To compare ordering provider uptime commitments, ask each provider to show you the written terms behind its percentage. Find out which parts of ordering are covered, how downtime is measured, what gets excluded, what remedy applies, and how your team will hear about an incident.
A strong headline number can hide a narrow definition. A commitment covering a website's home page, for example, may not tell you whether a guest can open the menu, complete checkout, or get an order through to the restaurant. Compare the full ordering path against the hours when your restaurant actually needs it.
Define what is covered and how it is measured
Start with the guest's journey. Someone opens your ordering page, chooses items, enters details, and submits an order. Your team then needs to receive that order in time to prepare it. Ask the provider to mark which steps fall inside its commitment and which depend on another company.
Be specific about channels and locations. Does the commitment apply to the ordering site, an embedded ordering page, or both? Does it cover every restaurant location in your agreement? If your restaurant website and ordering service are provided separately, ask where each provider's responsibility starts and ends.
Request a plain-language definition of an outage. A total failure is easy to recognize, but partial failures matter too. Guests might see the menu while checkout fails, or orders might be submitted but not reach the restaurant promptly. Ask whether those events count as downtime and how the provider detects them.
Then ask for the measurement period and calculation. Is availability measured monthly or over a longer period? What clock starts and stops the downtime count? Does the provider measure from its own systems, from a guest-facing check, or from another point? Ask for the calculation in writing, including how partial outages are handled.
Hypothetical example: A provider promises 99.9% monthly availability for a covered ordering service. In a 30-day month, there are 43,200 minutes. A 0.1% allowance equals 43.2 minutes of counted downtime. That is a calculation of the stated threshold, not a prediction that downtime will occur or a guarantee that all ordering problems count against it.
Ask the provider for the actual agreement and measurement method. Your manager can supply opening hours and the ordering steps staff rely on; the provider should identify the covered components and calculation. Put both sets of information on the same sheet before comparing offers.
Read exclusions and remedies together
Exclusions determine which interruptions are left out of the calculation. Ask for a complete list rather than accepting a summary such as “standard exclusions.” Look closely at scheduled maintenance, third-party payment services, internet outages, integrations, and events outside the provider's control. The agreement determines which exclusions actually apply.
For each exclusion, ask a practical question: if this component fails, can guests still place an order, and can staff receive it? A payment processor interruption may prevent prepaid checkout even if the ordering page is available. An in-store internet outage may keep staff from seeing orders even when the provider's service is working. Those are different causes, but both can interrupt service.
Ask how maintenance is announced and whether it can occur during your selling hours. A maintenance window that is convenient for the provider may coincide with your dinner rush. Find out whether maintenance time counts against availability and whether you can request a different window.
Next, read the remedy. A service credit and restored service are different things: a credit may reduce a future bill, but it does not recover an order that never arrived. Ask what triggers a credit, how it is calculated, whether there is a cap, and whether you must submit a claim. Note any deadline and the records required.
For a useful comparison, ask each provider to walk through the same scenario: guests cannot complete checkout for 45 minutes on a Friday evening. Which parts of that event would count, what evidence would be used, and what remedy could apply? The answer depends on the cause and the written terms. Have a qualified adviser review contract questions where needed.
Check incident updates and status history
Availability terms describe the measurement and remedy. Incident communication tells your team what to do while the problem is happening. Ask who is notified, through which channel, and whether the notice reaches a restaurant manager rather than only the account owner. Confirm whether alerts require an opt-in and who can change the contact list.
Ask what an update contains. A useful message identifies the affected service, the known impact, the time of the next update, and any action the restaurant should take. An initial message may not have a cause or restoration time; your team still needs to know when to expect more information.
Request a view of the provider's status history, if available. Check whether past entries distinguish investigation, mitigation, and resolution, and whether they identify affected components. Ask how long history remains accessible and whether the provider records incidents that affect only some customers or locations. A status page is one source of information, not proof that every guest-facing step is working.
Run a simulated communication check with your staff. Assume checkout is unavailable at 6:15 p.m. The manager reviews the provider's published status channel or a sample notice, tells the counter team what is known, and records when the next update is expected. Use a rehearsal message clearly marked as practice. Do not send a real guest message, take a real card payment, or dispatch a courier.
This exercise checks your team's response and contact list. It does not measure provider uptime or prove that an actual alert will arrive. Separately, ask the provider to demonstrate its notification setup or explain how an account receives incident updates.
Compare commitments against restaurant hours
A monthly percentage treats each counted minute alike. Your restaurant does not. An interruption during a quiet closed period has a different operational effect from one during a busy meal service. Ask whether the provider offers any terms specific to operating hours; do not assume the headline commitment does.
Map your opening hours, order acceptance hours, and busiest periods. Include late-night service, holidays, and any location with a different schedule. Then ask when maintenance is normally scheduled and how far ahead the provider communicates it. Compare those windows with your actual order-taking hours, not just the hours printed on the front door.
Also agree on a fallback before an incident. Decide who can pause an ordering channel, update staff, and check for orders that may have arrived late. Ask the provider what controls and order records are available during an interruption. If you plan to take orders by phone, make sure staff know how to handle them without assuming the online system will capture those orders.
Use the following sheet for each provider. The entries describe what to collect and how to judge it; replace them with the provider's written answer and your restaurant's operating needs. Keep the agreement, maintenance terms, and incident contact details with the sheet.
| Comparison item | Ask the provider | Restaurant decision |
|---|---|---|
| Covered service | Which ordering steps, channels, and locations are included? | Check that guest checkout and order receipt are addressed. |
| Measurement | What period, outage definition, and calculation apply? | Compare the same period and partial-failure treatment. |
| Exclusions | What maintenance and third-party failures are excluded? | Identify disruptions that could still stop service. |
| Remedy | What credit applies, and how and when must it be claimed? | Record the claim owner, deadline, and any cap. |
| Incident updates | Who receives alerts, how often are updates sent, and is history available? | Confirm the manager can access updates during service. |
| Hours and fallback | When is maintenance scheduled, and what controls remain available? | Compare with busy hours and assign staff actions. |
Do not score providers solely by adding up checkmarks. A narrower commitment may still be acceptable if you understand the risk and have a workable fallback. A higher percentage may be less useful if checkout failures are excluded or the claim process is difficult to manage. Record the trade-off and the person responsible for each follow-up.
Put the answers into your buying decision
Ask for the same documents from every provider: the service commitment, the relevant contract terms, the maintenance policy, and instructions for incident notifications and claims. Where an answer differs from a sales conversation, request clarification in writing before making your comparison.
Keep service availability separate from support coverage. Knowing when someone can answer a support request helps your response, but it does not define what availability is measured or what happens when the commitment is missed. Likewise, an availability commitment does not by itself tell you how quickly a person will respond.
If you are also comparing ordering features and operating costs, use the online ordering buying guide alongside this sheet. It can help you keep the broader purchase decision separate from the uptime terms.
Next step: Complete the sheet for your current setup, then ask the providers you are considering to answer the same questions in writing. Explore Fleksa's restaurant online ordering and ask about verified US availability and the applicable service terms for your location.
Frequently asked questions
Is a higher uptime percentage always better?
Not necessarily. Compare the covered services, measurement period, exclusions, and remedy as well as the percentage. A commitment is more useful when it addresses the ordering steps your restaurant depends on.
Does an uptime commitment cover payment failures?
It depends on the agreement and the cause of the failure. Ask whether checkout and third-party payment services are covered, and how a payment-related interruption affects the downtime calculation.
What should I save during an ordering incident?
Keep the times you first noticed and last observed the problem, the affected ordering steps, and any provider notices or incident references. Follow the agreement's claim instructions and avoid recording unnecessary guest payment information.
