Operations

How to Train Café Staff for a Mobile Ordering App Launch

By The Tany Team 10 min read

The software part of launching a branded ordering app is genuinely easy now — a café on Square can be live in about a day. The part that decides whether the launch works is what happens on the bar at 8:15 on a Tuesday, when the counter queue is nine deep and a mobile ticket prints for a drink that was promised in six minutes.

Almost every week-one order-ahead failure is an operations problem wearing an app costume: a prep time set optimistically, a pickup shelf nobody can reach, a ticket that arrived while the grinder was running. This is the runbook for avoiding all of that.

Why staff training is the actual launch risk

Order-ahead changes something structural about a café: it decouples when the order arrives from when the customer arrives. Every habit your team has — read the cup, call the name, next customer — assumes those two things happen together. Order-ahead breaks that assumption, and until the team has a new rhythm, tickets get made too early (cold drink, melted ice) or too late (customer waiting, watching).

For a sense of how large this channel becomes when it works: Starbucks reported that mobile order transactions were 33% of total transactions at US company-operated stores in Q2 FY26, up from 31% across all of FY24 and FY25. Roughly one in three orders arriving without a person attached to them is a fundamentally different service model, not a feature.

Starbucks is a 40,000-store chain with dedicated mobile stations, so do not treat their numbers as a forecast for your café. But their published service targets are instructive about what to measure. On the Q2 FY26 call, CEO Brian Niccol described targets of “4 minutes in the cafe, 4 minutes in the drive-thru and better than 12 minutes for promised times in mobile order pickup,” with about 80% of stores or better hitting them. The lesson for an independent operator is not the specific numbers — it is that mobile pickup gets its own service target, separate from counter service, and that a company at that scale still considers 12 minutes the standard worth managing to.

There is no credible published data on order-ahead throughput or labour impact for independent cafés specifically, so treat any vendor claiming a precise percentage lift with suspicion. What follows is an operations sequence, not a promise.

Two weeks before launch: set the machine up honestly

Set a prep time you can actually hit at peak

In your Square online ordering profile’s pickup options, you enter the number of minutes you need to prepare a pickup order. Every instinct will tell you to enter your best-case number. Do not. Enter your peak number, padded.

The reason is asymmetric: a customer who arrives to a drink already waiting is mildly pleased. A customer who arrives to a drink that is not started is actively annoyed and standing in your café being annoyed in front of other customers. Under-promising costs you almost nothing; over-promising costs you the customer.

At launch, take your realistic peak prep time and add 3–5 minutes. You will tighten it in week three with real data.

Cap the orders per window

The same Square settings let you limit orders per pickup window, which Square handles in 15-minute blocks. This is the single most important setting in the launch, and most operators skip it.

Work out how many mobile drinks your bar can absorb in 15 minutes on top of the counter queue during your busiest block. That number, minus a buffer, is your cap. It is far better to tell a customer the 8:15 window is full and offer 8:30 than to accept every order and blow every promise.

You can also set how many days ahead customers may schedule, require a lead time for advance orders, and set a daily cut-off that pushes later orders to the next day. If a meaningful share of your volume is scheduled rather than immediate, our guide to scheduling pickup orders on Square goes deeper on those settings, and setting prep times for Square online ordering covers the timing math in detail.

Build the pickup zone before the first order

Physical layout does more work here than any setting. Requirements:

  • Reachable without crossing the order queue. If a mobile customer has to push past the line, you have created a conflict at your busiest moment.
  • Visibly labelled. A sign that says “Mobile Pickup.” Customers who have never done this before need to be told where to go without asking.
  • Segmented by type. Hot drinks, cold drinks, bagged food. Cold drinks especially should be the last thing made and the first thing collected.
  • Names readable from a metre away. Whatever labelling you use, the customer should be able to identify their order without picking up three cups.

One week before: wire the tickets and rehearse

Decide how orders reach the line

An order that arrives silently is an order that arrives late. On Square, Order Manager gives you persistent full-screen notifications that chime when orders arrive, with a choice of notification sound and an alert duration you can set from 3 to 60 seconds or until dismissed, plus printer profiles for receipts, tickets and labels.

Set that alert deliberately. In a café with a loud grinder and a steam wand, a three-second chime is decorative. Longer alert durations and a full-screen notification are the correct choice at launch.

If your volume warrants a kitchen display instead of paper, Square KDS runs $30/month per device with Square Plus or $20/month per device with Square Premium. It is a good upgrade at volume and an unnecessary complication on day one. Our walkthrough of Square KDS for cafés covers when the switch is worth it.

Run twenty real test orders

Not a demo. Real orders, placed by your actual staff from their own phones, refunded afterwards. Cover these five scenarios at minimum:

  1. A simple drink.
  2. A drink with heavy modifiers — oat milk, extra shot, half sweet, extra hot.
  3. Food plus a drink, where the two are ready at different times.
  4. An order scheduled for later in the day.
  5. A mistake that needs a refund, so someone has done it once before a customer needs it. (Handling refunds on Square online orders covers the mechanics.)

Time each one from submit to handoff and write the times down. That list is your baseline, and it is what you will compare against in week one.

Write three scripts and post them at the station

Staff freeze on the sentences, not the steps. Give them the sentences:

  • The handoff. “Mobile order for Sam? It’s right here on the shelf — have a great one.”
  • The not-ready-yet. “You’re a couple of minutes early, we’re on it — grab a seat and I’ll bring it over.” (Bringing it over is worth the ten seconds. It converts a wait into service.)
  • The on-premise app order. “Just ordered on the app? I’ve got you — it’ll come up with the counter orders, I’ll call your name.”

That third one matters more than it looks. Even Starbucks has publicly described re-sequencing orders for customers who order via mobile while already in the store. In a small café the fix is simpler: train staff to recognize the case and make the drink in queue position rather than letting the ticket sit out its full prep time while the customer stands there watching.

Launch week: soft launch, then huddle daily

Soft launch to regulars only. Invite twenty or thirty regulars before you promote anything publicly. They are forgiving, they will tell you the truth, and the volume is absorbable. Public promotion comes after the operations are stable — the sequence in how to get customers to download your restaurant app works far better on a café that can already execute.

Five-minute huddle every morning. Review yesterday: how many mobile orders, what went wrong, the longest ticket time. Fix exactly one thing per day. A launch that improves by one small thing daily for ten days is transformed; a launch that waits for a big review meeting is not.

Watch four numbers in week one:

MetricWhat good looks like at launchWhat it tells you
Median submit-to-ready timeComfortably inside your promised prep timeWhether your prep time is honest
Orders arriving before readyFalling day over dayWhether pacing and prep time are aligned
Mobile orders per 15-min peak blockAt or below your capWhether the cap is set correctly
Remakes / complaints on mobile ticketsNear zeroWhether handoff and labelling work

You do not need software for this. A clipboard by the pickup shelf for one week is enough.

Two weeks after: tighten and promote

Once median ticket times are stable and staff stop hesitating, do two things in order. First, reduce the padded prep time toward reality — a six-minute promise that is consistently met is better marketing than a three-minute promise that is not. Second, start promoting properly: in-store signage, receipt messaging, your Google Business Profile, and a first-order incentive.

Only then is it worth turning on the retention mechanics — push notifications and loyalty — because they drive traffic to a system that can now handle it. Pushing a promotion at a bar that is still learning the rhythm is how you convert your best customers into your most disappointed ones.

Where the platform choice helps

Some of the friction above is unavoidable — it is the cost of a genuinely new service channel. Some of it is a platform choice. An ordering app that reads directly from your existing Square catalog means no second menu to maintain and no divergence between what the app sells and what the POS knows. Orders landing in the same place your counter orders already land means one workflow instead of two.

That is the setup Tany provides for cafés on Square: a branded iOS and Android order-ahead app plus web ordering, with loyalty and push built in, on your existing POS for $99 CAD/month per location with 0% commission. But the runbook above applies whatever you launch on. The app is a day of setup. The rhythm on the bar is two weeks, and it is the part that decides whether the channel sticks.

The takeaway

Treat the launch as an operations project with a software component, not the reverse. Set an honest prep time, cap your windows, build a pickup zone a stranger can find, run twenty test orders, write three scripts, soft launch to regulars, and huddle for five minutes every morning for a week.

Do that and by week three, mobile tickets will be as boring as counter orders — which is exactly the goal.

Sources

Step by step

  1. 1
    Week 2 before launch — set prep time and pacing limits

    In your Square online ordering profile, set the number of minutes you need to prepare a pickup order and turn on the limit for orders per 15-minute window. Base both on your actual peak throughput, not your best-case throughput.

  2. 2
    Week 2 — build the physical pickup zone

    Choose a spot for mobile pickup that a customer can reach without crossing the order queue. Add a sign, a shelf or counter, and a place for bagged items and cold drinks. Decide where labels go so names are readable from a metre away.

  3. 3
    Week 1 — decide how orders reach the line

    Confirm whether orders will land on a kitchen display, an order manager screen, or a printed ticket. Set the notification sound and alert duration so a ticket cannot arrive silently during a rush.

  4. 4
    Week 1 — run twenty test orders with staff

    Have each staff member place real orders from their own phone across different scenarios: simple drink, heavy modifiers, food plus drink, scheduled for later, and a mistaken order that needs a refund. Time each one from submit to handoff.

  5. 5
    Week 1 — write the three scripts staff will actually say

    Script the handoff greeting, the not-ready-yet line, and the customer-is-here-and-wants-to-order-on-the-app line. Three sentences each, posted at the station.

  6. 6
    Launch week — soft launch to regulars only

    Invite a small group of regulars before any public promotion. Volume you can absorb reveals the same problems as volume you cannot, without the reputational cost.

  7. 7
    Launch week — hold a five-minute daily huddle

    Each morning for the first week, review yesterday's mobile orders: how many, what went wrong, how long the longest ticket took. Fix one thing per day.

  8. 8
    Week 2 after launch — tighten prep times and promote

    Once median ticket times are stable and staff are comfortable, reduce the padded prep time toward reality and start promoting the app to your full customer base.

Frequently asked questions

How long does it take to train café staff on mobile order-ahead?
The mechanics take under an hour — most staff learn to read a mobile ticket and hand off an order in one shift. What takes two weeks is the operational rhythm: knowing when to start a mobile drink relative to the counter queue, where the order goes when it is ready, and what to say when a customer arrives early. Plan for a short training session plus a two-week settling period.
What is the most common mistake when launching order-ahead in a café?
Promising a pickup time the line cannot hit during peak. A prep time that works at 2pm will fail at 8:15am, and a customer who arrives to a drink that is not ready has a worse experience than one who waited in line. Set a padded prep time at launch, cap orders per pickup window, and tighten only once you have real timing data.
Can Square limit how many mobile orders come in at once?
Yes. In the pickup options for your Square online ordering profile you can set a prep time in minutes and turn on a limit on orders per 15-minute pickup window. You can also set how many days ahead customers may schedule, require a lead time for advance orders, and set a daily cut-off after which orders roll to the next day.
Do I need a kitchen display system to run mobile ordering?
No. Many independent cafés run order-ahead on a printed ticket or the order manager screen on their POS. A KDS helps once volume is high enough that paper tickets get lost or out of sequence. On Square, KDS costs $30 per month per device with a Square Plus subscription, or $20 per month per device with Square Premium, so it is worth adding when ticket volume justifies it rather than on day one.
How do I handle a customer who is standing in the café and orders on the app?
This is a real and surprisingly common case — even Starbucks has publicly discussed re-sequencing orders for customers who order on mobile while already in the store. The simple rule for an independent café is to treat an on-premise mobile order like a counter order and make it in queue position, then tell the customer where to wait. Train staff to recognize it rather than letting the ticket sit for the full prep time while the customer watches.