Resonance Engine adds four questions to the brand tracker you already run. For every brand in your category, it shows which customers are committed, which could walk, and which non-customers are already open to you.
About 10% on a used family MPV (Rumion vs Ertiga). On a city car (Vitz vs Celerio), nothing we could tell apart from zero. Same factory, same engine, different badge.
Two market leaders, two white commodities bought every week. One name charges 79% more per kilogram than the store brand next to it. The other, 5%.
South Africa's largest retail bank by clients. 38% of its active clients meet the bank's own fully banked bar, which requires regular monthly income paid in and everyday use of the account.
All three figures come from public market data. In each case the brand is well known and widely used, yet what the name is worth varies enormously. Awareness, usage and consideration, the measures most trackers stop at, can't explain gaps like these.
Resonance Engine adds the missing layer: for each person and each brand they know, whether they are committed to the name or just using it.
Three questions describe the person's relationship with each brand. The fourth asks how much brand matters in the category at all, because in some categories it barely decides the purchase.
When you need something in this category, does this brand surface, and is it clear what it stands for?
Not whether the brand is good, but whether the person sees it as a brand for them.
The warmth of the relationship. Brands with high affinity are the ones people would miss.
Asked about the category, not the brand. It tells you how much a strong brand can earn in this market.
We ask about competitors and store brands as well as yours, including brands the person doesn't buy. That is how we find people who are drawn to you before they become customers, and see your competitors' committed base next to your own.
We combine the answers with whether the person uses the brand. The result is a segment for each person against each brand they know, so your tracker shows the size of each group, wave by wave.
Use the brand and are committed to the name. This is the base that holds when a competitor makes a better offer.
Use the brand, but out of habit, price or convenience. They count as customers today and could walk tomorrow.
Don't buy the brand yet but are already drawn to it. Most trackers file them as non-customers and move on.
Neither use the brand nor feel drawn to it. Spend here builds familiarity first, and pays back slowly.
Most brand spend treats every non-loyal customer the same. Some are committed to a competitor and won't move. Others sit just below the line, and would respond to the right signal.
For each brand, Resonance Engine shows:
The four questions, the four segments, and the people sitting one step short of the line.
A survey measure is only worth tracking if it lines up with behaviour. We've checked Resonance Engine against real transactions in two categories that could hardly be more different.
The segments came from survey answers alone. When we matched them to the operator's account data, Partial players explored on promotional tokens and didn't deposit, while Core players stayed engaged without the incentives.
Using a major grocery retailer's loyalty transaction and shopper panel data, Core shoppers spent more each month and shopped more often than Partial shoppers, after allowing for household size and demographics.
The data belongs to the retailer, so we can't share the figures.
We sometimes publish simulated data to show what the output looks like. It is always labelled as simulated, and we never treat it as proof that the engine works. When we tested whether AI-simulated shoppers could stand in for real ones, they couldn't.
Each study uses market data anyone can check, and tests an idea behind Resonance Engine in the open.
Toyota sells Suzuki-built cars under its own badge. Matching trim, year and mileage, the badge is worth about 10% on a used family MPV and close to nothing on a city car.
What it taught us: brand equity depends on the category. That's what the Gravity question is for: how much brand decides the purchase in each category.
Across 14 food categories, leading brands charge 5% to 171% more than the store brand. We asked AI-simulated shoppers to rate the brands, blind to prices. They rated every leader above its store brand, but the size of the lead didn't track the premiums.
What it taught us: AI can guess which brands are famous. Only real buyers can tell you which ones they're committed to.
Capitec's active clients grew 17% in two years. Its fully banked clients grew 27%. Most banks publish customer counts, and few publish how many customers bank with them first.
What it taught us: a customer count tells you who uses you. It doesn't tell you who would stay.
Follow the series on LinkedIn, or read how the engine was first developed in our longer article.
No new study and no new sample. The four questions sit alongside your existing brand measures, and your supplier fields them as usual.
We agree the brand list with you, including store brands and smaller competitors, and send your supplier the question wording and routing.
The questions are asked for every brand each respondent knows. Nothing else about your fieldwork changes.
We score every respondent against every brand and report the segments with your other tracker measures, so each wave adds to the trend.
How committed each person is to each brand they know. We ask four questions about every brand: does it come to mind (Signal), is it right for someone like me (Fit), how do I feel about it (Affinity), and how much does brand decide what I buy in this category (Gravity). The answers, combined with whether the person uses the brand, place them in one of four segments for that brand: Core, Partial, Emergent or Static.
Awareness, usage and consideration tell you who knows you, who buys you and who might. They don't tell you which of your buyers would stay if a competitor made a better offer, or which non-buyers are already well disposed to you. Resonance Engine separates committed users from users who could walk, and open non-users from people who aren't interested.
Core: people who use the brand and are committed to it. Partial: people who use it but aren't committed, so they could walk. Emergent: people who don't use it yet but are drawn to it, so they are open to you. Static: people who neither use it nor feel drawn to it. Every respondent gets a segment for every brand they know, so you see your own base and your competitors' side by side.
Yes, in two very different categories. In online gaming, the segments from the survey matched real deposit behaviour: Partial players took the free tokens but didn't deposit, while Core players stayed engaged without the incentives. In grocery, using a major retailer's loyalty transaction and shopper panel data, Core shoppers spent more and visited more often than Partial shoppers. That data belongs to the retailer, so we can't share the specifics.
No. The four questions go into the brand tracker you already run, asked for every brand each respondent knows. We supply the question wording, check the brand list with you, and score the data after each wave. If you don't have a tracker, it can also run as a standalone study.
We tested that in public. AI-simulated shoppers ranked the leading brands above store brands, but they did not reproduce the price premiums shoppers actually pay. We use simulated data only to show what the output looks like, and we label it as simulated. Client findings come from real respondents.
Bring your current tracker questionnaire. In 30 minutes we'll show you where the four questions would sit, what the segments would tell you in your category, and what it would take to add them to your next wave.
The booking calendar is provided by Cal.com. Opening it loads content from Cal.com and may set third-party cookies.
30 minutes · Video call · With Greg Streatfield, founder of Knowsis
Prefer email? Send us a message
