongs inside: always-on social, programmatic execution, performance creative variants, measurement infrastructure. Work that needs an exceptional idea four times a year, or scale in a negotiation, stays outside. Procter and Gamble's "fixed and flow" model is the honest version of this compromise: a smaller retained core, with project work flowing to whoever is right for it.

The failure mode is subtler than cost. An in-house team of forty becomes a constituency in the next allocation debate, and it defends itself by producing volume. Asset counts rise, media weight behind each asset falls. Second failure mode: in-housing the buying and the measurement into the same reporting line. That is marking your own homework, and it is precisely why P&G's demands on third-party verification and accredited viewability measurement mattered more than the fee savings alongside them.

The cuttability trap

Rank your budget by days-to-exit and cost-to-exit, not only by return. Biddable digital stops inside twenty-four hours. Broadcast upfront commitments can be cancelled only in defined windows and usually only in part. Multi-year sponsorships and rights deals cannot be cut at all. Agency retainers carry ninety to one hundred and eighty days of notice.

So when the fifteen percent arrives with a quarter to deliver it, the reachable cut and the least productive cut are different sets of money, and the reachable one wins. That is how a company ends up defunding the channel with the steepest response curve because it was the only one with an off switch. Two responses: keep a liquidity mapmapUsing software to automate repetitive marketing tasks and campaigns, enabling personalisation at scale across channels like email, web, and social.View full definition → next to the returns map so you know in advance what a ten, twenty and thirty percent cut would actually have to hit, and treat cancellation rights as a priced term in every negotiation rather than boilerplate. Paying two percent more for a clause that lets you release a third of a commitment mid-year is cheap insurance you will use once in five years.

Governance that ends the annual fight

Reallocation becomes political when the rules are unwritten, because then the answer depends on who has the CEO's ear. Write the policy: who can move money, within what band, on what evidence, and at what cadence. Attach pre-agreed triggers so the shift is mechanical rather than negotiated (a defined fall in a leading indicator releases a defined tranche, in either direction).

The recurring arbitration is centre versus market. Local general managers want price promotion because it moves this quarter's volume and shows up in their bonus. The centre owns brand equitybrand equityThe commercial value your brand adds beyond functional product attributes: the price premium, preference and loyalty it generates.View full definition →, which no single market is compensated for protecting. Do not resolve this brand by brand each year. Ring-fence the equity-building portion at group level, give markets full authority below that line, and route disputes to a standing committee co-chaired by finance. Disclosure helps: when a company reports marketing investment as a visible line to investors, as Diageo does with a figure comfortably above two billion pounds a year, cutting it becomes an external decision rather than an internal favour.

Real-world cases

Case one: Procter and Gamble, from 2017. Marc Pritchard's January 2017 speech to the IAB set four conditions on the roughly eight billion dollars a year P&G spends on advertising: accredited viewability standards, third-party verification, transparent contracts, and action on fraud. In the quarter that followed, P&G cut more than a hundred million dollars of digital spend and reported no impact on business growth. That sentence will be quoted back at you by your CFO, so know what it means. P&G cut inventory that was unviewable, fraudulent, or hitting the same person too many times, plus over-narrow targeting that bought precision at the expense of reachreachThe number of unique people exposed to your message in a given period. Unlike impressions, reach counts each person once, no matter how often they see it.View full definition →. It did not cut reach. Alongside it came agency and production savings the company put at around three quarters of a billion dollars between 2014 and 2017, with several hundred million more targeted, achieved largely by roughly halving the agency roster. The money moved between the two categories the foundations lesson separates rather than leaving marketing.

Case two: Diageo. A portfolio of two hundred-odd brands across most of the world guarantees that allocation is contested, and Diageo's answer has been to make the comparison method shared rather than local. The company has talked publicly about a proprietary marketing effectiveness capability (Catalyst) used to compare returns across brands and markets, and it has repeatedly earmarked productivity savings for reinvestment in brand building instead of margin. Two consequences worth copying: the annual argument becomes a dispute about model inputs, which is smaller and faster than a dispute about deserving, and long-dated commitments such as Guinness's Six Nations title sponsorship since 2019 are decided as multi-year capital, not as an annual line anyone can raid. The risk is the mirror image: a central model nobody outside head office understands stops being authority and becomes resented. Publish the elasticities. Let markets challenge the inputs, not the arithmetic.

CMO action items

• Build the liquidity map before you need it. Every budget line rated by days-to-exit and cost-to-exit, with a pre-modelled answer to what a ten, twenty and thirty percent in-year cut would actually have to hit. Take it to your CFO before the planning cycle opens, not during it.
• Write the reallocation policy as one page: bands of authority, evidence required, cadence, and the triggers that move money without a meeting. Get it signed by finance. Anything not written down will be re-litigated annually by whoever shouts loudest.
• Price your in-housing decision properly. Fully loaded salaries, tools, recruitment and severance exposure against current fees, then ask separately whether the work is high-frequency or peak-talent. If the honest answer is that you are in-housing to hit a fee target, you are buying a fixed cost to solve a variable-cost problem.
• Prepare the brand defence pack out of cycle, with a share forecast attached to share of voice and an explicit recovery cost for the reach you would give up.

Common mistakes that kill results

Bringing efficiency to a growth conversation. A CMO who arrives with an improved cost per acquisitioncost per acquisitionCost Per Acquisition: the total cost to generate one customer or conversion, computed by dividing total spend by the number of acquisitions.View full definition → while volume declines has answered a question the board did not ask. Cheaper customers and more customers are different achievements, and only one of them shows up in the revenue line.

Assuming a brand cut is reversible on the timeline it was made. It takes a quarter to switch off and years to rebuild, and the rebuild happens against competitors who moved into the space. The decision is closer to closing a factory than to pausing a campaign.

Letting contract structure choose your cuts. If the only money you can reach is the money with the steepest returns, the discipline problem happened at signature, not at the cut.

In-housing without independent measurement. The team that buys the media should not be the only team grading it. P&G's insistence on outside verification was not bureaucracy, it was the recognition that self-reported numbers are worth roughly what they cost to produce.