Skip to main content

Migration and launch

Switching point of sale systems, including the parts that do not come with you.

A point-of-sale migration is not a data transfer. It is a rebuild with a deadline, run while the business stays open, and the risk is concentrated into about four hours on one morning.

This page is the honest version: what moves, what has to be built again, what is simply lost, and how the day itself is run so that nobody is learning the new screen in front of a queue.

Tell us what is happening on your floor →

The short answer

What happens when you switch POS systems?

Your menu, modifiers, prices, categories, staff and customer contacts get rebuilt in the new system, usually by hand from a clean source file. Gift card balances, loyalty points, stored payment cards and sales history generally do not transfer and need a separate plan. Then your team runs a test-order script, our team reviews the build, and you go live early in the week.


What moves

Most of the visible things come across, though almost never by a single button. Items, prices, categories, modifier structure and staff records are rebuilt from a clean source file, which we would rather produce ourselves than trust an export from a system that is about to be switched off.

Customer contact data usually moves, if it is exportable and if it was collected in a way that permits it. That is the asset most owners undervalue and the one worth being most careful with. It is also the thing a previous vendor is most likely to be sitting on.

The rebuild is a good moment to not carry rubbish across. Every migration we run deletes items nobody has sold in a year, merges duplicates, and rewrites modifier groups that grew by accretion. Carrying a bad structure into a new system is the most expensive possible way to change platforms.

A migration is the only free chance you get to throw away eight years of mistakes.

What does not

This is the section that gets skipped by everyone selling you a new platform, and it is where migrations go wrong.

  • Gift card balances. They almost never transfer automatically. Toast, for instance, only lets you adjust one gift card or one rewards account at a time in the back office, and needs specific permissions to do it; for a real stack of outstanding cards, the documented route is sending Customer Care a spreadsheet of card numbers and balances. That is a queue with a lead time, not a checkbox. Start it two weeks out.
  • Loyalty points. Members can often be bulk imported from a CSV through the new platform support team, and points sometimes come with them. Assume it needs a ticket, a file and a fortnight.
  • Stored cards and saved tabs. Card tokens belong to the old processor. They do not move. Anyone with a card on file will have to present it again, and house accounts need reconciling and reopening by hand.
  • Sales history. Your reports stay in the old system. Export everything you might want, including item-level sales, labor and tax reports, while the account is still open and before anybody cancels it to save a month of fees.
  • Integrations. Accounting, scheduling, delivery marketplaces, reservations, email. Every one has to be reconnected and re-authorized on the new platform, and each is its own small piece of work with its own login nobody can find.
  • Printer and screen configuration. Routing rules are platform-specific. Nothing about how your old system decided where a ticket printed comes with you.
  • Muscle memory. The unpriced one. Your fastest server is about to be your slowest for two days, and staffing the launch week as though that is not true is the most common planning mistake.

The test run

Before it is real, our team reviews the configured system against the source and a launch checklist, and your team runs a test-order script from the quick-start guide with the people who will use it. Not a demo. A service.

The script rings a full check with modifiers, fires it to the kitchen, voids a line, comps an item, splits a bill four ways, runs a partial refund, applies a discount, adds an automatic gratuity to a large party, closes a drawer and reads the report at the end. Then you look at the ticket that came out of the printer and ask the kitchen whether they would have made the right thing from it.

Almost every test run finds something. A modifier that does not print, a permission that stops a shift lead doing a refund, a tax that is applied to a takeaway order that should not be. Finding those with nobody waiting is the entire value of the exercise.

The day itself

Early in the week, into a slow daypart, never heading into a Friday night. A cutover heading into a weekend is how a good build gets a bad reputation, so it is a rule rather than a preference.

The old system stays available and reconciled through the first close, so a bad hour has somewhere to fall back to. Anything unexpected goes to the support team in your portal, which works from the setup record.

After launch, the small wrong things, the button in an awkward place, the modifier nobody thought about, go to the support team against the setup record. Automated follow-ups at day seven, day thirty and day ninety help you check progress; reply for support when you need it.

Which route you are on

The mechanics change depending on where you are coming from and where you are going. These are written up individually, with the specific traps of each pairing.

  • Clover to Square — usually a shop or a counter service leaving hardware-locked terminals behind.
  • Toast to Square — the one nobody writes about, and the move we run most often now: published pricing, no term, and a Restaurants plan that covers the floor.
  • Legacy POS to a modern system — a server in the back office, a support contract nobody has read, and an export format from another decade.

Should you move at all

Often the answer is no, and we would rather say it here than three weeks into a project. The platform is rarely the problem. The problem is usually the configuration: the modifier tree matching a menu from two summers ago, the ordering link pointing at a marketplace, routing rules nobody has touched since the install.

A migration is a bigger invoice for us than a configuration fix. We talk people out of migrations regularly anyway, because the alternative is charging somebody a lot of money to rebuild the same mistakes on a new screen.

The reasons that do justify a move are specific: the system cannot do something you genuinely need, like coursing or shared multi-location menus; the hardware is failing and locked to the vendor; the cost structure has changed materially against you; or support has stopped answering. If you are not sure which category you are in, a health check costs a fraction of a migration and answers the question.

Be honest with me

Do not migrate, or do not migrate with us, if:

  • You are unhappy with your current system but cannot name the specific thing it will not do. That is a configuration problem wearing a migration costume, and it will follow you onto the new platform.
  • You are in your busiest six weeks of the year. There is no safe cutover inside a holiday season, and any vendor who tells you there is has not run one.
  • You have thousands of outstanding gift cards and no card-number export. The balances are a manual reconstruction, and you should hear that cost before you commit, not after.
  • You want us to run a migration onto a platform we do not configure. We do Square, Toast, Shopify and Clover properly, and we will refer you rather than learn a fifth system on your invoice.
  • Your reason for moving is that a sales rep offered you a hardware deal. Work out the total picture over three years first, and we will help you do the math either way.

Questions we get

How long does a POS migration take?

Plan on a few weeks, most of which is waiting rather than working. The configuration build is days. The calendar is set by hardware shipping, gift card and loyalty imports queued with a support team, integrations being re-authorized, and finding a quiet morning early in a week that suits your kitchen.

Will we lose sales during the switch?

The goal is one slow shift that feels clumsy, not a day of lost revenue. That is what the test run and the early-week launch are for. Staff the first two services heavier than you think you need to, and expect your fastest people to be slow. Anyone promising you no disruption at all has not done many of these.

Can you migrate our customer list and email history?

The customer list, yes, where it is exportable and was collected in a way that allows it. Campaign history and engagement data usually stay behind with the old platform. Export what you have before anything is canceled, and treat the list as the asset it is. It belongs to you, and it should be in your account under your billing regardless of who does the work.

What about our gift cards that are already out there?

They are handled as a deliberate piece of work with a lead time. Depending on the platforms involved that means a bulk balance load through the new provider support team from a spreadsheet, a manual reconstruction card by card, or an honest policy decision to honor old cards manually at the till for a fixed period. Decide which one before launch, not during it.

Do you handle the hardware?

We configure, check and document the launch. We do not run cable, mount terminals or set up switches and access points. For most independents the platform ships hardware that works out of the box, and where a real network job is needed we will tell you to hire a low-voltage contractor and we will coordinate with them.

What if it goes badly?

The old system stays reconciled through the first close so there is somewhere to fall back to, and the support team in your portal works from the setup record through the first week, which is when the first week finds things. Revenue-blocking issues come first.


Tell us what the current system will not do.

Be specific about the thing that is actually blocking you, and we will tell you whether it is a migration, a configuration fix, or a setting somebody switched on by accident. The last two are cheaper for you and smaller for us, and they are the answer more often than not.

Platform pricing and features change; where a platform does not publish a number we say so rather than guessing.

Shop & Table connects with supported commerce platforms. Square, Toast, Clover and Shopify are trademarks of their respective owners.

Your preferences

Privacy choices

Choose which optional services can run on this website. Everything optional stays off until you choose. You can change your settings at any time.

Advertising tags are not configured on this website. These settings apply to this browser on this website; platform and ordering accounts have their own choices.

Make yourself comfortable

Accessibility

Choose what makes this page comfortable. Preferences stay on this device.

Navigate this page

Keyboard shortcuts

On a Mac, Alt is the Option key. Shortcuts do not run while you’re typing in a field, or with Control or Command held.

Alt + Shift + A
Open accessibility tools
Alt + Shift + S
Navigate by page structure
Alt + Shift + ?
Show keyboard shortcuts
Alt + Shift + T
Text size: next size
Alt + Shift + C
Contrast: next level
Alt + Shift + F
Focus outline on or off
Alt + Shift + L
Highlight links on or off
Alt + Shift + H
Highlight headings on or off
Alt + Shift + M
Stop animation on or off
Alt + Shift + G
Reading guide on or off
Alt + Shift + R
Reset everything

Policies & Choices

Open Page →