Two numbers, both from credible research, pointing in opposite directions.
91% of marketers expect budget increases in 2027 — Forrester, from a survey of more than 2,600 business and technology decision-makers. And marketing budgets have been on a plateau since 2022, currently averaging around 7.8% of company revenue, roughly 18% below where the allocation sat four years earlier — Gartner's CMO Spend research.
Both are true. Expectation surveys measure intent; spend surveys measure outcome. The gap between them has been persistent for several years, and understanding why is more useful than either number alone.
Why optimism keeps not arriving
The reconciliation isn't economic. It's evidentiary.
Gartner's research describes a cycle it calls a brand "doom loop": underfunded measurement produces unclear evidence of impact, which produces scepticism from the wider business, which produces tighter budgets, which further underfunds measurement. Research covering senior marketing leaders found 84% of companies caught in it — and those companies were half as likely to exceed growth targets as ones able to evaluate brand value properly.
What this means for your plan The constraint on your 2027 budget is not the market. It's whether you can show what the 2026 one did — which means the case was largely built or lost months before the planning meeting.
The accompanying prediction is blunt: Gartner expects that by 2027, more than 40% of CMOs pushing for larger brand budgets will lose influence with the C-suite because they cannot demonstrate sufficient returns. Not lose the budget — lose the standing to ask.
Which reframes the planning task. The question isn't how to argue for more. It's whether the evidence exists to make the argument land, and that's a capability question with a lead time.
Where the money is actually moving
Within a flat total, the mix has shifted meaningfully.
| Line | Direction | Why |
|---|---|---|
| Paid media | Up, to ~31% of budgets | Absorbing freed-up spend; largest single line |
| Agencies | Down | 39% of CMOs planned cuts; rosters streamlined |
| Labour | Down as a share | Productivity gains absorbed rather than reinvested |
| Martech | Down as a share | Consolidation after years of accumulation |
| First-party data | Emerging as its own line | Signal loss making owned data structural |
Two details worth pausing on. 22% of CMOs report that generative AI has reduced their reliance on external agencies for creativity and strategy — which is the client-side view of the pressure described in what agencies are seeing in Q4 budgets.
And paid media rising is not the same as paid media working. Media price inflation means each pound buys less than it did — the dynamic in rising CPMs. A line item growing while its unit cost grows faster is a defensive move, not an investment.
The AI line item is the wrong argument
Where a lot of 2027 planning is about to go wrong, and the research is unusually direct about it.
Forrester's stated position: the organisations that outperform in 2027 won't be those that spend the most on AI, but those that invest in the foundations that make AI effective — trusted data, strong governance, organisational readiness, and the ability to keep adapting as technology and customer behaviour change.
That's a genuinely awkward thing to put in a budget, because "trusted data" and "governance" don't demo well and don't have a vendor pitching them. A tooling line does.
But the logic holds under inspection: automation applied to unreliable inputs produces faster output that's wrong in the same direction, which is a more expensive version of the existing problem rather than a solution to it. The verification habit of checking what a model actually did is the small-scale version of the same point, and the governance side is covered in AI governance for marketing teams — which has now moved from good practice to regulatory requirement in several markets.
The practical version for a 2027 plan: budget the data work and the governance work as named lines with named owners, rather than assuming they'll happen alongside the tooling. They won't.
Systeme.io
Affordable all-in-one marketing platform with funnels email and course builder included
Best for: All-in-One Platform
The planning assumption that broke
Worth stating because it explains why more budget may not produce more result.
For a decade, B2B planning rested on assumptions that were reasonable: buyers were increasingly visible, engagement signals indicated interest, channels were manageable, and journeys could be observed and influenced.
Each of those has weakened. Buyers are harder to observe, discovery and evaluation are being reshaped by AI, traditional measurement signals are degrading, and buying groups keep expanding. Forrester has described this convergence as a go-to-market singularity, which is a dramatic phrase for a mundane and serious problem: the model most marketing organisations are built around describes a buying environment that no longer exists.
The consequence for planning is specific. More budget flowing into a model built for the previous environment produces more activity rather than more impact. Which means the useful 2027 question isn't allocation — it's whether the organisation can adapt faster than the market moves.
Concretely, three of those assumptions have already been documented breaking: discovery has moved partly into surfaces you can't see, as in where AI search traffic actually lands and measuring AI referral traffic; attribution has degraded as covered in why attribution keeps getting harder; and buying groups have widened, as in why committees trust peers over ads.
What to do in the six weeks before your plan is signed
The order matters. Evidence first, capability second, contingency third. A team that can't evidence last year gains nothing from a better contingency plan.
What's actually worth funding
Five things that come out of the research consistently, ordered by how badly they're usually neglected.
Measurement that survives scepticism. Holdout tests, incrementality, marketing mix modelling — the methods in why cross-channel measurement is a must. This is the doom-loop exit and almost nobody funds it deliberately.
First-party data infrastructure. Now a line in its own right rather than an IT concern, because the signals it replaces are the ones being lost.
Unit economics by segment. Knowing what you can afford to pay per customer per channel, as in calculating lifetime value properly, converts a budget request into a bounded commercial argument.
Proprietary evidence. Your own data published, for the reasons in why brands publish their own data — one of the few assets that keeps working while the surrounding content market compresses. That's a content investment behaving like infrastructure.
Governance and disclosure capability, which has moved from optional to obligatory across several markets during 2026 and is now genuine scope rather than an admin overhead.
What the optimism is probably right about
Worth balancing, because the picture above is one-sided and the Forrester finding shouldn't be dismissed.
The 91% figure follows a year of caution, and it sits alongside broad increases expected across technology, personnel and programmes rather than in one line. More than half of customer experience leaders anticipate growth of 5% or more, up sharply from the year before. That's a real change in posture, not noise.
The reasonable reading is that the appetite to invest has returned while the standard of proof has risen. Money is available to teams that can show what it does — and unavailable, or available at a lower level, to teams that can't. That's a more demanding environment than either a boom or a squeeze, because the differentiator is internal rather than external.
It also means the teams that spent 2026 building measurement capability rather than defending activity are about to have a materially easier year than the ones that didn't.
If you're a smaller team reading this
Most of the research above surveys enterprise CMOs, and the numbers don't transfer directly. Two things do.
The doom loop applies at any size, and it bites harder when the person controlling the budget is a founder sitting across the table rather than a committee. The dynamic is identical — unevidenced spend attracts scepticism, scepticism produces cuts — but the feedback arrives faster and less formally, which is an advantage if you're prepared for it.
And the measurement fix is cheaper for you than for them. A holdout test in one region, a self-reported attribution question at checkout, or a simple channel-level payback calculation are all achievable without a platform purchase. Smaller businesses can often produce better evidence than large ones precisely because the data is less fragmented and there are fewer systems disagreeing with each other.
The one thing worth resisting is the assumption that flat budgets and evidence pressure are a large-company problem. They are the same problem at a different scale, arriving through a shorter conversation.
Six questions to answer before you submit
- What did 2026 spend produce, in the vocabulary your CFO uses?
- Which claims can you evidence, and which are assertions?
- What would you cut at 80%, decided now?
- Which single measurement capability, if funded, would most change next year's conversation?
- Where is your model still assuming a buying environment that has changed?
- Is your AI line buying tools, or buying the foundations that make tools work?
The mechanics of assembling all of this sit in building a 2027 marketing plan, and the case-making itself in proving pipeline impact to finance. What this piece adds is the context: you're presenting into an environment where appetite is up and tolerance for unevidenced claims is down, simultaneously.
And the format matters more than usual — a standing monthly note showing what changed and why, of the kind described in a dashboard leadership will actually read, does more for next November than any deck produced this one.
The short version
91% of marketers expect budget increases in 2027, and budgets have been flat since 2022 while falling around 18% as a share of revenue. Both are true, and the gap is evidentiary rather than economic: 84% of companies are in a cycle where underfunded measurement produces unclear impact, scepticism and tighter budgets — with Gartner predicting that over 40% of CMOs pushing for larger brand budgets will lose influence by 2027 because they can't show returns. Inside a flat total, money has moved toward paid media and away from agencies, labour and martech, with 39% of CMOs planning agency cuts and 22% citing AI as reducing their reliance on external creative and strategy. Don't build the plan around an AI tooling line — the research points to trusted data, governance and readiness as what makes tools effective. And recognise that the assumptions the last decade of planning rested on have weakened, so more budget in an unchanged model buys activity rather than impact.
The constraint on next year's budget isn't the market. It's whether you can show what this year's did — and that case is built during the year, not in the meeting.
Planning 2027 without evidence for 2026?
We rebuild the measurement so next year's number is a commercial argument rather than a request.
Build my case →Frequently asked questions
Are marketing budgets actually increasing in 2027?
Expectations are strongly positive while the recent record is not. Forrester's survey of more than 2,600 decision-makers found 91% of marketers expecting budget increases in 2027. Gartner's CMO Spend data shows budgets on a plateau since 2022, averaging around 7.8% of company revenue — roughly 18% below the allocation of four years earlier. Both can be true, because expectation surveys measure intent while spend surveys measure outcome, and intent has not translated for several years running.
What is the brand doom loop?
A self-reinforcing cycle Gartner identified in which underfunded measurement produces unclear evidence of impact, which produces scepticism from the wider business, which produces tighter budgets, which further underfunds measurement. Research covering senior marketing leaders found 84% of companies caught in it, and companies in that position were half as likely to exceed organisational growth targets as those able to evaluate brand value properly.
Where is marketing budget actually moving?
Toward paid media and away from agencies, labour and martech as shares of the total. Paid media has risen to roughly 31% of marketing budgets, the largest single line. Around 39% of CMOs have planned agency cuts through eliminating relationships, streamlining rosters and renegotiating scopes, and 22% report that generative AI has reduced their reliance on external agencies for creative and strategic work. First-party data has emerged as a meaningful budget line in its own right.
Should 2027 budgets include a large AI line item?
The available guidance points the other way. Forrester's position is that the organisations outperforming in 2027 will not be those spending most on AI, but those investing in the foundations that make AI effective — trusted data, governance, organisational readiness and the ability to keep adapting. A large tooling line without those foundations produces faster output on unreliable inputs, which is a more expensive version of the same problem.
What is the biggest risk to a 2027 marketing budget?
Inability to demonstrate what the previous budget produced. Gartner has predicted that by 2027 more than 40% of CMOs pushing for larger brand budgets will lose influence with the C-suite because they cannot show sufficient returns. That makes measurement capability the constraint rather than market conditions — and it means the case for next year's budget is largely built or lost during this year rather than in the planning meeting.
