One of the topics we’re often engaged to explore for our clients relates to cannibalisation. Put simply, cannibalisation references a situation when sales at new locations are made to customers who would otherwise have shopped at one of the retailers’ other locations. So rather than attract new customers, some sales are being made to the same customer but at a different location.
It’s easy to understand why the company CFO is often terrified by cannibalisation. The net margin on sales being lost from the existing locations will be relatively high compared to the net margin on the first sales being made at the new location. So, a loss of even 10% of sales at one of the existing locations might account for a 50% drop in net profit – all the while the new location can be struggling to break even. That’s enough to give any CFO a headache!
But is this the disaster it’s often made out to be? Well, perhaps but that’s certainly not always going to be the case. Many of the most successful and profitable retailers across the globe recognise that an element of cannibalisation will be inevitable if their business is to achieve its sales potential. So perhaps the better question, rather than how to avoid local area cannibalisation, is how to avoid unexpected and excessive cannibalisation?
Understanding Trade Catchments
The first step in predicting cannibalisation is understanding the trading catchment of existing stores and across the network. This activity is for some retailers quite simple but for others, it can be complex.
Understanding and defining trade catchments for an existing store or network will never be as simple as saying the brand’s store catchment is a radius of 3km or a 5-minute travel time catchment. Surprisingly, many retailers persist with this type of over-simplification and consequently, they’ll make network growth decisions which will sometimes deliver a financial return well short of expectations.
As a starting point, defining the catchment area for a store requires an understanding of what’s termed the ‘customer travel path’. We’re interested to understand the travel path the customer makes immediately prior to and immediately after visiting the retailer. Did the journey commence at home or work or somewhere else? Where is the customer going immediately after visiting the retailer? Is the customer immediately returning to home or work or is the visit to the retailer part of a trip to a shopping mall or another destination or is the ultimate destination some other location?
This ‘trip pattern analysis’ allows us to define the retailer’s location types which are the starting point for understanding and predicting cannibalisation.
Not all catchment areas across a retail network will be similar and understanding the various location types that exist will be important in order to accurately predict the effects of cannibalisation. Most, if not all retailers will have a variety of location types in their network.
Consider a fast food or café retailer. Some locations will be what we call ‘neighbourhood’ locations, others will be ‘worker’ locations and others may be ‘shopping mall’ or one of a number of speciality location type categories. By allocating all stores within a network to a location type we can begin to group together stores with similar trading catchments and better understand cannibalisation.
Neighbourhood locations have a predominantly residential customer base and are convenient for customers travelling to/from home when visiting the retailer’s location and they perhaps bundle the shopping activity with a visit to another retailer in the neighbourhood.
Worker locations tend to be convenient to the customers’ work address and the trip pattern is typically from the workplace to the retailer and back to work – these catchments are often quite small and reflect that customers are often time-pressured whilst at work. Consider for example how far you may travel from your office to buy a mid-morning cup of coffee. For most of us, the coffee shop will be very close to the office and is often within a walking distance of just a minute or two.
Shopping mall locations are different again, particularly for food and beverage retailers in large shopping malls where customers are visiting the food and beverage retailer as part of a shopping trip (rather than the food and beverage retailer being the primary reason for visiting the shopping mall). For large shopping mall locations, the catchment area doesn’t extend beyond the mall itself.
Understanding and defining location types is important not just because there tend to be similar catchment sizes amongst a retailers’ outlets in each location type. Perhaps more importantly, cannibalisation impacts will vary enormously depending on the location types of the proximate locations. For example, if a food and beverage retailer opens a new location in a large shopping mall that’s within say 1km of an existing neighbourhood location there’s likely to be little or no cannibalisation. If the same retailer was to open a new neighbourhood location within 1km of an existing neighbourhood location the cannibalisation is likely to be very high. But if the new and existing locations were considered worker locations in a CBD environment the cannibalisation is likely to be very low.
For large retail networks, it’s possible to undertake a piece of analysis to determine whether there’s any broad evidence of cannibalisation across the network. If so, that may be an indication that it’s time to stop building store penetration in some parts of the market. To determine systemic cannibalisation, we look at the sales performance of the retailers’ stores that have 2 or more alternate locations within a small area (depending on the retailer that might be 500m or 1km or 2km or a combination of many options). Once we normalise for other performance factors, we’re seeking to determine whether sales at locations with many nearby alternate stores are relatively low compared to stores that don’t have alternate stores within the same close area(s).
How Much Cannibalisation is Too Much?
The answer depends not just on the client but also on the maturity of the client’s business. New popular brands will go through what’s often referred to as a honeymoon phase. During this period the location trading catchments are very large compared to what might be considered normal for that sector.
I recall when Krispy Kreme opened their first Australian location at Penrith Panthers in Sydney’s outer western suburbs. There was a drive-through queue requiring barricades and traffic management personnel. There was a queue outside the front door in excess of 100m in length. The opening period was exceptionally busy and the catchment area over those first couple of months would have covered half of Sydney. Today the catchment area for a Krispy Kreme neighbourhood location is unlikely to be more than 10 minutes.
When a new brand emerges and begins to grow some cannibalisation will be likely – just as it was with Krispy Kreme. In these situations, there can be a compelling case to grow the business irrespective of the cannibalisation that may occur. The objective is to select high sales potential new locations with limited cannibalisation on the existing network.
If you’re interested to learn more about cannibalisation there’s some good news – this is a field in the location intelligence community that is relatively well understood and researched. Consultancies like GapMaps will be only too happy to discuss your concerns and build a research brief to provide the insights and strategies required by franchisors to successfully grow their businesses whilst accurately predicting and managing cannibalisation concerns.