Brand Recall

For a long time, the way to ask how well known a brand was involved showing the name and asking whether people had heard of it. Slip your name into a list, ask respondents to tick the ones they recognize, and the number usually comes back high. Seeing a name makes almost anyone feel they have run into it somewhere.

The trouble was that the number didn't connect to the moment of purchase. In the same week marketing reported 40 percent awareness, sales came back from every meeting saying nobody had heard of the company. One of those looks wrong, but both were accurate. The two teams were asking different questions.

When a customer picks a tool, there is no list with your name on it. A need comes up, candidates surface from memory, and any name that doesn't surface never gets searched for or compared. Brand recall exists to imitate that moment. Take the options away, name only the category, and count how often your brand still comes out.

Give a sample the product category with nothing else, ask them to write down whatever brands come to mind, and divide the number who wrote your name by the number who answered. Because no options are offered, this is also called unaided recall.

The instrument is a survey — the same question put to many people so you can read the distribution of answers. It runs in four steps.

1. Decide who to ask. Not the general population, but people who would plausibly choose this kind of product. For Pagedeck that means office workers who have a say in which document tool their company uses. Widen or narrow that definition and the same product can score two or three times differently.
2. Put the unaided question first. Something like "which collaborative document tools come to mind? List as many as you can" names the category and offers nothing else. If any brand name appears earlier in the survey, a cue exists from that point on and the question is spent.
3. Divide the count by the number of respondents. Twenty-four people out of 400 gives 6 percent. Misspelled and abbreviated answers still count, which means a person has to read through the responses.
4. Record the order too. The people who wrote your name on the first line make up top-of-mind recall. If 5 of those 24 put you first, recall is 6 percent and top-of-mind is 1.25 percent.

Small samples can't carry much weight. Six mentions out of 200 is 3 percent, and two more mentions would make it 4 percent. There is no way to tell whether a point of quarterly movement is real or just a different set of respondents, so surveys you intend to compare should hold the sample size and the exact wording steady.

There is no absolute line above which recall counts as good. The same 5 percent means different things depending on how wide the category is and how many candidates compete in it. Five percent is weak in a market with three or four serious options and strong in one with dozens. So the number gets read against three other things rather than judged on its own.

The first is its ratio to brand awareness, which shows the name and asks whether people know it and therefore always runs higher. Awareness of 40 percent against recall of 3 percent means only one in thirteen people who know the name pull it up as a candidate. The name has spread without attaching to the category, and the fix is not more impressions but positioning or the tagline. Keep buying reach while it's unclear what the company is for, and only the awareness number moves.

The second is the last reading taken under the same conditions. Recall moves slowly. One or two points a quarter is a normal pace, and a number that doubles in a month points to changed wording or a changed respondent pool before it points to changed perception. To isolate what a single campaign contributed, use something like brand lift, which compares people exposed to the campaign against people who weren't.

The third is the distance to competitors named in the same survey. Most response sheets carry three or four brands, so competitor figures arrive with yours. Recall of 6 percent next to a competitor at 30 percent is a different situation from 6 percent next to 8 percent.

"Awareness went up but recall didn't move. I don't think more ad spend is the answer here."

The name is spreading without attaching to the category, so the message needs work rather than the budget.

"Twenty-four out of 400 respondents wrote our name without being shown any options."

Reporting recall of 6 percent by walking through how it was counted.

"Top-of-mind still belongs to the competition. We show up second or third."

We made the shortlist, but we're not the first name anyone reaches for.

At Pagedeck, Ravi runs the same survey every quarter and brings the number back. When Grace is assembling an investor deck and asks how well known the company is, Ravi reads awareness and recall side by side. If the gap between them widened since last quarter, what he puts on the table is not how much was spent on ads but what the ads actually said.

A low number doesn't start a budget conversation either. The first thing Ravi looks at is the language on the response sheets. When people write "that tool for organizing docs" instead of a name, the link between the brand and the category doesn't exist yet, and next quarter's work becomes one line of copy rather than more spend.

The most common mistake is reading low recall as a spending problem. More impressions do produce more people who know the name, so the assumption is reasonable enough. But this metric asks whether the name arrives when the category is mentioned, not whether people recognize it. Until what the company does is fixed in a single sentence, extra reach only makes the name drift past. Leave it alone and the budget doubles while recall sits still, and the same meeting happens again next quarter.

Reading recall straight through to revenue is the other one. Recall says a name reached the shortlist; whether it then gets chosen is a separate stage decided by price, features, and sales work. To measure what the name itself is worth, look at something like brand equity, which compares what people will pay with and without it.