Delivery Platforms Bring Orders, but Who Controls the Customer Relationship?
Receiving orders through GrabFood, foodpanda, Keeta and other platforms does not mean a restaurant has built a customer relationship. This article looks at how restaurants can use AI-assisted follow-ups and other touchpoints to build direct relationships while respecting platform rules and privacy boundaries.
A note before we begin: I previously wrote about how orders from three delivery platforms can enter one backend. That article focused on order integration: reducing the number of tablets in a restaurant, avoiding duplicate data entry and preventing missed orders.
Once those orders enter the same backend, however, a harder question remains. When a restaurant receives an order, has it also started a customer relationship?
What does a restaurant keep after a platform order is completed?
A restaurant receives a large order through GrabFood. The food is prepared and the order is completed.
The next time that customer wants a similar meal, they will probably open the platform again and choose a restaurant from the available options.
For the restaurant, the transaction is complete. Whether a customer relationship has begun is less certain.
The owner may not know whether this was the customer’s first order, what they bought before or whether they will return. The restaurant fulfils the order and the platform completes the transaction, but the platform still controls the main route to the next interaction.
That does not mean the platform has stolen the restaurant’s customer.
Platforms provide traffic, payments, delivery and after-sales support, and they bear the costs of providing those services. The restaurant pays commission or other fees in exchange for access to an established transaction channel.
In a platform transaction, however, the restaurant usually receives the information needed to fulfil the order. That information does not necessarily include a reusable customer identity or permission to contact the customer in the future. The difference appears in three areas:
- Identity: Seeing an order does not mean the restaurant has created a member profile confirmed by the customer.
- Data: The restaurant can see the transaction but may not be able to recognise behaviour across orders, outlets and channels.
- Re-engagement: When the customer next opens the platform, the restaurant must compete for visibility and selection again, unless the customer has chosen to join one of the restaurant’s own channels.
A platform can create an order opportunity without creating a customer relationship that the restaurant can continue to manage.
If the platform changes its commission, ranking or visibility rules, the restaurant’s acquisition cost and order sources may also change. This is why restaurants need a second customer entry point outside the platform instead of placing the entire relationship in one channel.
Why is this a product design problem?
Some might ask whether loyalty programmes, CRM and promotions are simply part of digital marketing.
In practice, the problem goes further.
At which touchpoint does the restaurant obtain consent? Can identities from different channels be linked? Which data from platform orders can be retained? How are loyalty rewards redeemed correctly at the POS? After a complaint is resolved, can the system tell whether the customer returned?
A marketing campaign cannot answer these questions. They involve product flows, data models, system integrations and permission boundaries.
Bringing delivery orders into one backend improves fulfilment. CRM begins when individual transactions can gradually become a relationship that the customer has agreed to and the restaurant can continue to serve.
Building customer relationships is therefore a set of product decisions, not a marketing slogan.
Three stages of the customer relationship
Restaurants may use the same delivery platforms while having very different relationships with their customers.
Stage 1: The relationship exists only on the platform
The restaurant receives and fulfils the order, but the platform still initiates the next interaction.
The system may show revenue by channel and the contents of each order, yet it may not be able to distinguish new customers from returning customers or identify repeat purchases across channels. When the customer next opens the platform, the restaurant must compete for attention again.
Stage 2: The customer voluntarily joins the loyalty programme
The customer registers, subscribes or links their identity through one of the restaurant’s own touchpoints. This might be a dine-in loyalty programme, direct online ordering, a reservation, click and collect, or an electronic receipt.
The customer should understand that they are forming a direct relationship with the restaurant and agree to the stated uses of their data.
Stage 3: The restaurant builds an omnichannel relationship
With consent, dine-in, collection, reservations, loyalty and after-sales service can gradually be connected. The restaurant begins to understand where and how the customer buys, what problems they have encountered and whether they return.
This does not mean every piece of data must be centralised. Nor should the system infer that two records belong to the same person without enough evidence. If an identity cannot be confirmed, the records should remain separate.
Five capabilities restaurant SaaS needs to provide
Moving from a platform transaction to a direct customer relationship requires at least five capabilities.
Step 1: Preserve the channel source
Every order should be labelled as dine-in, delivery platform, direct online ordering, reservation or collection.
Without that information, the restaurant sees only total revenue. It cannot tell whether a channel brings new customers, repeat purchases or short-term orders driven by promotions.
Step 2: Do not force customer identities together
Data left by the same person across different channels does not automatically prove that the records belong to one customer.
Phone numbers, email addresses, devices and membership details may be incomplete or restricted by platform rules. The safer approach is to let the customer sign in, link a membership or confirm their identity. The system should not merge records simply because two orders look similar.
Step 3: Build consent into the product flow
When a customer joins a loyalty programme, subscribes to messages or provides contact details, the system should explain who is collecting the data, how it will be used, whether it will be used for marketing, and how the customer can withdraw or change their consent.
Consent is part of the product flow, not a checkbox hidden at the bottom of a page.
Step 4: Connect CRM with restaurant operations
The customer relationship cannot remain inside the marketing team.
Complaints need to enter customer service and outlet workflows. Loyalty rewards must be redeemed correctly at the POS. Changes in repeat purchases should be compared across outlets, products and channels.
This follows the same principle as my previous article on turning online reviews into operational data. When a customer leaves a bad review, the system should record more than whether customer service replied. It should also track whether the underlying problem was resolved.
Step 5: Measure repeat purchases, not only reach
Membership numbers, message volumes and coupon redemptions are easy to increase, but they do not necessarily show that customer relationships are improving.
Restaurants should also look at whether customers buy again, which channel brings them back, whether they return after a promotion ends, and whether they continue to visit after a complaint has been handled.
The goal is not to send more messages. It is to keep serving the same customers without having to buy access to them again every time.
What can AI-assisted follow-up do?
Compared with sending another promotion, an after-sales follow-up is a more natural point of contact.
There is an important condition. The platform must provide an appropriate after-sales messaging channel, or the customer must have voluntarily provided contact details through the restaurant’s own channel and agreed to the relevant use.
With that condition met, AI can help by:
- sending routine follow-ups, inviting reviews and collecting feedback;
- organising replies and identifying the type of issue;
- spotting negative signals such as long waits, missing items or problems with taste;
- passing cases involving service recovery, refunds or liability to a person.
AI-assisted follow-up is one entry point, but it is not the only one. Restaurants can design several low-cost touchpoints within their own service flows:
- A QR code on a paper or electronic receipt: Customers can choose to join the loyalty programme, view their order or opt in to follow-up services.
- An after-sales and review channel: Every customer can provide feedback. Negative issues enter the outlet’s resolution process, while customers decide for themselves whether to share a positive review publicly.
- Email or SMS after consent: These channels can support order notifications, after-sales follow-up or loyalty messages that the customer has chosen to receive.
- A card in the delivery bag or at the table: Where platform rules allow it, the card can introduce the restaurant’s loyalty, collection, reservation or after-sales channels.
These touchpoints are not designed to force customers away from a platform. They show customers that they can also choose to maintain direct contact with the restaurant.
A restaurant does not need many touchpoints. Each one should be voluntary and explain clearly how customer data will be used.
AI reduces the cost of processing large volumes of replies. It does not obtain consent for the restaurant and should not promise compensation on its own.
If a platform provides contact information only for fulfilment, the restaurant cannot pass it to its own AI customer service system for off-platform marketing. Whether a touchpoint can be used depends on the platform’s terms, the customer’s consent and local privacy requirements.
Customer data is not there for the restaurant to take
Discussion of customer relationships can easily go too far. Completing a transaction does not mean the restaurant owns the customer’s data.
Platforms disclose different information under different terms, and privacy and marketing requirements vary by market. Data provided for fulfilment cannot automatically be used for another purpose.
Several boundaries should remain clear:
- Data provided by a platform for fulfilment cannot automatically be used for independent marketing.
- Consent given in one channel does not necessarily apply to another.
- Analysing operational problems does not create permission to identify and track every customer.
- SaaS can help manage consent and permissions, but it cannot decide which uses are lawful for the restaurant.
Building a customer relationship does not mean copying platform data into a private database. It means that a customer uses one of the restaurant’s own service touchpoints and gives clear consent to future interaction.
Platforms and direct channels are not an either-or choice
Restaurants should not expect to leave delivery platforms entirely.
Platforms are useful for attracting new customers and capturing immediate demand. Direct channels serve people who already know the brand and are willing to interact with it. SaaS should bring the operating results from these channels into one backend, then connect loyalty and service relationships where customers have given consent.
Restaurants do not need to push every platform customer into their own app, and they should not redirect customers in ways that violate platform rules. The aim is to create a second entry point and reduce dependence on a single channel.
Depending on a platform is not inherently a problem. Having no other customer entry point leaves the restaurant with fewer options. Deciding which customers deserve long-term investment is a separate question, discussed in how restaurant SaaS companies should choose customers abroad.
Conclusion: will the customer choose to speak with the restaurant again?
The first stage of delivery integration is bringing orders from different platforms into one backend.
The second is helping the restaurant understand what each channel contributes in revenue, cost, customer response and repeat purchases.
Only then does the customer relationship begin.
Receiving a platform order does not allow a restaurant to claim ownership of that customer. Better service and direct touchpoints can, however, give the customer a reason to return directly.
A customer relationship is not a phone number. It exists when the customer is willing to speak with the restaurant again and has agreed to do so.
After the order is complete, restaurant SaaS still needs to help the restaurant answer one question:
Did we gain another transaction, or another opportunity to build a customer relationship?
🔧 Admin
Comments · 0 comments
Please use your name and stay on topic; abusive or spammy content will be removed.
Loading…