Why Disconnected POS, E-commerce, and Loyalty Quietly Kill Your Unit Economics

Walk into most multi-outlet F&B chains running 5–50 stores and you’ll find the same setup: a POS at the counter, a Shopify or WooCommerce storefront for online orders, and a loyalty programme stitched in via a third-party SaaS. Three systems. Three databases. Three customer records for the same person.

For a while, everything works. Each system does its job. The POS rings up sales. The store handles online orders. Loyalty enrolments climb. But the unit economics drift in a direction nobody can quite explain.

Three databases, one customer

The shape of the problem is rarely operational — it’s structural. A customer who buys at your café on Tuesday and orders online Saturday looks like two different customers to your stack. The POS sees a transaction tied to a card token; the storefront sees an email tied to a Shopify customer ID; the loyalty app sees a phone number tied to a points balance.

None of those records know about each other. So your reporting answers questions that can’t actually be true. Your “average customer frequency” is wrong. Your “share of wallet” is wrong. Your churn metric is whichever-system-detected-it-first, which is usually the loyalty app, which only sees a fraction of revenue.

What the silos cost in practice

Attribution gaps. A customer joins your loyalty programme online, then visits the store and pays in cash. The visit doesn’t tie back to their profile, and the next morning your retention dashboard shows them as inactive. You re-target them with an acquisition campaign — paying twice for a customer you already had.

Channel cannibalisation, invisible. When your aggregator-driven delivery channel takes 25–35% of every order, you’re not just paying commission. You’re losing the customer record. The aggregator owns the relationship. When that customer comes back next week, they come back to the aggregator first — and you pay the toll again. Over twelve months, the same customer can be billed against your margin twenty or thirty times before you even know who they are.

Loyalty programmes that look healthy but aren’t. Most enterprise loyalty dashboards report enrolment growth. Almost none report active rate, redemption rate, or programme-driven repeat rate. Without those three, you have a discount mechanism wearing a loyalty badge — visible cost, invisible return.

What “unified” actually means in production

The marketing version of omnichannel is “all your channels in one place.” That’s a UI. It doesn’t fix the structural problem.

The version that moves unit economics is a single canonical customer record that every channel writes to and reads from. POS transactions, online orders, app installs, loyalty events, support tickets — all stitched to one identity. When something happens in one channel, the whole stack knows.

From there the practical wins start showing up:

• A purchase in any channel updates the loyalty profile in seconds, not at the next nightly batch.
• A visit to the storefront fires a check against in-store purchase history before the upsell engine recommends anything.
• Churn detection lives at the customer level — across channels — not at the loyalty app’s narrow lens.
• Direct-to-customer marketing has somewhere to write back to. Aggregator commission stops being your only delivery option, because you have the customer’s actual contact.

A three-question diagnostic for your current stack

If you’re not sure where your stack sits, here’s what to ask. The answers will be uncomfortable.

1. What percentage of in-store transactions can you tie to a known customer profile? Below 40% means the POS is operating outside your customer system. The loyalty programme is probably reaching a smaller audience than the founder thinks.
2. If a customer made their last purchase via an aggregator 90 days ago, can your stack send them a direct-to-store nudge? If the answer involves exporting a CSV from somewhere, the answer is no.
3. What’s your active loyalty rate (not enrolment rate)? If it’s below 15% of enrolled, the programme is decorative. If it’s above 35%, somebody’s done good work and the unified ledger is probably more functional than the org gives it credit for.

What replacing the silos looks like

The brands that get this right don’t necessarily use one tool — they use a stack where one system is the customer system of record and everything else writes to it. Sometimes the POS is the centre. Sometimes the CDP is. Sometimes — when a chain operates beyond 10 outlets and across countries — it’s a purpose-built omnichannel platform that owns the canonical layer and integrates the rest.

The decision isn’t a tooling choice. It’s a question of where the customer identity lives. Answer that and the rest of the stack starts to organise itself around it.

If your team is working through this question, our retail omnichannel platform was built for chains in exactly this position — five outlets and growing, three systems and noticing. Tell us where you are and we’ll show you where the gaps usually sit.