These Tests Are Positive

Management system software helps Wawa with the "what ifs"


From June 2009

By D. Gail Fleenor

 Sponsored by

Making decisions that impact the bottom line can be intimidating unless you have the utmost confidence that those decisions are fact-based. To improve its decision-making process, a leading convenience store chain opted for a way to test the effects of proposed changes before committing resources.

Wawa operates more than 570 c-stores in Pennsylvania, New Jersey, Delaware, Maryland and Virginia. The privately-owned chain, based in Wawa, Pa., is well known for its dedication to being on the cutting edge of technology and innovative operations but, like many other companies, its analysts used to pore over reams of spreadsheets. Assumptions were then made based on the best interpretation of data.

The company basically "took chances" when considering investments, according to director of planning and analysis Adam Schall. "You can avoid making bad investments with a more precise analytical tool than a spreadsheet."

Distance from sister stores, new competitors in the market, road closures and bad weather are just a few of the site variables that Wawa must consider before investing in a new location. "With a spreadsheet, you can't solve for many different variables at one time and, when doing a test, it is impossible to filter out all this variable noise," Schall says. "You really cannot determine true cause and effect."

What Wawa required was technology with a statistical regression tool and, after reviewing the offerings of several companies, Schall chose Test & Learn management system software from Applied Predictive Technologies.

With Test & Learn, retailers "can test ideas on a limited-risk basis in a small number of stores before a company-wide rollout," says Patrick O'Reilly, president of Arlington, Va.-based APT. More than 40 Fortune 500 retailers, financial institutions and consumer goods companies currently use the suite to help make more informed decisions.

Wawa ran simulations in six different areas before contracting with APT: labor tests were conducted to determine the effects of adding personnel during certain shifts; new product rollouts were tested to understand cannibalization effects on other products and categories; and SKU rationalization tests were used to determine the optimum product mix. "We also used the tool to help evaluate historic cannibalization results of our most recent store openings," Schall says.

What Wawa learned as result of those tests wasn't "always consistent with our hypothesis, which is O.K.," Schall says. "We're making decisions using valuable and limited resources. If you learn from a test that you're not moving the needle, that can be just as important as learning that you are succeeding so that you won't go down the wrong path of investment."

The most powerful aspect of the APT solution, according to Schall, is the ability to look at a situation from the macro level initially and then use the software to slice and dice results for deeper insight. "We are able to segment stores based on attributes such as demographics, location and accessibility."

If the overall test of a new product isn't positive, Test & Learn allows for further analysis of those stores where the product did achieve success. "You may discover a ‘nugget' that you never would have found with a spreadsheet," Schall says. "In today's economy, to be successful and gain market share, retailers must tailor different offers to different markets. The segmentation tool provides the road map for these offers."

Item-specific analysis
Wawa is also using APT's Market Basket Analyzer to capture transaction data by product, evaluate possible marketing efforts and analyze the ability of various products to drive additional margin. "It allows you to look at every POS transaction and provides insight into what comprises each ticket," Schall says, which allows Wawa to see relationships between products across all categories and identify those items that bring customers into the store and the companion items that were purchased.

"You can re-merchandise offers based on product relationships you learned about with the tool," he says. "It confirms some category strategies and challenges the validity of others."

Web-based Test & Learn provides retailers with a greater level of comfort in the analytical process, which allows them to spend less time debating decisions in a variety of areas, according to O'Reilly.

"One issue we are seeing a lot right now is that almost every retailer has store remodels in its capital budget, but returns on investment are unclear," he says. Test & Learn can help retailers "determine incremental sales resulting from a remodel and allows retailers to see the dollar increase for each dollar spent."

Not business as usual
It also can help determine which store profile will get the best return from a remodel. "Retailers can see what additions create the value, such as lighting, and then be smarter about how they spend their remodel dollars," O'Reilly says.

Retailers can also use Test & Learn to determine the ROI of freestanding inserts, allowing them to "reshape their media budgets and be smarter about spending," he says.

SKU rationalization helps retailers determine the "economies of products and also what additional purchases each product spawns," O'Reilly says. Loyalty to particular products can be determined, as well.

"It's not business as usual any more," Schall says. "We cannot be doing things the same way that we always have. The competitive landscape is more difficult and more challenging than ever; the economy is burdensome. We have to work smarter by mining data better and using it to maximize time and capital investment. This tool allows us to do that."

© STORES Magazine
325 7th St NW ·Suite 1100 Washington DC 20004 · 202-626-8101

Contact Us | Subscriptions | Advertising

Reprints | Copyright 2009 | Privacy