June 18, 2026

Knowing a Brand No Longer Means Choosing It

By Sandra Radlovacki

For as long as marketing has had a budget, the theory held firm: recognition does the heavy lifting. Get a customer to know your brand, trust it, and recall it at the moment of need, and the purchase tends to follow. Awareness fed preference, preference fed the sale. The chain is coming apart, and the thing pulling it apart is the AI assistant a customer now consults before deciding almost anything.

The behaviour is already common enough to measure. When someone asks ChatGPT, Claude, or Gemini what to buy, where to stay, or which provider to trust, the model does not honour the customer’s history with a brand. It runs the comparison fresh and recommends whatever fits the request. Recognition still earns a place in the customer’s memory but no longer earns a place in the basket.

Awareness No Longer Equals Sales

According to Amperity’s 2026 Consumer Priorities Report, 60% of consumers say AI has led them to choose a brand they had never previously considered, and only 23% go straight to brands they already know without weighing what AI recommends.

The majority of generative AI users research products or plan travel through these tools, and many of them sometimes or often act on the results by clicking, buying, or booking. The familiar brand is no longer the default. It is one option among several the customer asked a machine to rank.

None of this retires the loyalty programme. Amperity ranked programmes, rewards, and consistent cross-channel experiences among the strongest drivers of loyalty it measured. The mistake is treating the programme as insurance against switching. With 63% of consumers willing to move brands for a better offer, a points balance counts for little against a model that has just surfaced three cheaper, faster alternatives.

The loyalty programme best practices earn their keep because they start from customer behaviour and friction rather than reward mechanics, and predictive loyalty programmes go further, reading churn signals before a customer drifts instead of reacting once they have gone quiet. A programme works when it makes the brand the relevant answer to the customer’s question, never when it sits in the background as a sunk cost.

Personalisation Is The Trap

Here is where brands told that relevance wins will overcorrect, reaching for personalisation and getting it wrong. Seventy-two percent of consumers say personalised experiences are important when choosing a brand, yet 58% say invasive or “creepy” personalisation makes them less likely to choose one, and 78% engage only when they trust how their data is used.

The distance between useful and unsettling is thinner than most teams assume. On Valentine’s Day last year, Deliveroo sent select customers handwritten love letters from a secret admirer with a flowers voucher attached, and the goodwill curdled into unease the moment recipients asked how the company knew so much. Relevance built on data the customer never knowingly gave does not register as relevance. It registers as surveillance.

Earning The Choice Again

The task is not another campaign or a deeper rewards tier. Derek Slager, Amperity’s co-founder and co-CEO, said: “Personalisation isn’t the goal. Relevance is.”

Customers will share data when they get value back, he argues, provided brands use it responsibly. The brands a customer keeps choosing in an AI market will be the ones whose customer context is trustworthy enough to be surfaced by the model and accurate enough to be relevant in the moment, every moment, because the model keeps re-asking the question. Knowing the brand was once enough to win the sale. Now the brand has to earn the choice each time the customer, or the machine acting for them, decides to ask.