Guestit
English
ContactLog inIntroduction

Technology

8

min read

Rolling out new hospitality software without halting service

Roll out new hospitality software without service grinding to a halt: test three tricky scenarios, practise per role and start with a one-site pilot.

Illustration: two staff members in aprons stand at a POS screen on the bar of a restaurant; one of them points at something on the screen.

Do this today

Today, ask one colleague from the floor staff, one from the bar and one from the kitchen which two situations are the most difficult during a busy service, and write down the six answers.

The new POS passed the demo. On Friday at 19:15, table twelve wants to cancel one main course while the kitchen is already working on the order. Do your floor staff know which button to use? And does the kitchen see which line has changed? Those are the questions that decide whether your venue, or your next location, is ready to switch.

A demo shows what software can do. A rollout shows whether your team can work with it when it gets busy. In between lies work for you and your team: choosing what to test, cleaning up your item file, practising per role and agreeing who decides when things go wrong. After this article you will have a test sheet for your own scenarios, a worked example for the practice time, an outline for a pilot and one action for today. Still choosing? Then first read Choosing a restaurant POS? 14 key do's and don'ts; this article starts where the choice has been made.

Start with the order that goes wrong most often

Ask floor staff, bar and kitchen each for two situations that are difficult during a busy shift. That gives you six situations, taken from the daily practice of the people who will be working with the system. Then choose three processes that demonstrably need to work better. Think of adding a round via the handheld, passing a change to a single kitchen station or splitting a bill.

Why three and not everything? Because you need to be able to repeat a test. Three scenarios can be run again and compared at every test moment; test everything at once and it becomes a guided tour of the system.

Give every test an expected result, and write it down before you test. With a cancelled dish, for example, you check both the status in the kitchen and the amount on the bill. Without an expected result, any outcome can be explained away afterwards.

Check your item file before you migrate it

Check which dishes are still being sold, whether variants are correct and where each item belongs: which product group, which station, which price. A duplicate product line that barely stands out during the demo can lead in practice to two different prices or a ticket at the wrong station.

Do not throw away anything you are required to keep. According to the Belastingdienst (the Dutch Tax Administration), a retention period of seven years applies to basic records, including purchase and sales records (reference date 21 November 2026). So before the switch, ask how you can keep consulting the data from your old system, and let your accountant determine which data falls under this in your case.

Checklist: the test sheet per scenario

For each of the three scenarios, make one sheet with the same six lines. Below it is filled in for the situation from the introduction; replace the answers with those of your own venue.

  • Situation: a guest cancels one main course while the kitchen is already working on the order

  • Who carries it out: a member of the floor staff, without help from the supplier or the owner

  • What the kitchen must see: which line has changed, at the station that prepares the dish

  • What the bill must show: the amount without the cancelled dish, or with it after all, if the kitchen had already started and your agreement is that it then stays on

  • Who decides in case of doubt: the shift leader of that service, named

  • Signed off: date, test moment and the initials of the person who saw it work

The fourth line calls for an agreement: what happens to a dish that is already in the pan? That is not a software question but your choice.

Practise per role and appoint one point of contact

Have one staff member place an order, a colleague at the bar or in the kitchen process the ticket and a shift leader carry out the correction. Then switch roles where that makes sense: anyone who has stood on the kitchen side once understands why a note has to be on the right line. Record the three most difficult actions on a short card next to the POS. Give the weekend crew time to practise too. A training on Tuesday only works for the people who are there on Tuesday.

Worked example with made-up figures; fill in your own team and your own wage costs. Suppose your team consists of 16 staff members, 6 of whom work only at weekends, and that you plan 45 minutes of practice time per person.

  • Reach of the Tuesday training: 16 minus 6 is 10 staff members; 10 divided by 16 is 62.5% of the team

  • Who you miss: 6 divided by 16 is 37.5% of the team, precisely the crew working the busiest evenings

  • Practice time Tuesday: 10 times 45 minutes is 450 minutes, so 7.5 hours

  • Practice time Saturday before opening: 6 times 45 minutes is 270 minutes, so 4.5 hours

  • Total: 7.5 hours plus 4.5 hours is 12 hours

  • Wage costs of the practice time: 12 hours times €22 per hour is €264 (a made-up hourly wage cost)

Before go-live, agree who decides in an incident, where everyone reports it and which temporary way of working the team can use. One point of contact per service is meant to prevent three people from trying something different at the same time. Also record how you will check orders and payments afterwards if you have temporarily worked manually: who enters the tickets, and who compares the daily total with the card payments?

What a pilot can look like

A made-up example, not a customer story. A fictional chain with six locations chooses one venue with both table service and takeaway, so that both ways of working are part of the test. The team first practises outside opening hours, then runs a quiet lunch and then evaluates a busy dinner. Three scenarios at three moments make nine tests, each with its own test sheet.

During lunch, an order appears at the wrong station. The shift leader does not just fix that one ticket, but works out which items use the same configuration. The team corrects the configuration and repeats the same test before a second location goes live.

Define in advance when the pilot has succeeded. For example, have the shift leader sign off that all agreed scenarios work (in this example nine out of nine tests), that open incidents have an owner and that revenue and payment methods reconcile at closing. The next venue then gets a playbook that has been tested in real service. If you have one venue, the same works on a small scale: a quiet service first, only then the Friday evening.

Where it goes wrong

  • Testing what goes well: an ordinary order for an ordinary table says little; the test lies in the change, the cancellation and the split bill

  • The owner tests it personally: you know the system after three demos; the staff member filling in on Saturday does not

  • Go-live on the busiest day: a new way of working and a full venue at the same time leave no room to sort out a mistake calmly

  • No fallback option: anyone who has not agreed how the team works without the system for a while will be figuring it out during the outage

Do this today

Today, ask one colleague from the floor staff, one from the bar and one from the kitchen which two situations are the most difficult during a busy service, and write down the six answers. You can do this between two services, in a quarter of an hour. Then choose the three situations that come up most often and, for each situation, write down in one sentence what the kitchen must see and what the bill must show. No bar? Then ask two colleagues from the floor staff. You do not need new software for this.

Where technology helps

Guestit sells hospitality software itself, so compare your test sheets with our offering too. Guestit works with one central database for POS, ordering channels and kitchen. In the Guestit POS, a change to a dish that has already been sent arrives in the kitchen marked as a change. On the kitchen displays, the POS may lower a quantity or remove a dish as long as the kitchen has not started on it; otherwise the change is refused and the ticket shows what was requested. The kitchen displays send each dish to a station, and a routing change does not affect dishes that have already been started. The POS has a help function, Show me, that points out every step on the real screen. Whether that fits your scenarios, you will only see once you run them; bring your test sheets to an introduction.

How we put this together

This article is an elaboration of a text by Guestit itself and is our own reasoning, not research. The chain with six locations, the team of 16 staff members, the 45 minutes and the hourly rate of €22 are made up; we checked the sums with a calculation script. We do not cite a standard for how long a rollout takes or how much practice time is needed, because we found no independent source for that. We opened and read the Belastingdienst page on retention periods on 21 November 2026; that page does not say which POS data falls under basic records in your case. The occasion for the topic was a blog post by software supplier Nory, What I've learned from 100+ restaurant tech rollouts (published 25 February 2025, updated on 15 September 2026 according to the page data; read on 21 November 2026). Our article shares four general ideas with it: start at one location, clean up your data during the switch, train more than once and involve the managers on the floor. The pilot example also follows that first idea: one location first, fix the mistake, then expand. We did not take over sentences, headings, figures, customer examples or the description of their own rollout process. More articles can be found in the Guestit Academy. The image for this article is an illustration (AI-generated), not a photo of a real venue and not a screenshot of Guestit.

Sources

  1. Belastingdienst, Hoelang moet u gegevens bewaren? (last modified 24 June 2026 according to the page data; accessed 21 November 2026)

Want to see this work in your business?

Book an introduction and we’ll look together at which of these ideas fit your business. If Guestit isn’t a fit, we’ll say so.

Book an introduction
See the solutions

Shall we give you a call?

Guestit

The operating system for hospitality. Connects every step of the guest journey with the processes behind it.

© 2026 Guestit