FOR FINANCE directors of large companies, rolling out major initiatives company-wide always comes with high costs and risks attached. Never has this been more relevant than now, with the eurozone crisis and renewed speculation about a double-dip recession heaping further pressure on British companies.
Despite this, many directors plough ahead with big investments based on manual financial analysis. But there are many problems with this approach. Much more can be gained through rigorous business testing using advanced analytical technology, with the potential to increase profits by as much as 5%.
If yours is a retail chain, refurbishing all your outlets could cost millions. The same applies if you’re a hotel group considering opening in a string of new locations.
While you’re hoping for increased profits and customer satisfaction, the gains could fall short of the implementation costs. Worse still, you could find disastrous, unexpected losses on your hands. So how can FDs accurately predict the outcome of proposed investments?
Most adopt a fairly methodical manual approach to exploring a hypothesis. A retail group might try a promotion in a small number of stores. A bank might change the layout of a few branches. Analysts will study spreadsheets of resulting data and produce a recommendation for or against rollout for the FD to review with the rest of the board.
The fundamental problem is that such trials are difficult to run in a statistically reliable manner. The findings are too often distorted by all the surrounding ‘noise’ factors, such as weather, transport issues and competitor activity. Misleading findings can give false confidence to a company to implement a harmful change, or to withhold a beneficial one.
The good news is technology is now available to cut through this noise and enable fully meaningful trials, even when working with small samples. The latest analytical systems will help ensure trials are scientifically set up, with carefully chosen test and control groups, and will adjust for the many misleading factors. Not only will this provide an accurate ‘yes/no’ answer to whether an initiative is worthwhile, it will also reveal how it can be tailored for optimal effect.
A restaurant chain, for example, may learn that expanding its serving staff will increase profits in city centres, but not in suburban areas. A bank may find that upgrading branches for business customers will increase profits, while upgrading consumer-focused branches will only increase costs. Hence, companies can adapt initiatives for optimum gain based on certain products, locations or customer segments.
Never before has this approach been more relevant than in the current climate. Shoppers are filling their baskets with promotional products, but which offers will make the smallest dent in retailers’ profits? Companies are working hard to minimise expenditure, but which costs can be safely trimmed and which will harm the business?
Testing can provide the answers and bring huge benefits. Our experience shows that a company-wide ‘test and learn’ approach can boost profit margins by as much as 5%.
Testing in action
Some examples should help bring this to life. Wawa, the US convenience store group, planned to introduce a new flatbread breakfast product. CFO Chris Gheysens admits the firm used to take chances with such decisions. In this case, it was a good-quality item – it was well received in focus groups, so why not roll it out?
Testing the decision using a software-based approach revealed a completely different picture. Gheysens learned that the new item would cannibalise sales of higher margin products and reduce overall profits. He killed off the launch.
Wawa also analysed its long-standing belief that extra cashiers at stores with small car parks would move customers along faster and so boost sales. And testing showed this was unfounded. It revealed, on the other hand, that more staff at stores in large urban areas would produce dramatically higher sales, so the firm increased headcount at 50 such outlets.
Gheysens confesses that, until this point, the finance team had never felt truly confident approving or rejecting investments. He believes that previous, manual trials were so clouded by misleading variables that the firm relied on emotion more than fact. All this has now changed.
Meanwhile at US grocery group Big Lots, CFO Joe Cooper used advanced analytics tools to test, among many other activities, the number of printed circulars distributed to its stores. He found the number could be significantly cut without harming sales and the company has consequently saved millions of dollars a year. Cooper believes company-wide testing has helped deliver 14 consecutive record quarters, in spite of the tough trading environment.
It’s no coincidence that these examples are from North America: many CFOs there have been undertaking advanced testing for years. We have worked closely with Subway, InterContinental Hotels and McDonald’s, and all have embedded scientific testing deeply in their cultures.
But the first UK companies are waking up to the benefits. Boots is an innovative example from the British high street. It recently applied advanced testing software to its trial of a store refurbishment programme to see how the technology would perform. The company found the approach invaluable as even deciding how many stores to include in the trial required scientific calculation. Too many stores would have been too costly; too few would have produced inconclusive results.
The process generated accurate demographic analysis, information about which elements of the store refits were most productive, and how quickly Boots would recoup its investment. This enabled the company to make informed decisions about how to move forward with its programme. Based on this, Boots has licensed testing technology for a further five years and is using the approach across its business.
Jewellery retailers H. Samuel and Ernest Jones have made a similar move. Following a detailed pilot to try out advanced testing tools, they have now taken a three-year licence to use them business-wide. They’re trialling marketing initiatives, store and staffing changes, and gaining greater insight into shopper behaviour in the process.
The pitfalls, then, of failing to run sufficiently rigorous trials can be huge when it comes to large investments. Of course, creating a rigorous, company-wide test-and-learn culture itself requires investment in technology, training and senior management attention. But the benefits are typically overwhelming.
Given today’s tough economic outlook, which large company FDs can afford to ignore the potential to minimise risk and achieve a 5% lift in profitability?
Jim Manzi is chair of Applied Predictive Technologies