Hosted by David Blyth of eatbigfish Africa (represented by Delta Victor Bravo), in partnership with the Association for Communication & Advertising (ACA) and the Marketing Association of South Africa (MASA), the session brought together two people who have spent their careers on different sides of the brief: Stuart Dunk, founder of Don’t Forget to Look, who crossed from agency life into marketing procurement at brands like Betfair, Reckitt, Nike and Danone; and Johan van der Westhuizen, founder and managing director of Yardstick, who has been a consulting bridge between finance and marketing since 1998.
Good procurement, bad procurement
Dunk started with the version many of us have met. Bad marketing procurement measures and negotiates on time and rate cards, and treats every supplier in a way that becomes a race to the bottom. It usually comes from people who are junior, rushed, or generalists parachuted in from another category of buying, now running a pitch that spans strategy, creative and production. Good procurement does the opposite: it aligns RFP or pitch requirements to the marketing stakeholders it serves, links into the KPIs of marketing, and leans towards value-based outcomes and performance-based fees rather than hours.
Van der Westhuizen built on the point from the client side. Good procurement understands that marketing creates business value, stays close enough to know which questions to ask, and still holds the line on governance; bad procurement applies a generic, time-and-rates process to everything, and the result is horse-trading. Both kept returning to one flaw: incentives. Procurement is usually incentivised with a bonus on savings, which pulls decision-making toward the cheapest answer; give it the same targets as the marketers it serves, Dunk argued, and the tension disappears. Good procurement also front-ends risk before it gets to legal, as evidenced by data, IP and the AI clauses now appearing in master services agreements.
It starts with the brief
If procurement is going to buy outcomes, someone has to be able to say which outcome they are buying, and that is where it often falls apart. Chasing outcome-based pricing off the back of a vague brief, as Blyth put it, is like playing darts in the dark. A brief that can be answered five ways should never leave the building.
Dunk was blunt about what a usable brief needs: what you achieved last year, what you want this year, and a realistic budget range. Withhold those and you waste everyone’s time. He offered a statistic that landed: “80% of brands think they’re good at writing briefs, and only 10% of agencies believe that to be true.” Van der Westhuizen pointed out that the gap often sits with marketing, not procurement: when marketers cannot define what they want or what separates one supplier from the next, weak guidance produces a generic process.
Stop leading with price
The clearest theme was sequence. If the first conversation is about cost, you are already in the wrong place. Have the effectiveness conversation first; the commercials come second. Both panellists see the weighting slowly shifting in that direction.
Dunk has seen price account for 40 to 50% of a pitch scorecard; the better processes he now runs put the commercials at around 30%, with no points for being cheapest. Van der Westhuizen put the private-sector norm at 20 to 30%, noting that public entities governed by the PFMA clear a technical threshold before price enters at all. The direction is the same: away from buying hours, towards buying value.
Out of the sea of sameness
There is an irony inside the paradox: agencies whose entire reason for existing is differentiation routinely show up looking interchangeable. Dunk spends his days on agency websites, and they blur together with claims they think set them apart but are simply the cost of entry. “The homepage says we’re creative-led, we’re partnership-focused, we use data,” he said. “That’s par for the course. It doesn’t differentiate you and it doesn’t help you win – instead it keeps you interchangeable.”
His fix is what he calls radical accountability: codify your services in your own way, lead with case studies that headline results rather than activity, and define tightly who you are for. Van der Westhuizen sees the same failures from the buyer’s seat: one pitch deck recycled for every client, case studies with no relevance to the brand in the room, and a strong value story the agency then fails to carry into the negotiation, where it becomes the first thing procurement trades away. Relative differentiation has to survive all the way to the contract.
AI is collapsing the existing economic model
AI ran underneath much of the conversation, and its first casualty is the billable hour. “Work that used to take days now takes minutes,” Dunk said, and that collapse is accelerating the end of time-based pricing. The danger is inexperienced procurement assuming everything must now be 30 or 40% cheaper, without working out what AI changes or how to measure it. Better units are emerging, like cost per output or deliverables per unit of time.
Van der Westhuizen added a warning for agencies banking on AI margins: the belief that it unlocks easy super-profit ignores basic economics, because where there is super-profit to be made, new operators pile in and compete it away. The honest move, both agreed, is for agencies to bring the AI conversation to the table themselves and explain: here is what this used to take, here is what it takes now.
From outputs to outcomes
Asked how to move to outcome-based pricing in practice, both resisted the silver bullet. Van der Westhuizen described output-based remuneration as the realistic first step: define the deliverables, agree a cost per output, then layer multipliers and performance on top. He noted how few people on either side know what things cost; ask people for a ballpark on a typical set of outputs and many draw a blank. Get that right and it informs budgeting, planning and negotiation at once.
Dunk’s answer highlights a useful and defining habit: “Ask the client what they’re trying to achieve, and what achieving it is worth to them.” A good digital agency, he said, does not price a website on rounds of design; it asks what the site is for, floats the KPIs that matter, for example, load time, drop-offs, cart abandonment, or revenue – and agrees the from and to move with the client. Where he started on the client side, all revenue came through a website, so agencies were paid against sign-ups and active users. The limit flagged here, however, is that this is far easier in digital than in traditional marketing. Where a brand is running six or seven campaigns a year, it’s better to agree the overall strategic direction first, and then add a performance structure to each brief.
The fight worth having
By the end, the gatekeeper-versus-growth-blocker framing looked like a false choice. Procurement blocks growth when it is distant, junior and fixated on price; it drives growth when it is senior, close to marketing, aligned on the same targets, and focused on value creation and risk mitigation. The closing advice pointed at the same fix from each side: Dunk urged agencies towards radical accountability: to own commerciality, positioning and results, and so negotiate from strength. Van der Westhuizen, on the other hand, asked simply for marketing and procurement to collaborate. The tension that gives marketing procurement its bad name is not structural – it’s an alignment problem, and alignment is a choice. DM
About the series
This was the second session of “We’re Picking Fights”, Season 3 of the Daily Maverick Marketing Masterclass, hosted by eatbigfish Africa (represented by Delta Victor Bravo) in partnership with the Association for Communication & Advertising (ACA) and the Marketing Association of South Africa (MASA). Six fault lines, one platform for tough debate. The series runs as a one-hour webinar every second Thursday from 11 June to 20 August 2026. The next episode asks whether customers really care about your brand, and takes place on 9 July. To view the series, visit Daily Maverick Events.