A manager can identify which message generated a tracked visit when each message has a distinct link and the reporting system preserves that link’s campaign label. Identifying a completed order takes another step: the visit must be connected to an order confirmation event. A click alone does not establish that a guest ordered, paid, or visited the restaurant.
Before buying software, ask to see the path from message link to landing page to order report. The useful question is not simply whether a dashboard shows campaign numbers. It is whether a manager can explain what each number counts and find the orders behind it.
Map the link to the order confirmation
Start with one destination and give each message its own link. For example, an email and an SMS can both lead to the same online ordering page, but their links should carry different campaign labels. A label might identify the campaign, channel, and message version. Use a consistent naming pattern so the lunch email does not appear under several slightly different names.
The path has three separate checkpoints:
- Link: The guest follows a campaign-specific URL from a message.
- Landing page: A visit is recorded with the campaign label, subject to the measurement setup and the guest’s privacy choices.
- Confirmation: A completed order event is recorded, and the report connects it to the earlier visit if the system supports that connection.
Ask the vendor to define a visit and a completed order. Does a visit mean a page load, a session, or a unique person? Does an order count when it is submitted, accepted, or paid? Can a canceled or refunded order remain in the campaign report? Those definitions matter when the manager compares campaign reports with the order system.
Also ask where the label is stored as the guest moves from the landing page into ordering. A label visible in the first page URL is not proof that it survives checkout. If the order page opens on another domain or in another browser, ask how the connection works and what happens when it does not.
Your restaurant management software may be one place to review orders, but confirm whether campaign labels appear there, in a separate report, or in an export. Do not assume that two screens sharing an order total use the same attribution rules.
Use a worksheet to separate clicks, visits, and orders
The following figures are fictional inputs for a simulated campaign, not results from a restaurant or a claim about any product. Both messages promote the same ordering page. The manager wants to know which message generated tracked visits and which visits were connected to confirmed orders.
| Measure | Lunch email | Lunch SMS | What to check |
|---|---|---|---|
| Recipients | 80 | 60 | Count people sent each message; do not treat this as visits. |
| Tracked link clicks | 24 | 12 | Clicks are link actions, not necessarily distinct people. |
| Recorded landing visits | 18 | 10 | Check the visit definition and why some clicks did not become recorded visits. |
| Confirmed orders attributed | 3 | 2 | Find each confirmation and check the attribution rule. |
In this scenario, the report records 36 clicks, 28 landing visits, and five attributed confirmed orders across the two messages. The email accounts for 18 of the recorded visits and three of the attributed orders; the SMS accounts for 10 and two. These are counts of different events, not interchangeable measures. A person might click more than once, and the two recipient lists might overlap.
The gap between clicks and visits deserves a question, not an automatic conclusion. A click might not lead to a completed page load, or the click and visit tools might count differently. Ask for the timestamps, link labels, and event definitions needed to investigate the gap.
Next, select one attributed order in the report and trace it to an order confirmation. A useful report should let the manager distinguish that order from an abandoned cart or a page visit. If the restaurant POS system is the source used to check order status, ask how its records are matched to the campaign report and whether cancellations are reflected in both places.
For the worksheet to pass, the two labels should be distinguishable, the landing visits should appear under the intended labels, and the five simulated confirmations should be traceable under the stated attribution rule. It fails if an order is counted merely because someone clicked, or if a manager cannot tell what the reported order count represents.
Test missing tags, shared links, and privacy limits
Run these checks in a vendor demonstration or a safe test environment before using a live campaign. Use simulated events and test records. Do not send a guest message, charge a real card, or dispatch a courier as part of practice.
- Missing label: Open a simulated link without its campaign label. Pass if the visit appears as unassigned or under another clearly explained category. Fail if it is silently credited to the email or SMS without a stated rule.
- Shared link: Simulate one recipient forwarding the email link to another person. Pass if the report’s documentation makes clear that the visit can retain the email label even though a different person opened it. Fail if the dashboard presents that label as proof the original recipient visited.
- Privacy restriction: Simulate a visit where the measurement method cannot record the relevant event. Pass if the reporting behavior and resulting gap are explained. Fail if the tool implies every visit will always be attributed.
- Interrupted journey: Simulate a guest opening the link on one device and placing an order separately on another. Ask whether and how the system connects those events. Pass if an unconnected order remains unassigned or follows a documented rule; fail if an unsupported match is presented as certain.
Agree on the expected result before each check. Ask the vendor to show the raw event or report entry as well as the dashboard total. A plausible-looking chart is not enough if the team cannot see how an untagged visit or order was classified.
Contact information is a separate issue from permission to send marketing. Ask how the system handles audience selection, opt-outs, and access to contact data. For questions about the legal treatment of a particular campaign or tracking setup, consult a qualified adviser rather than relying on a dashboard label.
Read attributed orders without claiming campaign lift
An attributed order tells you that the reporting system connected an order to a message according to its rules. It does not prove the message caused that order. A guest might have planned to order already, seen another promotion, or received both messages. A forwarded link can bring in someone who never received the original campaign.
Ask which touch receives credit if a guest follows more than one link. For example, does the report use the first recorded campaign touch, the last, or another rule? Also ask how long after a click an order can still receive credit. Compare reports only after checking that those rules, event definitions, and reporting periods match.
A holdout comparison addresses a different question: whether a campaign produced additional orders compared with a similar group that did not receive it. Link attribution alone cannot answer that. If you later run such a comparison, plan the groups and measurement separately rather than treating the attributed order count as incremental sales.
For a buying decision, write down the minimum evidence your manager needs: distinct message labels, a clear visit definition, a traceable confirmation, an understandable treatment of unassigned activity, and an export or report the team can actually use. If a vendor cannot demonstrate one of those steps, decide whether a manual check is acceptable or whether the gap rules out the workflow.
Use that checklist to compare your requirements with Fleksa’s restaurant marketing automation, and confirm current availability and attribution behavior before choosing a workflow.
Frequently asked questions
Does a campaign link click mean a guest placed an order?
No. A click records a link action. An order requires a separate confirmation event, and the reporting system must connect that event to the campaign under a stated rule.
Can a shared campaign link identify the original recipient as the visitor?
Not reliably. A forwarded link may retain its campaign label when someone else opens it. The label can identify the link’s source without proving who visited.
Can attributed orders prove a campaign increased sales?
No. They show orders connected to a campaign by the reporting rules, not whether those orders would have happened anyway. A properly planned holdout comparison addresses the separate question of incremental impact.
