Marketing budgets go up. The board asks what it bought. If the answer is a few thousand more impressions, you’ve already lost the room. This is about choosing the number that survives that question, and agreeing it before the money goes out the door, not after.
Every marketing director has sat through this meeting. Spend is up, the dashboard is full of green arrows, and somebody on the board asks what it actually got them. Reach, impressions and click-through rate don’t answer that. They describe activity, not outcome.
This isn’t a new problem, but it’s a sharper one now. PwC’s Pulse Survey of CMOs found marketing is often the first budget line scrutinised when spend comes under review, and that proving ROI sits in the top three priorities for a large share of marketing leaders. Separately, the Spring 2025 CMO Survey found 63% of CMOs reported increased pressure from their CFO specifically, up from 52% just eighteen months earlier. The board isn’t getting more patient. It’s getting more specific about what it wants to see.
In our own work with marketing directors across their ideal customer and persona planning, we see where this usually goes wrong. Traffic metrics get the attention because they’re the easiest numbers to pull and the quickest to present. But when a campaign isn’t converting, the real cause is rarely a lack of visitors. It’s usually a conversion problem or a value proposition that isn’t landing with the people actually arriving on the site. Chasing more traffic on top of that just gets you more people not converting, faster.
So we asked other agency leaders, across performance, growth and PR, the same question: what’s the one KPI you trust most when you need to prove marketing ROI to a board? The answers didn’t agree on much. That disagreement is, itself, the point.
Alex Ford, founder of the Manchester data consultancy DataMoon, goes straight for the commercial link. His answer is revenue or sales tied directly to the campaign that produced them, which only works if the CRM or e-commerce backend is actually feeding that link.
That caveat is doing a lot of work. Alex has spent a decade in data roles across the NHS, retail, education and agency sectors, and says the gap he sees most often isn’t a lack of data. It’s data sitting there, uncaptured, because tracking wasn’t set up or linked correctly in the first place. Wasted spend, in his experience, is rarely a strategy failure. It’s a plumbing failure.
He’s also seen the vanity-metric trap from the client side rather than the agency side. Working in retail, he watched a board get drawn into brand analytics (how many people know about us) while the number that mattered (how many of them actually bought something) went unexamined. His view on what a good answer to “what did we get for the spend” sounds like is blunt: ROI tracked directly through the customer journey to a sale or a deal, with attribution doing the joining up.
Guy Levine, founder and CEO of the Manchester growth agency Return, pushes the revenue-first case further and adds a cost. His trusted number is true ROAS, which factors in agency fees, fulfilment costs and margin, not just top-line sales. It’s easy to show growth, he points out, when nothing supporting that growth is counted against it.
Pressed on the single number he’d take to a board, Guy lands on net profit. How much was generated, what it cost to generate it, and what was left afterwards. Without that worked out, he says, everything else reported alongside it is misleading. He’s watched clients chase aggressive growth in what he calls “vanity mode,” only for an owner, investor or PE house to eventually discover the business is making little money, or losing it. A good answer to the board’s question, in his view, looks like cost per new customer, cost per returning customer, and a clear profit position against break-even, so everyone can see what headroom there actually is to push harder.
There’s a wrinkle worth naming here. Treating a direct, tracked line from spend to sale as simply a data problem to be fixed isn’t the whole picture. Attribution models built on last click, or on any single touchpoint, are structurally weighted toward the channels that are easiest to track right at the point of sale. That can undervalue the upper-funnel activity that helped get the customer there in the first place. Worth keeping in mind before treating “just fix the tracking” as the complete answer.
Alex doesn’t think clients and agencies are usually aligned on measurement at all. In his experience, the gap sits with agencies that report on their own platform metrics rather than the client’s wider business objectives. Easy to measure isn’t the same as meaningful.
Rick Guttridge, founder and CEO of the Manchester PR, influencer and social agency Smoking Gun, sees a version of the same gap from the communications side. A lot of his work, by his own account, still starts with explaining what PR actually is, because it isn’t just media relations, and clarifying how social and influencer channels get measured, because that varies hugely by client.
The pressure he sees most often isn’t the client being unreasonable. It’s the client’s own stakeholders wanting one clean leading number, without much appetite for the nuance sitting underneath it. He gives a concrete example: a client asking for media coverage volume to keep rising year on year, as a sign of progress. Publishing is moving the other way. Reach plc has made repeated cuts to editorial jobs, which makes landing coverage harder. Its CEO has also said the company is moving away from volume and towards original content on topics distinctive to each brand. That leaves fewer routine articles, but more room for individual stories that PRs and brands can target. If a board doesn’t understand that context, a flat or falling coverage chart reads as failure when it might just be the industry around it changing shape.
This is where Rick draws a harder line. Clients still ask Smoking Gun to report advertising value equivalent, a metric the agency has refused to provide since it was founded in 2010, years before the CIPR and PRCA formally banned it from industry awards. The Barcelona Principles began that process the same year, and it’s been reinforced by the industry repeatedly since. AVE hasn’t quietly fallen out of fashion. It’s been formally rejected, more than once, for over a decade. And it’s still requested.
None of this means traffic, engagement and coverage numbers belong in the bin. The contributors are consistent on that too, just not in the way a simple “ignore vanity metrics” rule would suggest.
Rick’s preferred board number isn’t revenue at all. It’s key message penetration: whether the right story is reaching the right humans, and increasingly, the right AI systems summarising and surfacing that story elsewhere. He argues for a mix of leading and lagging indicators looked at together, with a real warning attached. Too many data points and too many charts create what he calls “data puke,” where the volume of reporting obscures rather than clarifies. He also points to last-click attribution as having caused what he calls an unhealthy obsession with short-term performance at the expense of brand health, and reflects that the longer he’s worked in marketing, the more he thinks marketers themselves get bored of a campaign and move on faster than the creative or the strategy actually wears out.
His version of a good answer to “what did we get for the spend” isn’t a single figure at all. It’s showing the full set of outputs and activities, and how each one feeds the next: has a PR campaign earned coverage, generated social conversation, and did that conversation move web traffic, search visibility and sentiment in a positive direction. The leading indicators aren’t the enemy of the revenue number. They’re the evidence trail that explains it.
Guy’s shift over his career points the same way, even from a performance background. What’s changed, in his view, isn’t the arrival of new tools so much as a clearer understanding of client cost structures and better attribution models for how channels work together, rather than treating each one as if it won the sale alone.
Alex frames the industry-wide shift in similar terms: data needs to be centralised and tracked properly, because the days of agencies getting away with reporting trends and vanity numbers in isolation are over. The spotlight now sits on outcomes.
Three agency leaders, three different answers: revenue tied to the customer journey, true profit after real costs, and message penetration read alongside a basket of leading indicators. None of them would accept reach or impressions alone as proof of anything. All three would reject a single number reported without the context around it.
What they’d agree on is the timing. The KPI that proves marketing ROI has to be settled before the campaign starts, not reached for afterwards to justify whatever numbers came back. A marketing manager with less channel experience can mistake a familiar metric for an important one simply because it’s the one they know how to pull. The board doesn’t need a familiar number. It needs the one everyone already agreed would count.
We’re taking this exact question to the floor at Crow Wood Hotel & Spa on Wednesday 7th October 2026. If you’re setting next year’s reporting framework and want to pressure-test it against people who build this for a living, that’s the room to do it in. Registration details are on the event page.
Our latest articles, features and ramblings.
We explore performance marketing, AI, communications and optimisation.
