Marketing KPIs for CMOs: the metrics that measure growth

Too many metrics, too little meaning

Most marketing dashboards share one flaw: they measure everything measurable and none of it drives a decision. A screen full of numbers looks responsible, but it's really a defensive posture. When you track fifty metrics, one is always going up, so there's always something to report, and yet nobody knows whether the company is winning or losing.

A simple test for any metric: does its movement change a decision of yours? If it spiked or crashed, would you do something differently? If not, it belongs in an appendix, not on the dashboard. A good metric isn't the one that looks nice on a slide, it's the one you can act on. And there are surprisingly few of those.

Leading vs lagging: this is where it's decided

The key is to split metrics into two families. Lagging metrics tell you how the past turned out: revenue, margin, market share. They're true, but they arrive too late to steer a campaign. Leading metrics predict what's coming: how available the brand is in the buyer's mind, how loudly it speaks versus rivals, how many people search for it. Those move before revenue does.

These leading metrics were exactly what the original LinkedIn list missed, and they're the ones that buy you time to react. Mental availability (how easily a buyer recalls you in a buying situation) is, per the Ehrenberg-Bass Institute, one of the main engines of growth, because brands grow mainly by acquiring new buyers, not through loyalty. When mental availability rises, revenue tends to follow.

Brand metrics with a foundation in data

Brand building long hid behind "you can't measure it". That's no longer true. We have several metrics with hard evidence:

Share of Voice and ESOV. Per Les Binet and Peter Field (an analysis of 996 campaigns from the IPA Databank), a brand grows when it speaks loudly relative to its size. Specifically: every 10 points of excess share of voice (ESOV) yields on average about 0.5 percentage points of market-share growth per year. It's one of the few genuinely predictive equations marketing has.

Share of Search. A clever, cheap proxy. It's your brand's share of all category searches, and you measure it for free via Google Trends. Les Binet presented it at IPA EffWorks 2020 as a fast, predictive metric that usually precedes market share by 6 to 12 months. Mind the context though: it works beautifully in consumer categories, in B2B it's less reliable and you need to clean the search composition (brand vs product vs competitor). Treat it as a useful proxy, not a hard target.

Distinctive brand assets and mental availability. How well people recognise you without the logo, and how many buying situations bring you to mind. That measures the brand's work better than an abstract "awareness" figure. The method is laid out by Jenni Romaniuk of Ehrenberg-Bass in Building Distinctive Brand Assets.

Business metrics: the scoreboard, not activity

Lagging metrics are still necessary, just don't confuse them with steering. On a CMO dashboard belong: revenue growth, but above all gross-profit and margin growth (revenue without margin is vanity), growth versus the category (are you outpacing the market or just riding it?), ideally with market share, and penetration, the number of active customers. Penetration, not loyalty, is per Byron Sharp and Ehrenberg-Bass the main lever of brand growth.

Worth adding a point from the discussion under the original post: revenue per employee (is the company growing efficiently, or just by hiring?) and, for companies that scale, a breakdown of conversion and CAC by client size and segment. An average across the whole funnel hides the fact that enterprise and self-service behave completely differently.

Efficiency: count lift, not the last click

The trio of CAC, LTV to CAC and CAC payback belongs on almost every growing company's dashboard. Mind two things, though. First, LTV is sensitive and easily inflated in early stages, so take it with a grain of salt (and treat it more as customer-base analytics à la Peter Fader than one magic number). Second, and this is crucial: count return through incrementality, not last-click attribution. Last-click ROAS credits whichever channel happened to be last and systematically overvalues the bottom of the funnel. How to do it without cookies (MMM, geo-experiments, incrementality) we cover in a separate article.

What to demote or cut outright

Now the unpopular part. Some beloved metrics don't belong on a CMO dashboard as growth predictors.

NPS as a growth predictor. The longitudinal study by Keiningham et al. in the Journal of Marketing (2007), on data from 21 firms and 15,000+ interviews, failed to support the claim that NPS is a clearly superior predictor of revenue growth over other satisfaction metrics. The paper, incidentally, won the MSI/H. Paul Root Award. Track NPS as an operational gauge if you like, but don't run strategy or budget off it.

Last-click ROAS as proof of efficiency, vanity metrics (likes, impressions, "buzz" with no tie to the business) and Share of Search used blindly in B2B. All of these soothe more than they steer. And one for the room, raised in the original discussion too: AI efficiency is more an internal operational parameter than a marketing growth metric. Useful to know, but it doesn't belong on a CMO dashboard next to market share.

It depends on the model (and stage)

There is no universal KPI list, and that's actually the main point. A sales-led model adds sales-cycle length, pipeline coverage and CAC payback; self-served focuses on activation, retention and time to first value. B2B versus B2C changes the weight of the brand: per the 95-5 rule from the LinkedIn B2B Institute and Ehrenberg-Bass, roughly 95% of B2B buyers aren't in the market at any given moment. Most of your audience just needs to remember you for later, so mental-availability and brand metrics carry far more weight than a short-term performance dashboard would admit.

And since a CMO leads people, not just campaigns, it makes sense to track team eNPS too. Here, unlike customer NPS as a growth predictor, the purpose is different: the health and stability of the team that has to deliver the growth.

Sources: Keiningham, Cooil, Andreassen & Aksoy, A Longitudinal Examination of Net Promoter and Firm Revenue Growth, Journal of Marketing (2007) · Binet & Field, The Long and the Short of It and ESOV work (IPA) · Les Binet, Share of Search as a Predictive Measure (IPA EffWorks 2020) · LinkedIn B2B Institute & J. Dawes (Ehrenberg-Bass), How B2B Brands Grow (the 95-5 rule) · Byron Sharp, How Brands Grow and Jenni Romaniuk, Building Distinctive Brand Assets (Ehrenberg-Bass) · Peter Fader (Wharton) on customer lifetime value. This article was prompted by a public LinkedIn discussion (Ivo Mrena and others); thanks for the inspiration.

Related to what we do: marketing strategy, growth & performance and RevOps & CRM. Why funnels and AIDA don't work is in a separate article, and measurement after cookies in this one.