Menu Intelligence
New Item Incrementality: Added Sales, or Moved Sales?
“Our new item is selling. Is it actually growing the business, or just moving sales around the menu?”
A new menu item is incremental only if the total business grew beyond where it was already heading — beyond its pre-launch trend and its normal seasonal pattern — and that growth traces to the item rather than to momentum that was coming anyway. The item's own sales figure cannot answer this, and neither can its category's: a category can grow while the rest of the menu shrinks by the same amount. Quantiiv measures what a launch really added, net of what it took from the menu around it.
The readout separates three stories that look identical in a sales report: genuine new demand, where new and returning guests show up above the brand's normal rate or existing guests visit more often; trade-up, where the same visits spend more; and substitution, where the same dollars land in a different line item. Only the first grows the business. The second can still be worth having. The third is a launch cost with no return, and it is the most common of the three.
Sound Familiar?
The launch number flatters every launch
The new item did $X in its first month is a gross figure. If most of that volume traded out of the items it sits next to, the menu got more complex and the business got nothing. Almost no launch recap nets this out.
Category growth masquerades as business growth
The new item's category is up, so the launch is declared a win. But if total sales grew no faster than they were already growing, the category gained at the rest of the menu's expense, and the win is a mix shift wearing a growth costume.
Verdicts arrive before the data can support them
Repeat purchase and stickiness get judged in week two, at brands whose guests visit once a month — guaranteeing a verdict of 'no repeat behavior' that says nothing about the item. Early reads answer early questions; they cannot answer retention.
How Quantiiv Answers It
- 1
Establish what would have happened anyway
Before crediting the item with anything, we build the honest comparison: the trend the business and category were already on before launch, the seasonal pattern for the period, and comparable stores where the story played out without the launch. Growth that was already coming does not belong to the item.
- 2
Test the total business, not the category
The core question is whether the whole business grew beyond its baseline, not whether the item's category did. This one test catches most false positives: launches that reshuffled demand across the menu while the topline did exactly what it was going to do anyway.
- 3
Net out cannibalization item by item
Which existing items lost volume while the new one grew, and by how much against their expected path — confirmed at the basket level, where substitution is visible directly. The launch's net contribution is its sales minus what it took.
- 4
Identify who bought it and what changed
Where customer identity exists, buyers split into new, returning-lapsed, and existing guests — and the split is judged against the brand's normal rates, because every period has new faces. Existing buyers split again: visiting more often, spending more per visit, or simply reallocating the same spend.
- 5
Time the verdict to the brand's rhythm
Trial velocity and customer acquisition can be read early. Repeat behavior can only be judged once guests have had a realistic chance to return, which depends on how often they visit. The readout says what can be concluded now, what needs more time, and when to look again — including after removal, where the category's response is the final word on what the item was carrying.
Why Quantiiv
The bar is total-business growth beyond baseline
Most launch analyses stop at item sales or category lift. Quantiiv's incrementality standard requires the whole business to beat its own expected path — the same honest test we apply to pricing actions and promotions.
Honest about what the data can say yet
A launch read two weeks in gets conclusions about trial and acquisition, not retention. Every claim in the readout is labeled with what supports it, and premature questions get a date for the answer instead of a guess.
Frequently Asked Questions
How do you measure whether a new menu item is incremental?
Compare total business performance after launch against what it was on track to do anyway — its pre-launch trend, its seasonal norm, and comparable stores without the item. Then net out what adjacent items lost, and check who bought it: new and lapsed guests above the brand's usual rate, or existing guests visiting more or spending more in total, are the signatures of real incremental demand.
Our new item's category is growing. Isn't that success?
Not by itself. A category can grow entirely at the expense of the rest of the menu, leaving total sales exactly where they were heading. Category growth is consistent with a great launch and with pure substitution; only the total-business test against a fair baseline distinguishes them.
How do you detect cannibalization from a new item?
Track the items the new one competes with against their expected path from pre-launch trend, and confirm at the basket level: which items stopped appearing in baskets where the new item shows up. Volume those items lost below expectation is the cannibalization cost that nets against the launch's gross sales.
How soon after launch can you judge a new item?
It depends on the question. Trial velocity and whether the item is attracting new or lapsed guests can be read within weeks. Repeat purchase can only be judged after typical guests have had a realistic opportunity to return, which for many brands means a month or more. Judging stickiness before that window produces confident, wrong answers.
Do we need a loyalty program to measure this?
No. The incrementality and cannibalization reads come from transaction data alone. Customer-level analysis — who bought it and how their behavior changed — is deeper where loyalty or digital-order identity exists, and every customer-level finding is reported with the share of sales it covers, so the strength of the evidence is always explicit.
Related Problems We Solve
LTO & New Item Analysis
“Our LTO sold well. But did it grow the business, or did it just cannibalize the core menu?”
Read moreMenu Engineering
“Which items actually drive our revenue and margin, and how should the menu change?”
Read moreCustomer Analytics
“Who are our customers, are they coming back, and which ones are actually worth investing in?”
Read moreStop Fighting Your Data.
Start Using It.
Transform fragmented restaurant data into actionable insights—with just an email to ROGER.
400+
Locations Served
15+
POS Integrations
10hrs
Saved Per Week
Powered by Quantiiv
Enterprise Restaurant Intelligence