Customer and food
Customer intelligence
Changing customer needs, behaviours, and commercial implications.
This week’s key read
Confidence: highGK Software describes real-time basket context as a layer beyond loyalty integration
GK Software describes a retail model in which loyalty identifies members and applies rewards, while customer-data, live basket and availability signals can support a next-best action before payment. This supplier-described model has not been independently validated for catering.
Why it matters
For caterers with consented, digitally identified transactions, the live ordering moment may be an opportunity to test more relevant menu recommendations.
Recommended action
Run a four-week, consented pilot at one digitally identified site: show one explainable complementary recommendation based on current basket, prior purchases and available menu items. Compare attach rate, average spend, repeat purchase and recommendation acceptance against a matched control period; do not assume improvement before results are available.
More from this section
Validated finding
Confidence: mediumRetail survey indicates preference for transparent, first-party personalisation
Attentive's February 2026 survey of 1,050 opted-in shoppers reports acceptance for recommendations based on past purchases, saved preferences, cart contents and reorder patterns, and reported discomfort when brands appeared to infer sensitive information or use data from elsewhere without clarity. Its relevance to contract catering remains unproven; it does not support inferring health or weight goals.
Why it matters
A transparent preference model may offer a lower-risk way to test relevance while handling food, dietary and allergen-adjacent data responsibly.
Recommended action
Test a voluntary three-question preference capture flow: dietary style, menu goals and communication preference. Display a plain-language benefit and separate consent for marketing from use of preferences in on-site menu filtering. Measure completion, opt-out, recommendation engagement, repeat purchase and complaints; do not infer allergies, medical conditions or weight goals.
