Coordinating staff across two store locations

The second location is not twice the first

Opening a second location feels like a scaling problem and behaves like a communication problem. The operational work roughly doubles, which is expected and manageable. What surprises people is that the coordination work grows faster than that, because every question that used to be answered by looking across the room now requires somebody to ask somebody.

With one location, a great deal of coordination happens through ambient awareness. You can see that the delivery arrived. You can hear that a customer is upset. Nobody has to be told, so nobody has to remember to tell anyone. That mechanism is free, invisible, and it does not survive a second site.

The teams that struggle are usually the ones who were coordinating well informally, because they have no explicit process to extend. The teams that cope had already written things down, often because somebody worked remotely.

The four things that break first

Stock that exists somewhere else

The most immediate one. A customer wants an item, it is out at this location and in stock at the other, and the answer to whether that helps depends on questions nobody has agreed on. Who can authorise a transfer. How long it takes. Who tells the customer. Whether it is worth doing for one item.

Without an agreed rule, each case is decided individually and inconsistently, and staff learn to just say no because saying no is never questioned.

Decisions that used to be made by whoever was there

With one site, the person on shift decides, and if they are unsure they ask the owner who is usually around. With two, the owner is at one of them, and the other location either escalates constantly or stops escalating entirely. Both are bad, and the second is worse because it is quiet.

The pace of one location being invisible to the other

One site is slammed while the other is quiet, and neither knows. Staff cannot be moved, help cannot be offered, and the busy site absorbs it while the quiet one wonders why the owner seems stressed.

Information arriving at one site and not the other

A supplier calls the first location about a delay. That is now known at one site and unknown at the other, and there is no mechanism that makes it travel. The second location finds out when a customer asks.

The structure that prevents most of it

Three things, in order of how much they help per unit of effort.

1. A shared place where store-wide things are said

Not a group chat per location, which is what most teams create first and which guarantees information stays where it landed. One channel that both locations read, for anything true of the business rather than of a site.

The rule that makes it work is about subject rather than sender: supplier news, stock arriving, promotions, policy changes and anything a customer might ask either site about goes in the shared place. Site-specific logistics stay local. If you split by team instead of by subject, the split collapses within a fortnight because the same person has both kinds of thing to say.

2. Written decision limits, per location

The single highest-value thing on this list. Whoever is senior at each site needs to know, in writing, what they can decide alone: refund up to a value, reship up to a value, discount on a service failure, transfer stock under some threshold.

Set these higher than feels comfortable. The cost of a slightly wrong decision made quickly is almost always lower than the cost of a customer waiting while somebody tries to reach the owner, and the second cost is invisible so it never gets counted.

3. One place to see both locations at once

Shopify already knows stock by location. The gap is usually not data but habit: nobody looks, because looking means opening a report nobody has bookmarked. Whatever you use, the test is whether somebody serving a customer can answer "do we have it at the other shop" in under fifteen seconds without phoning anyone.

The stock transfer rule, written out

Transfers cause more friction than their volume justifies, so it is worth settling once rather than case by case. A workable default:

• Customer is waiting and the item is at the other site: the senior person on shift can authorise a transfer without asking, up to an agreed value.
• Who moves it: agree this in advance. A weekly run is usually better than ad hoc trips, because ad hoc trips get volunteered for and then not made.
• Who tells the customer: the person who took the request, with a date rather than a promise of a call back.
• Below what value it is not worth it: say the number out loud. Staff will otherwise either transfer everything or nothing.

Write it once, put it where both sites can see it, and revisit it after a month. The first version will be slightly wrong and that is fine.

What not to do

Do not create a channel per location and assume things will cross between them. They will not. Location channels are useful for logistics specific to a site, and they are where store-wide information goes to be forgotten.

Do not route everything through the owner. It works at two locations and it is the reason a third is painful. The owner becomes the integration layer between sites, which is a full-time job nobody has agreed to do.

Do not assume the quieter site is fine because it is quiet. Silence from a location is ambiguous, and the most common reason for it is that nobody there is sure what is worth mentioning.

The weekly rhythm that replaces ambient awareness

The thing you lost when you opened the second site was not information, it was incidental contact. People used to absorb context by being near each other. Once they are not, the only replacement that works is a small amount of deliberate, scheduled contact, and the emphasis is on small.

A shape that works for most two-site operations:

• A written open, once a day, per site. Three lines, not a report: what is short, what is unusual today, anything the other site should know. It takes two minutes and it is the single highest-value habit on this list, because it catches the supplier call nobody thought to pass on.
• One short call a week between the two senior people. Fifteen minutes, same time, on the phone rather than in writing. Written updates are good at facts and bad at tone, and tone is how you find out a site is struggling before the numbers say so.
• A monthly look at the decision limits. Not the numbers themselves, but whether anybody has hit them. A limit nobody has ever reached is too high to be informative. A limit somebody hits weekly is too low and is generating escalations you do not need.

Resist adding more than this. The failure mode for multi-site coordination is not too little process, it is a weekly meeting that grows to an hour, stops being useful, and gets cancelled, taking the daily habit down with it.

When the third location arrives

Most of the above still works, with one exception: anything that relies on the owner being the link between sites stops working entirely. Two sites can be held together by one person paying attention to both. Three cannot, and the attempt is how owners end up working every weekend.

The change that has to happen before the third site opens is that sites communicate with each other rather than through you. In practice that means the shared channel becomes genuinely shared rather than an owner broadcast, and the senior people at each site have each other's numbers and use them without asking permission first.

Common questions