The CFO asks a fair question: “What did that brand campaign return?” The marketing team gives the usual answer: “Brand is hard to measure.” Both walk out frustrated. The CFO thinks marketing is hiding. Marketing thinks finance doesn’t understand. Neither is quite right.
Summary
No. You can measure brand marketing, just not with last-click reports that only count sales in the next few days. Brand work pays off slowly, through more people searching for you, choosing you first and accepting your price. Track signals like branded search, share of search, direct traffic, brand lift studies, holdout tests and marketing mix modelling, and judge them over months, not days.
Internet Culture · Overthinking Marketing · 006
The belief: “Brand marketing can’t be measured, so it’s a gamble.”
The honest answer: It can be measured. It just works on a slower clock than performance ads.
Why this is worth ten minutes
- Teams that can’t measure brand marketing end up cutting it, then wonder why ad costs keep rising.
- “It can’t be measured” usually means “we’re using the wrong ruler”.
- A handful of simple signals, most of them free, can show whether brand spend is working.
1 The room
A D2C brand runs a TV and YouTube campaign during a cricket series. Two weeks later, the performance dashboard shows almost nothing. Few clicks. Few attributed sales.
Finance wants to cut brand spend next quarter. The CMO argues back but has no numbers. The budget gets cut.
Six months later, Meta CAC has crept up. Branded searches are flat. The sales team says fewer people seem to know the brand when they call.
Nobody connects the two. They were measured on different clocks. The cut looked smart in week two. It looked expensive by month six.
2 Why does brand marketing look impossible to measure?
Because most dashboards were built to measure something else.
Performance marketing works fast. Someone sees an ad, clicks, buys. Tracking connects the click to the sale. Clean and quick.
Brand marketing works slowly. Someone sees your campaign today. Weeks later they need your category. Your name comes to mind first. They search for you directly, or walk past five rivals on a shelf and pick you. No click links the two moments.
So last-click reporting gives brand work almost no credit. It isn’t that brand had no effect. It’s that the report can’t see it.
This is the core of the brand vs performance marketing debate. They aren’t competing for the same job. Performance captures people who are ready to buy now. Brand makes sure more people pick you when they’re ready later.
3 How do you measure brand marketing? 6 signals that work
01. Branded search. How many people search your brand name on Google? You can see this in Google Search Console’s branded queries. If brand work is landing, more people look you up directly.
02. Share of search. Your brand’s searches divided by total searches for all brands in your category. Les Binet, presenting work through the IPA in 2020, showed share of search often tracks closely with market share, and moves before it. You can estimate it for free using Google Trends.
03. Direct and organic traffic. People who type your website address or arrive without an ad. Rising direct traffic usually means more people remember you.
04. Brand lift studies. Meta and Google offer studies that compare people who saw your ad with a similar group who didn’t, then ask both about awareness or intent. They show short-term memory effects.
05. Holdout tests. Run the campaign in some cities and not in matched others. Compare sales, branded search and installs over the following months. It’s slower, but it’s the closest thing to a clean answer.
06. Marketing mix modelling (MMM). A statistical model that looks at sales over time alongside spend on each channel, plus seasons and prices. Free open-source options like Google’s Meridian and Meta’s Robyn have made it more accessible, though it still needs clean data and someone comfortable with numbers.
THE SLOW PROOF RULE
Brand marketing proves itself slowly. Measure it with signals that move early, judge it on outcomes that move late, and never judge it with a ruler built for clicks.
Rule 1. Pick early signals. Branded search, share of search and direct traffic move within weeks.
Rule 2. Judge late outcomes. Market share, CAC trend and how well you hold price move over months.
Rule 3. Agree the clock before the campaign. Tell finance up front when you’ll report, and on what. Surprise deadlines kill brand budgets.
4 The mechanism stack
01. Brand makes performance cheaper. When more people know you, your ads get more clicks and convert better. That’s why cutting brand spend often shows up later as rising CAC. The damage lands in the performance report, not the brand one.
02. Brand protects price. Customers who know and trust a brand are less likely to switch for a small discount. If you can hold your price while rivals discount, brand is working.
03. The effect builds and fades slowly. Les Binet and Peter Field’s research for the IPA, published as The Long and the Short of It in 2013, found that brand effects build over longer periods, while sales activation effects are strong but short-lived. Their often-quoted guideline was roughly 60% brand and 40% activation, though they later stressed the right balance varies by category and situation.
04. Memory, not clicks, is the product. Brand marketing works by making you easier to recall when someone needs your category. That’s hard to see in a click report and easy to see in search behaviour.
5 The counter-weight
Some brand spend really is wasted. “It can’t be measured” is sometimes an excuse for campaigns that didn’t work. Measurement protects good brand work and exposes bad brand work. Both are useful.
Small brands may not have the data. Share of search needs enough search volume to read. MMM needs months of steady data. Early-stage brands should start with branded search, direct traffic and simple city holdouts.
Every method has blind spots. Brand lift studies measure recall, not sales. MMM depends on its assumptions. Holdouts are slow. Use two or three methods together and look for agreement.
Performance-first can be right early on. A new brand with little cash often needs to capture demand that already exists before building more. Brand spend makes most sense once there’s something worth remembering.
Internet Culture verdict desk
Brand marketing isn’t impossible to measure. It’s impossible to measure in two weeks, and most teams only ever give it two weeks.
6 Steal this
01. Set up branded search tracking in Google Search Console today. Note your current monthly branded clicks and impressions.
02. Build a simple share of search chart in Google Trends: your brand plus your top three rivals, past five years.
03. Before your next brand campaign, agree with finance on three signals and a reporting date three to six months out.
04. Run your next campaign in some cities and hold out matched others.
05. Track your CAC trend alongside brand spend over twelve months. Look for the link.
06. Report brand and performance on the same slide, so leadership sees how one affects the other.
7 Frequently asked questions
Can you measure brand marketing?
Yes. You can measure brand marketing using signals such as branded search, share of search, direct traffic, brand lift studies, geographic holdout tests and marketing mix modelling. The key is judging results over months rather than days, because brand effects build slowly.
What is share of search?
Share of search is the percentage of all brand searches in a category that are for your brand. Les Binet’s work through the IPA in 2020 showed it often tracks closely with market share and can move before it. You can estimate it for free with Google Trends.
What is the difference between brand and performance marketing?
Performance marketing aims for quick, trackable actions like clicks and purchases from people ready to buy now. Brand marketing builds memory and preference so more people choose you later. They work on different timelines and support each other.
How do you calculate brand marketing ROI?
Brand marketing ROI is usually estimated rather than calculated precisely. Common approaches include geographic holdout tests, marketing mix modelling, and tracking how brand spend relates to sales, branded search, CAC and pricing power over time.
What is marketing mix modelling?
Marketing mix modelling is a statistical method that estimates how much each marketing channel contributed to sales, while accounting for factors like seasons, prices and promotions. Open-source tools such as Google’s Meridian and Meta’s Robyn have made it easier for smaller teams to try.
