The Perfect Store? – Co-op and Unilever use Virtual Stores to Design the ‘Perfect’ Neighbourhood Store

by Michael Letchford

So What is a Perfect Store Layout?

In 2003, as part of their ‘Partners for Growth’ programme, Unilever conducted extensive consumer research into shopping ‘missions’ in convenience stores. The study revealed that 67% of shoppers mostly used convenience stores for ‘complementary’ shopping and that their most important shopping ‘missions’ in these stores were Newsagent, Top Up, Food for Now and Need it Now. It was also obvious that the average convenience store layout did not efficiently support these shopping missions and, because shoppers were so focussed on completing these missions, the poor layout meant that they made fewer additional purchases than expected whilst in the store.

Unilever felt that they could address these shortfalls but needed a partner – a retailer who would collaborate with them to test their ideas on category adjacencies and zoning stores by shopper mission. We introduced them to United Co-op, a specialist in this format, who agreed to work with them to co-develop a more effective store and merchandising layout. Coincidentally,United Co-op were also reviewing their convenience store formats. Each store at that time was slightly different from stores opened before it, which had led to too much variation in layouts. Fifth Dimension acted as matchmaker in putting the two teams together so that they could both achieve their goals.

existing-co-op-bradwell-storeThe project team selected a typical local store, Bradwell in Stoke on Trent, and used a third party to conduct interviews with shoppers to compare their actual and perceived behaviour whilst in the store. Analysis showed that while the store was strong in Need It Now and Newsagent missions it significantly under-traded in Top Up, Food for Now and Meal for Tonight items.

Shopper Profile for the Convenience Store

Post visit store interviews at the Bradwell store revealed that;

  • 80% of shoppers live within a mile of the store
  • 64% of shoppers were female, 83% of whom shopped alone. 48% of these women were at least 45 years old, with 41% in the 25 to 45 age range
  • 62% of shoppers came from home and went home after visitng the store. 10% visited on their way home from work
  • 87% of all shoppers interviewed visited the store at least once a week with 73% visiting the store every 2 to 3 days
  • 38% of shoppers claimed items were purchased for themselves, 34% said they were for the family and 16% for themselves and their partners
  • 70% of shoppers pre-planned their visit to the store
  • 61% spent uder £5 per visit and 24% spent between £6 and £10
  • Most shoppers were purchasing for immediate or later that same day consumption

Types of Shopping Mission

The shoppers were visiting this store to fulfill these main missions;

  • 26% Top Up (fresh items 14%, cupboard items 9%, fridge items 2%, freezer items 1%)
  • 16% Food for Now – lunch, snack or drink
  • 15% Run Out (food items 8%, household 5%, ingredient 1%, other items 1%)
  • 15% Newsagent – cigarettes, newspapers, magazines or lottery
  • 8% Need it Now – purchasing items to complete a meal for the same evening
  • 6% Other – buying for others

Shopper Traffic Flow

A traffic flow analysis also revealed that 51% of shoppers entered the store and immediately turned left and walked down the first aisle of the store, 10% turned left into the second aisle, 13% turned into the third aisle and 12% headed straight for the tills. Of the shoppers visiting the first aisle, where fresh meat and produce were located, a large percentage turned right down the first cross-break rather than continuing up the first aisle – so less than 33% of customers shopped this aisle fully.

Customers’ first interactions were focussed on Newspapers (19%), Pasta and Sauces (9%), Produce (6%) and Bread and Rolls (5%) all of whch were located along the first aisle to the left when entering the store.


The diagram shows the general penetration around the existing store. From it you can see that the majority of movement was around the perimeter of the store, the confectionery aisle being the most visited. The Lunch item presentation, positioned just to the right of the entrance, performed very poorly as did the Chilled Wines area.

Many areas of the store were under visited. Analysis showed that areas of poor shopper penetration correlated well with poorer spends in the categories that comprised the underperforming mission baskets i.e. Top Up, Food for Now and Meal for Tonight.

Further analysis showed that there was no obvious pattern of movement around the store. Shoppers’ paths diverged early in the visit, people turning left immediately on entering the store or moving directly to the tills. Penetration was spread across the store with the cross aisle receiving over half of the traffic.The back of the store was eventually visited only by a third of people. The obvious conclusion was that the existing layout, category adjacencies and merchandising was obstructing the efficient completion of the four main missions.

Shoppers spent, on average, 4 minutes 45 seconds in the store and interacted with various categories for an average of only 2 minutes and 50 seconds of that time. 73% of shoppers spent under 6 minutes in the store. The Top Up shopper spent the longest time in the store, while the Newsagent mission shopper had the shortest visit. Top Up shoppers had the largest basket with an average of 4.5 items, the Newsagent shopper having the smallest with only 2 items.

Solving the Problem?

It was obvious that the existing format needed radical improvement and so a great deal of effort and careful thought was invested in re-planning and developing a proposed Perfect Store layout:-


Design changes were focussed on improving the customer flow and penetration around the store. It was discovered that shoppers did not purchase from the store’s gondola end promotions, so the Perfect Store layout was redesigned to avoid traditional gondola ends and instead have staggered aisle gangways so that promotions could be sited opposite an aisle break, forcing shoppers to directly face the promotions.

Aisles were reworked to improve customer flow and ‘break-up’ the regularity of the existing typical store layout. Categories were relocated; for example, Confectionery was moved to the fourth aisle to pull customers further in-store. Category adjacencies were re-worked to better facilitate shopper missions. The store was ‘de-cluttered’; floor standing displays relocated to improve access and an ATM was introduced, in line with customer feedback from the study, to occupy the underperforming area to the right of the entrance.

Testing Shopper Reaction

proposed-co-op-unilever-bradwell-virtual-storeOur brief was then to build an exact replica of the Bradwell Store, in our Virtual Store research technology, as a control, together with a completely revamped Perfect Store layout incorporating all the proposed innovations. Our Virtual Bradwell Store is shown on the right – from the outside. We were asked to test these models on the Bradwell shoppers to determine their reactions to the changes.

During the field work, shoppers were recruited from outside the store and offered the opportunity to participate in the study; they were asked why they were visiting the store, to determine which mission they were on. The quota was based on the shoppers filmed in the third party study earlier in the process.

The embedded movie shows a brief overview of the virtual test. It may take a while to download, so please be patient.

The Virtual Store

perfect-store-virtual-store-big-night-in-end-cap-dsiplay-200Our testing centre was set up in another part of the existing store. 550 respondents were invited to participate in the tests. They were introduced to the study and, after some familiarisation, invited to shop in the Virtual Store as they would do normally. 50% saw the control store (the existing Bradwell layout) and 50% the Perfect Store layout. All shoppers began their visit from outside the relevant Virtual Store.


The interior of the proposed Perfect Store layout is shown above


This view shows the proposed mid-store cross-aisle break that presents one of the new promotional areas face on to the shopper.

Overview of the Virtual Store Study Results

The results of the study confirmed that shoppers did indeed shop by mission. The control Virtual Store shoppers’ overall shopping trip times compared well with the average times observed in the real store in the earlier study, and so provided a good foundation for comparing the Perfect layout Virtual tests.

Shopper flow and penetration in the control Virtual Store also validated well with the real store tests, being concentrated in both cases in aisle 1, 2 and 3. The average number of products and categories purchased also validated well with the real store tests.

improved-shopper-flowThe two diagrams show the shopper flow improvements achieved in the Perfect Store layout (top right) versus the current store (top left).

improved-shopper-penetrationThese diagrams show the customer penetration improvements achieved in the Perfect layout (bottom right) versus the current one (bottom left). Marked improvements in penetration can be seen throughout the store but particularly in the bottom right hand corner, centre of store and upper right corner.

The Perfect Store layout was seen as far easier to shop, the in store times comparing well with those of the control store layout in all missions except for Food for Now, which was slightly lomger than before.

The combined impact of the layout, category adjacency and visual merchandising changes resulted in a larger basket spends in less time in-store. Analysis also showed that shoppers purchased as much from the new in-aisle promotional units as they did from the gondola ends. Interestingly, there were far fewer promotional displays in the Perfect Store than in the original layout, and so the average spend per promotional display was higher in the redesigned store tests.

Real Store Implementations – a Perfect Success?

perfect-store-results-the-grocer-0306-2006-200Following the studies, the Co-op refitted the real Bradwell store, incorporating all the innovations successfully tested in the Virtual Store tests. The store enjoyed considerably enhanced takings over and above comparative refits of other very similar stores. Bradwell’s turnover increased 9% in the weeks following its refurbishment. Average weekly sales have now stabilised 8.9% up overall. Average spend per customer is 9% higher and the store’s gross profits have increased by 9.7%.

The changes proven at Bradwell were then rolled out to many other similar stores across the country – with dramatic effect.
The redeveloped stores saw similar improvements in overall sales. After completion, during weeks five to twelve, sales were up 8% on average, and subsequently stabilised at a 5% overall increase. The changes also improved the completion of the four main missions, as well as increasing the value of the mission baskets. Average weekly sales in the Meal for Tonight mission rose by 27%, that of the Top Up mission by 9% and for the Food for Now mission by 8%. Perfection indeed!

Virtual Stores – A Real Payoff

It was crystal clear, to everyone involved in this series of studies, that store redevelopment is a costly, risky and time consuming business. Doing this kind of research and piloting potentially successful changes in real stores is lengthy, expensive and difficult. By testing in a Virtual Store environment, which fully validates with real life, the chances of success increase dramatically and the difficulties become manageable. The results speak for themselves.

For interested readers, The Grocer periodical followed the project and the text of the two articles it published are here;

The Business of Perfection – Text.
The original Grocer article can be found here : The Business of Perfection


The follow on article – In Search of Perfection – Text
The original Grocer article can be found here : In Search of Perfection

Unilever’s Partners for Growth website can be found here : Unilever : Partners for Growth

{ 3 comments… read them below or add one }

tracy June 16, 2010 at 7:30 pm

could you let me know what software the virtual store layout was created in? thanks!


Michael Letchford October 20, 2010 at 6:25 am

Hello Tracy,

The project you are asking about was created in a virtual shopping environment called ‘Discovery’, which is an application designed and built by Fifth Dimension Ltd. Since ‘Discovery’, Fifth Dimension has created new and more advanced technologies for Virtual Store research and Shopper Insights work. The most advanced of these is currently ‘Revelation’, which enables Fifth Dimension to create the largest Virtual Stores in the world and offer a fully multi-dimensional shopping experience within the Virtual Store. You can reach Fifth Dimension from their web site at



fauzan muda July 6, 2010 at 8:12 pm

this info is suitable with my project. Store Profiling. amazing


Leave a Comment