Cannibalization Analysis: Protect Retail Revenue 2026

Retail cannibalization analysis measures how much of a new store's or new product's sales are transferred from something you already own instead of added to the business. It turns a vague worry about overlap into a number you can weigh against incremental profit, before the lease is signed or the SKU reaches the shelf.

What retail cannibalization analysis measures

Retail cannibalization analysis measures four things: product cannibalization, location cannibalization, the cannibalization rate, and net impact. The same four parts apply whether the new unit is a storefront or a shelf item, and only the last one, net impact, says whether the business grew or just moved existing sales from one unit to another.

• Product cannibalization: a new item pulls demand from an existing one.
• Location cannibalization: a new store draws customers from a store you already run.
• Cannibalization rate: sales an existing unit lost, as a percentage of the new unit's sales or the existing unit's prior base.
• Net impact: incremental sales minus cannibalized sales.

Cannibalization almost always happens, so the real question is whether it was a trade made on purpose. Retiring an aging product for a better one is a hit somebody picked; a lease three miles from an existing store because the corner looked strong is not. For a plain-language primer on the concept, start with our guide to internal cannibalization.

The stakes are concrete in this market. Coresight Research's midyear 2026 review puts the full-year count at roughly 6,428 US store closures against 4,482 openings, fewer on both sides than it forecast in January. Closures are slowing, but openings still trail them by nearly 2,000 locations, so growing brands keep picking sites where somebody just handed one back, and every one of those calls carries an overlap question.

I'm Clyde Christian Anderson. At GrowthFactor I work with expansion teams on exactly this question. Cavender's Western Wear alone has evaluated more than 2,000 sites on the platform, and the cannibalization question comes up on nearly every one.

How to run a cannibalization analysis in five steps

A cannibalization analysis runs the same five steps whether the unit is a store or a SKU: define what you are measuring, build a control set, estimate transfer against that control, convert transfer into profit, and fix the decision rule before you see the result. Most teams skip the second one, which is where the analysis quietly stops being evidence.

1. Define the unit, the window, and the affected set

Nail down three things before you pull any data. The unit is the new store or the new item. The window is the comparison period, commonly 8 to 13 weeks after launch against the same span a year earlier, long enough to clear the opening bump. The affected set is the specific existing stores or items you expect to take the hit.

Choosing the affected set in advance keeps the analysis honest. Affected stores picked after seeing which ones declined are a story, not an effect, so write the list down and date it.

2. Build a control set, because before-and-after lies

There's a dozen reasons a store's sales move in any 90-day window: weather, a road closure, a competitor opening, a category slowdown. Call the before-and-after gap cannibalization and the new store gets credit for all of them.

The fix is a control set. Pick stores or items that look like the affected ones on the variables that matter (format, volume band, trade-area profile, seasonality) but sit far enough away that the new unit cannot reach them. The change in the affected set minus the change in the control set over the same window is your cannibalization estimate. Analysts call the design difference-in-differences. When no single unit is a clean match, a synthetic control built as a weighted blend of several donor units does the same job. PyMC Labs walks through the Bayesian version, which returns a credible range for the cannibalization share instead of one brave number.

Use several control units per affected unit where you have them, because a single control brings its own noise, and freeze the control set before launch just as you froze the affected set.

3. Estimate transfer, not correlation

A control set tells you that demand moved. It does not tell you that it moved to your new unit. But that's a separate measurement, and the method splits by unit type.

For stores, the evidence is customer origin. Map where the affected store's customers live or commute from, project the new site's draw, and calculate the share of the existing store's origin volume inside it, using loyalty records, transaction ZIP codes, or foot traffic patterns. What does not work is a distance ring, because two stores five miles apart can fish the same commuter pool while two stores a mile apart serve entirely different crowds.

Then weight it, because two trade areas can share a thin slice of area and a thick slice of demand. CARTO's cannibalization tutorial measures overlap as the share of an existing store's catchment population inside the new site's catchment, not the share of map area. Lead with the demand-weighted figure and label the convention.

For products, the evidence is the basket. The volume is what makes this worth systematizing: NielsenIQ counted more than 3,500 new brands and sub-brands launched across five European markets in 2025, and only about a third reached 1 percent household penetration. Look at customers who bought the existing item before the launch and see what they bought after: the new item, the old one, or nothing from the category. Switching rates from that group are your transfer estimate. Category-level basket analysis fills in the rest, separating customers who traded across from customers who were newly attracted to the category, which is halo rather than cannibalization.

Halo is the part teams forget about. A new unit that lifts total category volume while pulling from a neighbor is a different call from one that only splits an existing line, so measure both directions.

One more fix while you're in the data: stop treating the answer like two buckets. A new store's forecast demand arrives from five places: transfer from your own nearby units, customers taken from a competitor, trade-area demand nobody was serving, relief for a unit already at capacity, and a residual whose origin you cannot attribute, which deserves its own line rather than a quiet rounding into the others. An 18 percent transfer rate on a site that also wins heavy competitor share is a very different deal from 18 percent transfer and nothing else, and a cannibalized-or-incremental read cannot tell the two apart.

4. Convert transfer into net profit

Transfer is a sales number, and decisions get made on profit. Three conversions take it from there.

1. Cannibalization rate (%): for locations, estimated sales lost by the existing store divided by its sales before the opening, × 100. For products, divide by the new item's sales instead.
2. Incremental sales: the new unit's estimated sales minus the sales it cannibalized.
3. Net profit impact: incremental sales × margin, minus the lost profit on cannibalized sales.

For a step-by-step walkthrough of each formula with worked examples, see our guide to calculating cannibalization rate. To ask the broader question of whether a market can hold another unit at all, see market saturation analysis.

The math on a sample store looks like this:

5. Set the decision rule before you see the number

Decide what result would make you walk away before the model runs. A rate set in advance is a gate; a rate interpreted after the fact is a negotiation, and a great-looking corner is good at talking a team right past a threshold nobody wrote down.

A workable rule names three things: the overlap rate above which the site needs a different answer, the minimum net profit the opening has to clear anyway, and who gets to override it. Ten to 20 percent of new unit revenue is a common tolerance band, but margin and category move the line. Our prevention playbook covers how to build those gates into site selection.

Product cannibalization analysis vs. store cannibalization analysis

Product cannibalization analysis and store cannibalization analysis share one skeleton, difference-in-differences against a frozen control set, and differ in almost everything else. Product analysis reads SKU and basket data over a few weeks and feeds assortment calls that can be reversed. Store analysis reads customer origin before a lease that cannot be reversed for a decade.

The reversibility row is why store cannibalization gets analyzed before commitment: a product launch that eats its neighbor is a bad quarter, while a store that mostly serves customers who were already yours is a bad decade, and no amount of post-opening analysis un-signs the lease.

How to run a product cannibalization analysis

A product cannibalization analysis follows the customers who bought the existing item before the launch and records where their spending went afterward: to the new item, back to the old one, split across both, or out of the category. That destination mix, measured against a control cohort in stores without the new item, is the transfer estimate.

The cohort comes from household-level data: a loyalty program, or a household panel from a provider such as NielsenIQ or Circana. Freeze the household list before launch, then read the destinations by margin rather than units. A buyer who moves to a cheaper pack size is transfer that costs money; a buyer who moves to a premium version is the trade a line extension exists to make. One blended switching percentage hides the difference, and the difference is what changes the assortment decision.

Without household data, the fallback is cross-price elasticity from store-week sales: the percentage change in the existing item's unit sales divided by the percentage change in the new item's price. A positive value means the two items substitute for each other, so promotions on the new item pull from the old one. A value near zero means they serve different needs. It needs price variation, so it reads best after at least one promotion cycle.

Product teams also get a tool store teams never have: the staggered rollout. When the new item reaches some stores weeks before others, the late stores are a ready-made control set, and comparing the existing item's sales across the two groups isolates the launch from the season. A store network cannot open the same location in half its markets first, which is why store analysis leans on customer origin instead. The product-side rate formulas are in the cannibalization rate guide.

What triggers cannibalization

Cannibalization is triggered by four usual causes: pricing that invites a trade-down, assortment overlap, too little differentiation between two units, and proximity without a plan. Knowing which one is in play tells an analyst where to look before the sales numbers arrive, and which of the four produced the loss once they do.

• Pricing: a new item priced just below a higher-margin comparable invites a trade-down, and discounts do the same on a shorter clock.
• Assortment overlap: a line extension that replaces demand instead of creating it makes the older item redundant.
• Insufficient differentiation: two units, on a shelf or a map, that do the same job for the same customer split that customer rather than add one.
• Proximity without a plan: the most expensive version, because the lease is long and the overlap is baked in on day one.

The cost of missing one shows up in the closure column a year or two later. Grocery Outlet opened 42 stores in fiscal 2025 and then identified 36 with no viable path to sustained profitability, two dozen of them on the East Coast, or 30 percent of that region's store count. CEO Jason Potter told investors on the March 2026 earnings call that "it's clear now that we expanded too quickly."

Not every cause is a mistake, though. Procter & Gamble knew Tide would take sales from its own soap brands and launched it anyway, because the category shift was the point. The difference from the closure lists is that somebody priced the trade in advance.

The tolerable distance varies by brand, not only by geography. OpenStreetMap's September 2026 data shows that 137 of Manhattan's 166 mapped Starbucks sit within a quarter mile of another Starbucks, with a median distance of 0.13 miles to the nearest location. In that borough, the median gap among the 48 mapped McDonald's locations is 0.37 miles. Chick-fil-A's median is 0.44 miles, and only 2 of 16 restaurants have a twin within a quarter mile. That spacing shows how much overlap a brand has. It doesn't tell the committee what the overlap costs, which is the figure it needs.

Where cannibalization analysis goes wrong

Cannibalization analysis usually goes wrong in four places, and none of them involve the math: trade areas drawn as rings, a rate quoted against an undefined denominator, a before-and-after comparison with no control set, and an overlap estimate nobody can open. Each produces a confident number the next review cannot defend.

Ring-based trade areas. A three-mile radius is a circle drawn for convenience, and it lines up with real customer origins less often than teams expect. A VP of Real Estate at a multi-location veterinary group put the broader point this way: "We thought we understood what made a good site. GrowthFactor showed us our primary KPI was hiding the real story, and the variables we'd been ignoring were the ones that actually predicted revenue." If the trade area is wrong, every overlap number downstream of it is wrong too.

The wrong denominator. Cannibalization rate can be quoted against the new unit's sales or against the existing unit's prior base, and those produce very different-looking percentages from identical facts. Pick one, define it in writing, and make every site use the same one. Comparing a 15 percent from one convention against a 25 percent from another is how a portfolio review goes sideways.

No control set. The most common single defect: a before-and-after with no control credits your new store with the weather.

A number with no owner. An estimate that nobody can open is an estimate nobody will defend when a CFO pushes on it. If the analysis cannot show which inputs produced the overlap figure, the meeting turns into a debate about the tool rather than the site.

Measuring what actually happened after you open

Measuring cannibalization after a store opens means re-running the pre-opening analysis on real sales at 90 days and again at four quarters, with the affected set and control set that were frozen before launch. The gap between the projected and realized rates is the correction the next forecast needs, and it is the step most expansion teams skip.

At each review, pull the same affected set and the same control set you froze before the opening, and compute the realized cannibalization rate the same way you computed the projected one. If you projected 12 percent and got 21, the gap tells you something specific about how your model reads that market type, and you can correct it before the next five sites.

Strategic cannibalization: when the trade is worth making

Strategic cannibalization is overlap a brand accepts on purpose, and it takes four recognizable forms: replacing an aging product with a better one, blocking a competitor from a corner the brand would rather not concede, moving customers to a higher-margin format, or reaching a customer segment the existing unit cannot serve.

Steve Jobs is widely quoted as saying, "If you don't cannibalize yourself, someone else will." What separates strategic from accidental is whether the number was on the table when the decision got made, not the size of the hit. A 20 percent overlap priced into the business case in advance is a trade; the same 20 percent discovered in a quarterly review is a miss.

From analysis to decision

Cannibalization analysis pays only when the overlap number changes what gets approved, which means it has to arrive with the site packet rather than after the committee meets. GrowthFactor models overlap between a proposed site and every existing unit as part of scoring the location, with each input visible and adjustable rather than buried behind a score.

Cavender's Western Wear cut analyst time per site evaluation in half this way and avoided three locations the analysis flagged, worth roughly $2M. Across customers, roughly 80 percent fewer underperforming locations show up once the workflow is in place (JAN2026 customer survey).

The same data runs the other direction too. A high cannibalization rate isn't automatically a dealbreaker, and some operators find that closing an underperforming unit lifts network profitability as sales consolidate into stores that convert them better. That is portfolio optimization working off the overlap figures you already built, alongside site selection and retail location analysis that read network effects, and demand forecasting that counts what a new unit takes as well as what it adds.

GrowthFactor checks every site against your own stores and shows why. Plans run from Enterprise for teams evaluating sites week to week up to Labs for network-wide optimization, and our retail expansion planning software covers the workflow around the decision.

Ready to put the overlap number on the table before the committee meets? See how expansion teams use GrowthFactor.

Frequently Asked Questions about retail cannibalization analysis