At ten stores, local marketing is one helpful person in the brand team. At four hundred it is a queue nobody can clear. This is the shift that fixes it: stop producing per store, start producing per campaign, and measure the one number that actually moves network volume.
At ten stores, local marketing is a person. Somebody in the brand team knows each operator, builds what they ask for, and sends it over. It works, and it does not look like it is about to stop working. Then the network doubles, and the same person is still in the middle of every request.
Every local campaign starts with somebody asking for something. At scale the queue is longer than the campaign is relevant, so the good ideas arrive late and the partner stops asking.
One campaign, four hundred stores, each with a different name, address and phone number, across five ad formats and two print sizes. This is where design budgets quietly disappear.
Results come back from three ad accounts and a printer, and somebody rebuilds the network view by hand every month. By the time it is finished, it is old.
A third of the network runs nothing, and nobody notices, because the reporting only shows the stores that did run something.
Partners who cannot wait start boosting from their own phones with their own creative. That activity is now completely outside your view.
Every campaign rebuilt by hand is targeted slightly differently. Two neighbouring partners end up bidding against each other and nobody planned it.
The unit of work stops being a store and becomes a campaign. Head office builds a package, and the package already knows everything except the two things only the partner can decide.
Targeting is the part that usually does not scale, because doing it properly per store is slow. Amplaro fits each order to that store's real service area automatically, so the four hundredth store is targeted as carefully as the first. See delivery area targeting.
The instinct is to give partners everything. The result is a partner scrolling through forty campaigns, picking nothing, and closing the app. A catalogue is a menu, and menus work when they are edited. Most networks need something in each of these slots and not much else.
| Package type | Who runs it | When |
|---|---|---|
| Always on | Every store, continuously | Background presence, low budget, never off |
| Seasonal or national tie in | Most of the network | Timed to sit under a national flight |
| New store opening | New and reopened stores | First six to eight weeks after opening |
| Competitor response | Stores under pressure in one suburb | On request, short bursts |
| Winback | Stores with a lapsed customer problem | Quarterly, or when the data says so |
| Local community | Any store with something genuinely local on | Editable dates, fixed brand frame |
Swipe the table sideways to see the rest.
Building a quarter of campaign packages in any console is the same work every time. Type the title, paste the copy, set the dates, pick the channels, set the budget range, repeat. For a marketing team planning a year across several regions, that is hours that produce nothing new. The thinking already happened in the planning document.
Amplaro is, as far as we can tell, the first franchise marketing platform with an MCP connector. In practice that means you connect the platform to Claude, attach the plan, and ask for the packages and calendar entries to be created. They arrive as drafts for a human to check and publish. Dates are never guessed, only read, and running the same document twice does not create duplicates. Everything it does is written to the audit log as Claude, with the person who asked.
The full picture, including what it will and will not do on its own, is on building campaigns with AI.
At network scale the biggest lever is not campaign performance. It is the share of stores running anything at all. A network where sixty percent of stores are active is doing roughly twice the local volume of one where thirty percent are, and no amount of creative testing closes that gap.
Which means head office needs one report more than any other: the list of stores that have launched nothing. Not a chart, a list, with names, so somebody can call them. Amplaro reports active stores, spend and results per store and per network, and it does not hide the quiet ones. See reporting and rolling out.
The share of the network that has launched anything at all. The single most useful number head office has, and the one most reporting leaves out.
Partners who ordered a second time. A first order can be enthusiasm or a favour. A second one means it worked for them.
How long from invite to live, per store. If this is measured in months, the problem is onboarding, not marketing.
You stop producing per store and start producing per campaign. Head office builds a campaign package once, with the creative, copy, channels, targeting rules and budget range already set, and every store orders against it. The work of adding the four hundredth store is then the same as adding the tenth, which is the only version of this that survives contact with a real network.
A campaign package is one approved campaign, ready to run, that any store can order for their own area. It carries the creative, the ad copy, which channels it runs on, the dates it may run between, the budget floor and ceiling, and the targeting rules. The partner picks it, picks a budget and a date, and the store details are filled in automatically. See how it works.
Fewer than you would expect. A catalogue of six to ten packages that are always relevant beats forty that a partner has to read through. If a partner opens the app and cannot decide in under a minute, the catalogue is too big. Retire packages the way you retire menu items.
Make the approved route the fastest route, then chase the quiet stores. Every network has a group that never signs in, and no amount of software fixes that on its own. What software can do is tell you exactly who they are, in a list, this week. Most rollouts stall because nobody was watching that list. See rolling out.
Yes. Amplaro exposes an MCP connector, so you can connect the platform to Claude and build packages by describing them, or load a whole quarter of calendar entries by attaching the plan you already wrote. It creates them as drafts for a human to check. Building a season of campaigns by hand in a console is hours of work. This is the same work in one pass. See building campaigns with AI.
That request should go to head office, not to a local designer. In practice most of these turn out to be the same handful of requests across the network, which means they belong in the catalogue as a package rather than as a one off. The ones that are truly local, a school fundraiser or a street festival, are the case for a flexible package with editable dates and a fixed brand frame.
It replaces the part of the agency work that is repeated four hundred times. Strategy, brand campaigns and the big national creative are still a job for people who are good at it. Resizing that creative for every store and rebuilding the same targeting every month is not.
By deciding what may run rather than approving each thing that does. Guardrails are set once in the package: fixed creative, copy from the catalogue, budget limits, real end dates on offers, and territory targeting per store. An approval queue does the same job but adds a person who becomes the bottleneck at exactly the moment the network gets big. See brand control.
The same problem, sector by sector: franchise marketing by industry.
Connect Amplaro to Claude and load a whole quarter from the plan you already wrote.
How the local layer should sit under the national plan instead of competing with it.
Importing stores, inviting partners, and chasing the ones who never sign in.