Who Is Your Customer?

Twenty percent of your revenue comes from aggregators. Zero percent of your customer data does. That asymmetry is a strategic problem — and it is solvable.

The aggregator bargain

When a customer orders your biryani through Talabat, Talabat books the transaction. They collect the payment. They own the customer. You get an order ticket, a commission invoice, and no way to contact the person again. Nothing in that arrangement prevents Talabat — on a competing restaurant's page — from recommending an alternative to your customer next Tuesday.

This is the trade. It works until you want to do something with the relationship: resolve a complaint directly, launch a loyalty programme, send a voucher for a bad experience, invite someone to a new branch opening. At that point, you discover that the customer is not yours.

The standard workaround does not work

Most operators try to solve this by putting a QR code or a discount code in the delivery bag — "scan this to save 10%." The uptake is almost always below 5%. By the time the bag arrives, the customer is eating; the relationship they want is with the food, not with a CRM onboarding flow.

“Three complaints this week. We had no way to call the customer back.”

— Brand owner, restaurant group — Qatar

What aggregators actually share

The answer varies by aggregator and by market. Some platforms provide customer phone numbers and addresses at order time as part of their operator API. Some obfuscate phone numbers behind a masked forwarding layer. Some share customer identity only after the order is delivered. Some never share it at all. A serious integration has to be granular about this — it is not a one-size-fits-all field.

What HornbillOS does

• For every aggregator we integrate with, we surface customer information to the extent the aggregator's terms allow. Where identity is available, it flows into your CRM attached to the order.
• Repeat-order attribution stitches together direct, walk-in, and aggregator orders for the same customer — so you can see lifetime value across channels, not per channel.
• Per-region privacy posture is explicit. Operators can see exactly what is available on which platform and configure visibility accordingly.
• The CRM is not a bolt-on. It is the same customer table that drives your loyalty programme, marketing campaigns, and support workflows.

What a reclaimed customer enables

• Dispute resolution — a customer complains. You call them back within the hour. Public reviews turn private; 1-star reviews turn into retained customers.
• Direct remarketing — a high-value customer has not ordered in 60 days. A single outreach (WhatsApp, SMS, email) brings them back without sharing the revenue with anyone.
• Loyalty enrolment — tier progression, stamp cards, and points wallets work across aggregator orders, not just direct ones.
• Multi-channel attribution — stop asking "was this customer from Keeta or from the website?" and start asking "what is this customer worth across every channel they use?"

Two caveats, honestly stated

First: the data that can be surfaced depends on each aggregator's API and terms of service. Our job is to make the most of what is permitted — never more. The capability matrix on the Aggregator Control page is transparent about full vs partial coverage.

Second: customer data is a responsibility before it is an asset. Every piece of information that flows into your CRM comes with expectations around consent, retention, and use. HornbillOS exposes controls for this — what gets stored, for how long, who on your team can see it — because regulators and customers both expect it.

The strategic question

If 20% of your revenue comes from a channel that owns the customer, you are not running a restaurant — you are running a kitchen for someone else's platform. The shift starts by acknowledging that aggregators are distribution, not the relationship. The relationship is the food, the experience, and the brand — and those belong to you. Reclaiming the customer is how you make sure the business belongs to you too.

That reclaiming is not an aggregator integration. It is an operating system decision. And it is the reason HornbillOS was built as a whole OS, not just a sync engine.