How to roll out a new menu item across a franchise system

To roll out a new menu item, test it first in three to six stores for four to eight weeks, with a matching group of control stores that do not get it. Measure attach rate, the change in average check against the control stores, food cost and ticket times. Fix what the test finds, then launch across the system in waves, with recipe training, POS buttons and supply in place before day one.

After launch, check attach rate store by store every week for the first month. A store far below the rest almost always has a setup problem, such as a hard-to-find button or an item that keeps running out, not a demand problem.

Decide what the new item is for

Every new item should have one job. Write it down before the test, because it decides what you measure.

• An add-on, like a side or dessert, should raise average check. Measure attach rate and check lift.
• A traffic driver, like a limited-time offer, should bring in more orders. Measure order count against control stores.
• A margin item should replace something less profitable. Measure food cost percentage and what it pulls sales from.

Then set the pass marks in advance. For example: attach rate of at least 6% in most test stores, check lift of at least $0.25 per order over control, food cost at or under 30%, and ticket times up no more than 15 seconds at peak. Pass marks set after the results come in tend to bend toward a yes. Our guide on how to calculate food cost percentage shows how to cost the recipe.

How to test a new menu item in a few stores

Pick test stores that look like the system, not just your best operators. Mix high and low sales stores, different markets and different service mixes. Then pick a similar set of control stores that keep the current menu.

1. Set up the POS the same way in every test store. Same item name, same button position, same modifiers.
2. Train the build and the register line before the first day, not during it.
3. Confirm supply for the full test plus a buffer, so the item never runs out.
4. Run it for four to eight weeks. Ignore week one when you judge it.
5. Visit each test store at least once at peak to watch the build and the line.

Field teams are useful here. A visit during the test tells you whether the item fits the line in practice. Our guide on how to run franchise field visits covers what to look at.

How to measure attach rate

Attach rate is the share of orders that include the item.

Attach rate = orders containing the item ÷ total orders, over the same period.

Here is a four-week test of green chile fries, a $4.50 add-on, at Switchback Grill in Castle Rock, a fictional Mountain West burger chain.

Attach rate tells you how many people bought it. It does not tell you whether they would have spent that money anyway, on regular fries for example. For that, compare the change in average check at test stores with the change at control stores over the same weeks.

Control stores rose $0.15 on their own, from price, season or weather. Only the $0.30 above that belongs to the new item. Across 8,400 orders at Castle Rock, that is about $2,520 of new sales in four weeks, less than the $3,213 the item sold, because some of those dollars would have been spent on other sides. Our guide on how to increase average check covers other ways to move that number.

Go, fix or stop

Compare the test results with the pass marks you set at the start.

• Go: it clears every mark in most test stores.
• Fix: it clears the demand marks but misses on speed or cost. Change the build, the hold time or the portion and run two more weeks.
• Stop: attach rate or check lift are well under the mark in most stores. Pull it and keep what you learned.

In the Castle Rock example, ticket times rose 12 seconds at peak. That is inside a 15-second mark, so it passes. A bigger jump would point to a prep or station change before launch. Our guide on how to improve speed of service has more on keeping the line moving.

How to launch a menu item across a franchise system

A good rollout kit means every store opens day one the same way. Include:

• A recipe card with build photos, portions, prep steps and hold times.
• POS setup done centrally where you can, with the same item name and button position in every store, so attach rate can be compared.
• Supply confirmed through your distributors before the launch date, with a plan for the first reorder.
• A training date for every store before launch, and a short video for new hires.
• One short register line, the same in every store.
• Marketing timed after stores are ready, not before.

Launch in waves, by region or by distribution center, a week or two apart. The first wave finds the problems the test missed while they are still small. Tell franchisees early. Show them the test results, the expected food cost and what it should add per order, so they can plan next week's schedule and their ordering.

For a limited-time offer, also set the end date, the last order date for supply, and a plan for leftover inventory. Our guide on how to reduce food waste covers the last part.

Measure the rollout store by store

For the first four weeks after launch, look at attach rate for every store, not just the system average. If most stores sit near 8% and two sit near 3%, look at those two stores first. The cause is usually something about the store's setup: the button is on a second screen, the item ran out, or the register line was never trained. Fix the setup before drawing conclusions about the market.

Our guide on how to benchmark franchisee performance fairly covers how to compare stores without punishing the ones in harder markets.

Common mistakes in a menu item launch

• Testing with no control stores, so a busy season looks like a hit.
• Judging the item on week one.
• Different item names in different POS systems, so nobody can compare stores.
• Launching before supply is confirmed, then running out in week two.
• Counting item sales and calling it new money, without checking what it pulled from.
• Ignoring ticket times at peak.
• Launching everywhere on the same day, so a problem hits every store at once.

A weekly habit for new menu items

For the first eight weeks after any launch, take ten minutes each Monday. Pull attach rate by store for last week, rank the stores, and look at the bottom three. Check average check against the same weeks last year and food cost on the item. Pick one store to call or visit about setup. After eight weeks, fold the item into your normal weekly numbers.

Franlever's AI COO compares each store with its own recent weeks and with the middle store in its brand, so a store falling well behind the rest after a launch can show up as that store's one lever for the morning, sent to whoever runs it. See AI COO.

General information for restaurant and franchise operators, not legal, tax, accounting or financial advice. Figures in examples are illustrations, not predictions or promises of results. Laws and platform policies change and differ by location, so check the current rules and talk to a qualified advisor about your situation.