Marketing

Own Your Customer Data: Building a Restaurant CRM From First-Party Orders

A restaurant CRM from your own orders: marketplaces keep the customer, direct orders give you the data. RFM, segments (new, regular, lapsed, VIP), GDPR.

Ordering.ToolsSeptember 23, 20269 min read
Restaurant owner reviewing customer orders on a laptop at a dining table

You have served the same customer forty times this year. You know their face, maybe their usual order. And if every one of those forty orders came through a marketplace, you know almost nothing else — no name, no email, no phone, no way to reach them the next time you want to fill a slow Tuesday. The marketplace knows all of it. You paid a commission on every order and rented the relationship.

This is not a piece about email tactics or how to turn a first-timer into a regular — those are their own topic. This is about the layer underneath: the customer list itself. Who owns it, what sits inside it, and how to slice it so that the right offer reaches the right person instead of one loud message blasted at everyone. Get the infrastructure right and every campaign you ever run gets sharper. Get it wrong and you are marketing blind.

Why the Marketplace Leaves You With Nothing

When an order arrives through a delivery marketplace, the customer belongs to the marketplace, not to you. The name on the ticket is often a first name and a partial address. The email and phone are proxied or hidden. You cannot message that person, cannot see whether they ordered from three competitors last week, cannot tell a loyal weekly regular apart from a one-time tourist. On top of the commission — typically a 15–30% cut on each order — you hand over the single most valuable asset a restaurant builds over time: knowing who its customers actually are.

Direct ordering flips this. When a customer orders through your own branded page, the transaction is yours end to end. With their consent you keep an email, a phone number, an order history, and a running picture of how often they come back and how much they spend. That is a first-party customer list — data you collected directly, on your own channel, that no third party can revoke or price you out of. It is the difference between renting attention and owning a relationship you can return to for free, again and again.

The Data That Actually Matters: RFM

You do not need a data science team, and you do not need to collect everything. A tiny handful of facts about each customer tells you almost everything worth knowing. Marketers call the core framework RFM — Recency, Frequency, Monetary value — and it works because it is built from the order history you already generate. Track these:

  • •Recency — how many days since their last order. A customer who ordered yesterday behaves completely differently from one who vanished four months ago.
  • •Frequency — how many orders total, and how often. Once ever, or the same Friday every week? This one number separates a fan from a stranger.
  • •Monetary value — total spend and average order size. A €12-per-order customer and a €60-family-order customer deserve different offers.
  • •Order history — what they actually buy. Someone who only ever orders the vegan bowl should never get a steak promotion.
  • •Channel and context — pickup vs delivery vs dine-in, weekday vs weekend, lunch vs late night. Habits hide in the pattern.
  • •Consent status — whether they agreed to marketing email, SMS, or push, and when. This is data too, and the most important field of all.

None of this requires a survey. Every field above falls out of the orders themselves, the moment those orders run through a channel you control. The job is not to collect more — it is to look at what you already have.

Rule of thumb: if you can answer "when did they last order, how often do they order, and how much do they spend" for any given customer, you have enough to market intelligently. Everything past that is a bonus, not a prerequisite.

Segmenting: Why You Should Never Blast Everyone

A single message sent to your entire list is the most common marketing mistake a restaurant makes. The "20% off this weekend" you send to win back a lapsed customer is money burned when it lands in the inbox of a regular who was going to order at full price anyway. Segmentation is simply splitting your list into groups that deserve different treatment. Four buckets cover almost every restaurant.

New, Regular, Lapsed, VIP

New customers ordered once, recently — the goal is a gentle nudge to a second order before they forget you, not a hard discount. Regulars order steadily; they need recognition and the occasional reason to try something new, not a coupon that trains them to wait for one. Lapsed customers used to order and stopped — this is where a real win-back offer earns its keep, because the alternative is losing them entirely. VIPs are your top spenders and most frequent visitors — they deserve early access, a genuine thank-you, maybe a standing perk, and almost never a blunt discount.

From Segment to Campaign

Once the buckets exist, the campaigns write themselves. A lapsed segment gets a "we miss you, here is something to come back for" message. New customers get a warm second-order prompt. VIPs get the first look at the new seasonal menu. Each message is small, relevant, and cheap, and because it is targeted it converts far better than the same offer sprayed at the whole list. The Marketing Campaigns feature page shows how these segments plug straight into sends, and the Loyalty Program feature page covers how to reward the VIP tier without leaning on a permanent discount.

Segments Are Not Set-and-Forget

Customers move between buckets constantly. Today's new customer is next month's regular, and this quarter's VIP can quietly slide into lapsed if you stop paying attention. The value of a real CRM is that it re-sorts everyone automatically as new orders arrive, so the lapsed list you email is always genuinely lapsed — not a snapshot from a spreadsheet you exported and forgot in March.

Consent and GDPR, Done Without Fear

Owning customer data comes with a duty to handle it properly, and in the EU that means GDPR. The good news is the rules are practical, not scary. Collect a clear, opt-in consent for marketing at checkout — an unticked box the customer chooses to tick, not a pre-filled one. Store what that consent was and when it was given. Make unsubscribing genuinely one click. Only keep data you have a reason to keep, and be ready to delete a customer's record if they ask. Do these things and consent stops being a legal headache and becomes exactly what it should be: a list of people who actively want to hear from you, which is worth far more than a bigger list of people who do not.

A platform that captures consent cleanly at the point of order, records it, and honours unsubscribes for you takes the compliance weight off your plate entirely. See the Customer CRM feature page for how order history, segments, and consent live together in one place — the same place your ordering already runs, so the list builds itself with every order instead of being a separate chore. Own the channel, own the data, and the marketing you do next is aimed instead of hoped.

Frequently asked questions

Why should a restaurant own its customer data?

Because data you collect on your own ordering channel is yours to reach again for free, forever. When orders come through a marketplace, the marketplace keeps the customer name, email, and phone, so you cannot re-market to people you have already served. Owning first-party data turns every past customer into someone you can win back on purpose instead of by luck.

What is restaurant customer segmentation?

Segmentation is splitting your customer list into groups that behave differently and deserve different messages — typically new, regular, lapsed, and VIP. Instead of one offer blasted to everyone, each group gets a message that fits where they are. A win-back discount goes to lapsed customers, not to a regular who would have paid full price anyway.

What is RFM in a restaurant context?

RFM stands for Recency, Frequency, and Monetary value — how recently a customer ordered, how often they order, and how much they spend. These three facts fall out of your order history automatically and are usually enough to sort customers into meaningful segments without any surveys or extra data collection.

How do I collect customer data without breaking GDPR?

Collect a clear opt-in consent at checkout using an unticked box the customer chooses to tick, record what they consented to and when, and make unsubscribing one click. Only keep data you have a reason to keep and be ready to delete a customer record on request. Done this way, GDPR compliance is routine, not a barrier.

Is a spreadsheet enough for a restaurant CRM?

A spreadsheet is a static snapshot that is out of date the moment you export it, and it will not re-sort customers as they move between new, regular, lapsed, and VIP. A proper CRM tied to your ordering updates every segment automatically as new orders arrive, so the list you market to is always current rather than a copy you forgot to refresh.

Try Ordering.Tools for Free

Set up your digital menu in under 2 minutes. No credit card required.