Cut vs. Create: How Marketing should define ROAI in 2027

September is upon us, bringing the all-too-familiar juggling act of Golden Quarter execution and budget planning for 2027. As budgets take shape, the conversations we’re having suggest that ROAI, what marketers get back from their AI spend, is fast becoming central to board budget submissions.

It’s a fair question. For some time, the technology has promised all manner of transformation. As knowledge and capability mature, its success is increasingly being measured by the roles, hours and fees it has removed. The problem is that most answers start in the same place: faster tasks, cheaper content, reduced agency costs and fewer people.

Two FT stories on consecutive days caught my eye this week, coinciding neatly with 2027 budget planning season and illustrating why the “cut” narrative has taken hold in boardrooms. One described companies using AI to reduce their reliance on large consulting teams and bring work in-house. The other showed law firms being pressed to lower fees because research, document review and contract assessment can now be completed in fewer hours. Both industries have long relied on large teams billing large numbers of hours. When AI removes those hours, it attacks the unit being sold. Naturally, the return is framed as lower cost.

Marketing does a different job doesn’t it?

I’ve spent more than 20 years watching marketing and agency teams being asked to do more with less. Sometimes it creates focus. More often lately, it strips out the resource and thinking that made the work good. Marketing is not a list of tasks. It helps businesses grow, find new customers, create more reasons to buy, building mental availability and investing marketing money better. A marketing function that becomes 20% cheaper while the brand becomes less distinctive has not become 20% better.

The results we’re seeing with our clients at TAU reflect wider industry evidence of a broader definition of return. In work with leading CMOs, BCG has seen 20–30% cost improvements alongside a threefold increase in marketing ROI and a tenfold improvement in campaign cycle times. PwC found its most AI-ready companies generated AI-driven revenues and efficiencies 7.2 times higher than the rest. The message here is clear: the strongest ROAI comes from cutting waste and creating growth.

This is why I think marketing needs to look at ROAI through two simple lenses: Cut and Create.

Cut is obvious. Automate repetitive, labour-intensive data processing. Produce a first draft in minutes. Stop moving numbers between tech platforms or rebuilding reports in spreadsheets that were out of date before the meeting.

Create is bigger. It uses AI to tackle marketing problems that have become too slow, complex or fragmented to optimise properly. It redeploys the time saved to explore fresh perspectives, ideas and opportunities.

Take media planning. Cut produces a media plan faster. Create frees planners to explore far more routes to growth, bringing together market, audience, brand, sales and performance signals that normally live in different systems, with different specialists and different versions of the truth. It can test more audiences, investment levels and channel combinations. It can model scenarios in real time, like what happens when brand ATL increases, or when the plan chases Q4 conversions by increasing performance activity.

Create means finding growth that current planning cycles miss. Joining CRM with market intelligence, customer behaviour, search patterns, audience segmentations and business outcomes can unearth overlooked buyers, emerging needs, new purchase occasions or places where the brand has permission to play. AI can spot likely lapsers, what might bring them back and where loyalty is being lost. It can connect those signals to media and messaging, avoiding another blanket discount for people who would have bought anyway.

Used properly, AI can help make better decisions across the whole paid media plan. It helps planners see trade-offs earlier, challenge the easy answer and find the gold in the insight. Create is not about fewer media planners. It is about planners doing more valuable work.

The same applies to creative. Building a hundred average assets more cheaply is a Cut. Higher value outcomes come from exploring more cultural signals, strategic territories and creative routes, then learning what builds attention, memory and action, all in real time. That’s Create.

In measurement, a tech platform can use AI to make its own ROAS look better, but that does not create incremental growth. Joining planning, activation and independent measurement shows a true picture of what changed behaviour, feeds learning into the next decision and improves the plan with every cycle. That’s Create.

Cut matters. Bank the savings and remove the waste. But use some of what comes back to fund Create. If AI frees the capacity of four FTEs and the response is simply to remove four FTEs, the return stops at the cost of these. Reinvest that capacity in understanding customers, exploring possibilities, building ideas and improving decisions, and the business creates growth that can compound.

This is where our work at TAU starts: not simply making existing marketing processes faster, but helping teams find new audiences, optimise media across brand and performance, and make every planning cycle smarter than the last.

The 2027 marketing ROAI conversation should not only ask, “How much can AI Cut?” It should also ask, “What can AI help Marketing to Create and grow that it could not before?”