What a business truly owns in its brand is whatever has stuck in customers' memory through each encounter, and that part is built over years but lost after one hasty change. This chapter is about seeing what the business actually holds in customers' memory before discussing what it wants to hold. Most of the craft's most expensive mistakes begin with changing something without knowing what it is currently worth.
Quick read: the core statement, when it comes up, and the test to take home. Full read also opens the reasoning, the limits and the sources. Each principle comes before its practices, so the numbers do not run in sequence.
Every encounter writes into the customer's memory, including the message a staff member answers on their own; a brand is also made in the encounters nobody designed.
When it comes upWhen a business defines brand work as making a logo and an identity, then is surprised that customers still remember it differently.
Resolution
Look first at the places customers meet the brand most often and at the heaviest moments (the decision to buy, a failed delivery, a complaint), not at what is easiest to sign off; how to split money by frequency and weight is in 6.02. The worst encounter rarely happens but can outweigh a hundred ordinary ones, so it is examined first. The mark is still needed, but it must not absorb all the effort while the places customers meet every day are left alone.
Full reasoning
Two goods pulling against each other
On one side is investment in what is visible and can be signed off: the mark, the colours, the set of printed materials, things with a handover date and a signature. On the other are the quiet touchpoints nobody designed: the reply to a customer message, the invoice, the way the delivery driver talks, the sales page staff post themselves. A touchpoint is any place a customer meets the brand.
Why it holds
Memory of a brand is a network of associations: the name linked to images, to feelings, to buying situations. Each encounter adds or thickens a link, and memory does not distinguish an encounter approved by the creative director from one a new employee made up. Marty Neumeier put the root idea briefly: a brand is not what a business says about itself but what people say about it. This principle goes one step further and asks which encounters are producing what people say. Encounters do not weigh the same: an encounter with strong emotion, especially a bad one, is recorded more deeply and retold, so it enters the memory even of people who never met the brand. A business cannot write directly into anyone's memory; it can only choose what to put in front of customers, how many times, and whether it stays the same. So the right question for the craft is not what the brand should mean, but through which encounters that meaning reaches customers' memory, and what the remaining encounters are putting there.
When it fails
A business with no customers yet has no memory to read; then the first encounters weigh more than any later ones and the initial design decides more than this principle suggests. In a business that sells to a few large clients through personal relationships, memory attaches to specific people more than to images, and most of the brand work lies in how those people work.
What it costs
Brand work becomes hard to package as an item with an end date; it has to discuss things beyond design, touching operations and the people inside, and business owners do not always want to hear that from a designer.
Sources
Marty Neumeier, The Brand Gap, New Riders, 2003
Kevin Lane Keller, Conceptualizing, Measuring, and Managing Customer-Based Brand Equity, Journal of Marketing, 1993
Jenni Romaniuk and Byron Sharp, How Brands Grow: Part 2, Oxford University Press, 2016
Ask ten real customers what they remember about the business, without prompting, before writing the brief.
When it comes upBefore writing the brief for a brand project, when everything known about customers comes from the business owner's words.
A test to take home
(1) Choose ten customers: six or seven recent buyers, mixing regulars and newcomers, plus three or four who buy rarely or have not bought for a long time; do not choose the owner's close customers. Record the two groups separately for comparison, because heavy buyers remember most clearly while most of a brand's customers are light buyers. The business contacts the customers itself, or tells them in advance and asks their consent for the interviewer to call; notes carry no names or phone numbers. (2) Ask in person or by phone, without showing images or reading the slogan: When you think of [name], what is the first thing that comes to mind? then: Anything else? (3) Write down the exact words, do not summarise. (4) Mark what three or more people mention. Pass: two or three things recur, and the brief builds on them or says plainly why it wants to change them. Fail: the answers repeat the business's own self-description word for word, usually a sign the interviewer prompted. If nobody remembers anything visual, the test has not failed; it is a finding: the business's visual identity has not entered customers' memory. If customers cannot be asked yet, the brief marks which parts are the owner's unchecked words and treats them as hypotheses.
Sources
Sources
Byron Sharp, How Brands Grow, Oxford University Press, 2010
Vietnam Personal Data Protection Law No. 91/2025/QH15
Decree 356/2025/ND-CP guiding the Personal Data Protection Law
The sentence customers use to introduce the business to people they know is the measure of the brand that actually exists.
When it comes upWhen the business already has a positioning statement and wants to know whether it lives outside the meeting room.
A test to take home
(1) Collect real introductions: ask customers When you recommend this place to someone you know, what do you say?, and read the natural recommendations that already exist (comments, map reviews, messages customers sent to others and chose to share). (2) Copy them word for word and mark recurring phrases. (3) Put them next to the business's positioning statement. Pass: the reason customers use to recommend it is the reason the business wants to be chosen for. Fail: customers recommend it for another reason, such as near home, cheap, knowing the seller; that is the brand that actually exists, and the brief must start there, not from the positioning statement. This test turns Neumeier's idea quoted in 1.01 into something measurable. What customers say can also be used as material for writing copy (7.08), but here it is a measure before it is material.
A brand asset is what people still recognise when they cannot see the name; everything else is doing another job, such as signalling the category or helping reading, and is not yet an asset.
When it comes upWhen weighing whether to keep or drop a colour, a shape, a packaging style or a familiar phrase of the business.
Resolution
What is already recognised beats what is prettier, unless it is recognised as a sign of the whole category rather than of this business alone.
Full reasoning
Two goods pulling against each other
On one side is the new, pleasing to today's designers, on trend. On the other is the old, sometimes clumsy, that customers have learned to recognise.
Why it holds
An element becomes an asset only when two things hold: many people link it to this brand, and few people link it to other brands. Missing the first, the element is unknown; missing the second, it is shared, it evokes the category rather than the business. Removing the name is the most direct way to separate the two, because while the name stands there everything next to it looks recognised. People inside the business are especially prone to this mistake: they see their name everywhere, so they never learn what their visuals say on their own.
When it fails
A new business has no assets yet; then this principle becomes a question of choosing candidates: choose elements that can become distinctively yours, not the shared codes of the category. An element recognised because of an incident or scandal is a negative asset; being widely recognised does not mean it should be kept. An element that is recognised but hard to register as exclusive, such as a single colour, is still an asset but an easily copied one: keep it, and always pair it with a protectable element such as the name or the mark.
What it costs
It forces the business to keep things the people inside are tired of and designers find dated; the project looks less new than the money spent.
Sources
Jenni Romaniuk, Building Distinctive Brand Assets, Oxford University Press, 2018
Cover the name and ask customers whether they still recognise it; an element recognised only when the name stands beside it is not yet an asset.
When it comes upBefore deciding to keep, fix or drop an identity element; before any refresh project.
A test to take home
(1) Choose five to eight elements in use: the core colour, the mark's shape, the typeface, the pack shape, the slogan, the character, the pattern. Isolate each in its own image and remove every trace of the name and brand lettering. Test a colour on a neutral shape from the category, not as a flat swatch, because a flat swatch almost always fails even when the colour is already an asset. (2) Show them to about ten buyers in the category, not employees, not relatives, and ask an open question: Which brand do you think this belongs to? Do not offer a list to choose from. Show one image at a time and change the order between people, because earlier images prompt later ones. (3) For each element record two numbers: how many people named the right brand, how many named another brand or only the category. Pass: most name it correctly, very few name another brand. Fail: most name only the category (a shared code of the industry) or name another brand (an element advertising for someone else). Ten people is not a statistical study; it shows a direction. Large decisions, such as dropping a colour used for many years, should be measured with a larger sample.
Sources
Sources
Jenni Romaniuk, Building Distinctive Brand Assets, Oxford University Press, 2018
Changing an identity spends the recognition already built; change only when what is kept is worth less than what the change gains.
When it comes upWhen a business wants to redo its entire identity, especially when the stated reason is that it looks old, feels stale, wants to look younger.
Resolution
The reason to change must lie in the business and its customers, not in the feelings of people inside; and the change should be only as large as what is actually wrong.
Full reasoning
Two goods pulling against each other
Keeping protects the recognition already built. Changing can send a new signal and fix what is truly wrong: the business model has changed, the customers have changed, the mark does not work on small screens.
Why it holds
Recognition builds up through the number of times people meet the same identity asset again. When that asset changes, the counter for the changed part drops close to zero: regular customers look for the old pack on the shelf, cannot find it, think it is out of stock or buy the product next to it. People inside get bored before customers have had time to remember (12.01), yet customers are the ones holding that recognition. A new pack may not be ugly at all; it simply erases the things customers used to find the product (the often cited case: Tropicana orange juice in 2009, where the old pack was brought back in under two months after regular customers complained).
When it fails
A mark tied to a crisis or scandal, a mark that cannot be protected or is in a rights dispute, a merger or a complete change of industry: then the old recognition may be a burden and a bold change is right. When a business still has very few customers, what has been built is small and so is the cost of changing.
What it costs
The owner spends money and gets back the feeling that nothing is new; the practitioner loses the chance to make something showy for the portfolio.
Inventory the identity assets in use and sort them into keep, fix and drop before drawing a single new line.
When it comes upThe first step of any refresh or standardisation project for a business that has been operating for some time.
A test to take home
(1) Collect every touchpoint in use as real photographs: signs, vehicles, uniforms, packs on the shelf, invoices, web pages, profile pictures, email signatures. Do not take them from the original design files, because the files do not show what has drifted after passing through many hands. (2) List the elements that recur across those touchpoints. (3) For each element ask two questions: is it recognised (by 1.04, or because customers name it themselves), and does it work at every touchpoint (small size, one colour, difficult materials). Recognised and workable: keep. Recognised but technically broken: fix, keeping exactly the part that makes it recognisable. Recognised by nobody and not distinctive: drop. Pass: each element has a group and a one-line reason, and the design brief is written from that table. Fail: the table has only a drop column, or the inventory is done after a new direction has been chosen.
Refresh in steps: keep what is recognised, adjust what has aged, and move to a higher step only when the lower one cannot solve the problem.
When it comes upWhen a refresh has been decided and the question is how much to change.
A test to take home
(1) Arrange the degrees of change into steps: step one is technical refinement (strokes, spacing, one-colour version, small-size version), customers barely notice; step two refreshes what surrounds the mark (secondary typeface, photography, layout, patterns), keeping the mark and core colour; step three redraws the mark while keeping the shape that makes it recognisable; step four changes everything. (2) Write the real problem in one sentence, then ask from step one: does this step solve it? Move up only when you can say why the lower step is not enough. (3) Before settling on a step, count the transition cost: packaging and printed stock, signs, vehicles, uniforms; put the stock run-out schedule into the transition plan. That cost is a legitimate reason to stop at a lower step. (4) At steps one to three, test the new version with the name-covering test (1.04) on regular customers. Pass: regulars still name it correctly and can say it looks newer. Fail: regulars ask whether the business has changed owners, or cannot name it. (5) At steps one to three, replace touchpoints gradually along their natural cycles, packs and stationery with each print run, signs later, so customers meet the new beside the old without a break. At step four, do not mix gradually: a completely different old and new version on the same shelf easily makes customers suspect a counterfeit. Set a switch date, the day the business stops printing and shipping the old version; inform distributors in advance, exchange or recall stock at key outlets, and keep a shelf notice about the new pack for as long as old stock remains in small shops, along with announcements on every channel.
