Brand loyalty: what it is and how it is formed

Behavioral and affective loyalty are not two levels of the same phenomenon, but different logics

Behavioral loyalty is what is visible in cash register receipts: the customer buys from you again and again. Affective loyalty is what happens in the customer's head: he chooses a brand not because it just happened to be that way, but because he believes in it. The mistake of most companies is to think that the first gradually turns into the second. In practice, this is not the case: you can buy the same laundry detergent for years simply because it always stands on the same shelf in the neighboring store, and yet feel nothing toward the brand.

For long-term business, affective loyalty is more important for one reason: it does not depend on whether a cheaper alternative appears nearby. Behavioral loyalty lasts until the first serious temptation — a competitor's discount, product shortage, or logistics changes. Affective loyalty withstands both, because the customer's decision is made not at the moment of purchase, but much earlier, at the level of their relationship with the brand.

Loyalty matrix: how to distinguish true attachment from habit

The most common methodological mistake is to consider loyal any customer who purchases frequently. In reality, frequent purchases with low satisfaction are not loyalty, but dependence on the lack of alternatives, and this breaks down at the first moment when an alternative appears. The matrix of 'satisfaction — frequency of purchase' helps distinguish these cases.

An example of false loyalty is a customer who for years takes the same banking product not because they are satisfied, but because changing banks requires effort. An example of latent loyalty is a person who sincerely loves a particular clothing brand but buys from it once a year because the store is far from home. The difference between these cases is critical: latent loyalty can be converted into true loyalty by removing the barrier (for example, by offering delivery), whereas false loyalty cannot be changed — in that case, the product itself needs to be altered.

The threshold, after which behavior can be called loyalty and not randomness, is at the level of the majority of purchases in the category (when a customer prefers your brand in most cases rather than switching between competitors). If the share is below that, you are dealing with a 'switcher' who chooses between several brands without a clear preference.

What actually forms loyalty: product versus service versus emotion

The debate over whether price, quality, service, or emotional connection is more important is not resolved by 'either-or,' but rather by a sequence. Product quality is an entry ticket, not a competitive advantage. Consumer behavior studies show that simple customer satisfaction is insufficient for building loyalty. Only the effect of exceeding expectations works, when perceived quality (P) is notably greater than expected (E). Simply 'good, as promised' results in zero emotions and zero attachment.

• Product — a necessary but insufficient condition; without it, everything else becomes meaningless.
• Price — only effective for behavioral loyalty and instantly loses power when a cheaper alternative appears.
• Service and staff — amplifying factor: employee qualifications enhance brand attractiveness, but do not compensate for a weak product. A good salesperson will not save a bad product, and will even spoil the impression of an excellent one.
• Emotional connection — a defining factor specifically for the brand as such, unlike loyalty to a retail outlet, which is based on functional aspects: convenience, discount programs, omnichannel experience.

Here it is worth distinguishing two different objects of loyalty, which are often confused. Brand loyalty is built on emotions — trust, recognition of oneself in the brand, habit of recommending it to friends. Loyalty to a specific network or sales point is built on functionality: it is close, convenient, has the right size, and the app works. These are different constructs, and an attempt to purchase emotional loyalty through functional tools (discounts, bonuses) usually yields only false loyalty from the top row of the matrix.

Survey metrics and CRM data show different pictures — and this is not a bug, but a norm

NPS and CSI measure declaration — what the customer says about their attitude at a specific moment, often right after a positive experience. CRM data measures fact — what the customer actually does over months. The gap between them is not an exception, but a regularity: in real databases, it is often found that customers with neutral ratings behave more stably than customers with maximum scores — the latter gradually reduce activity after the emotional surge.

The reason is simple: a high score in a survey often reflects a momentary impression (a fortunate consultation, a pleasant discount), not stable behavior. Relying solely on survey metrics is like making managerial decisions based on noise. What should really be tracked:

• Repeat rate — the share of customers who returned for a repeat purchase within a specified period.
• The share of repeat purchases out of the total number of transactions — with a threshold of 67% as the boundary of true loyalty.
• Churn rate — the rate of churn, especially in dynamics by cohorts.
• CLV (Customer Lifetime Value) — the total customer value over the entire interaction period.
• Purchase frequency over a fixed period — not occasionally, but as a trend for at least 3-4 purchase cycles.

The right practice is to correlate NPS with behavioral data for the same customers, rather than publishing them separately. If a high NPS is not supported by an increase in repeat rate among the same respondents — the company is measuring politeness, not loyalty.

Where exactly does loyalty begin: not in a points program, but in the first touchpoints

A common misconception is that loyalty is formed through a loyalty program. In reality, the program only records what has already happened at the level of impression: it works well for customers who already have a positive relationship, and almost does not work for those who simply choose by price.

The real starting point is the first order and how much the experience exceeded expectations at that moment: delivery speed, support's reaction to an unusual question, how the packaging looks. Then comes the stage of repeat choice — the moment when a customer decides whether the first impression was a coincidence or a pattern. It is at this stage that it is decided whether the customer will fall into the category of true loyalty or get stuck in false loyalty, continuing to buy on inertia.

A good working example of how this looks in practice is the Telegram bot of the ZARINA brand, embedded in the loyalty program. Through it, part of the content is available only to channel subscribers — 38% of the program's audience receives information exclusively this way. The result is a 30% increase in revenue per program member and a 20% increase in purchase frequency. This works not because the bot is a trendy tool, but because it creates exclusivity and a constant reason for contact, rather than just a discount for points.

How to put this together into a working system

If we reduce everything said to practice, it turns out to be the following sequence of actions for assessing real loyalty in the company:

1. Calculate the share of repeat purchases in the customer base and compare it with the threshold 67%.
2. Distribute customers according to the matrix satisfaction/frequency — true, false, latent, lack of loyalty.
3. Compare NPS/CSI with behavioral metrics for the same customers, not averages.
4. Evaluate the emotional component (willingness to recommend the brand without stimulus) and the functional component (convenience of the sales point) separately.
5. Check whether the actual customer experience exceeds expectations — not just meets them.

Such an approach does not yield a pretty number for the report, but it does provide an honest picture: where loyalty is real and where the company simply benefits from the lack of alternatives for the customer. The second situation is riskier than the first precisely because it looks like success until an alternative appears.