When the route to completing an online sale is long and complex, it can be difficult to know what exactly clinched the deal, as Matthew Valentine discovers
Department store pioneer John Wanamaker, who opened his first outlet in 19th-century Philadelphia, is widely credited with coining the popular adage: “Half the money I spend on advertising is wasted; the trouble is I don’t know which half.”
Tim Denison, director of retail intelligence at Ipsos Retail Performance, may recognise Wanamaker’s dilemma. “Leading retailers are moving towards being truly multichannel,” says Dr Denison. “Many of them are now embarked on consolidating separate delivery operations into single functions and single teams, which is a difficult nettle to grasp. The disciplines of and approaches to online retailing and conventional retailing are very different to one another, demanding separate skills. The need for consolidation seems obvious, but the path to success is challenging.
“Inevitably the transition to multichannel throws up problems that were not problems before now. Sales attribution is one of these. In a bricks-and-mortar world, unit sales and value per store, footfall and conversion rates are very straightforward metrics and ones that are fundamental both to performance evaluation and remuneration. How though can they accommodate e-retailing and m-retailing, when shopping behaviour regularly involves decision-making including more than one channel – researching online and buying in-store, for example?
We require a new mindset and fresh approach that is less dependent on geography
“Some retailers have sought to attribute online sales to nearest stores, but this has obvious flaws. Many continue to report them separately, but this is equally unacceptable, especially when you start to consider how to deal with returns and so on.
“In truth, believing that the conventional architecture for sales attribution can accommodate the multichannel format is misguided and certainly suboptimal. We require a new mindset and fresh approach that is less dependent on geography. Such are the influences on decision-making and purchasing behaviour that attributing sales back to stores is losing its former relevance. Perhaps, indeed, sales attribution per se has had its day.”
Some retail sectors have paths to online purchase that are among the longest on the web. Choosing an expensive item can take from 20 to 25 days, meaning the sales funnel customers negotiate on their path to a purchase is long and complicated. Last-click rewards do not acknowledge this.
More effective measurement and analysis of the customer path will let retailers assign revenues and use marketing budgets more effectively. But measuring the path to purchase is notoriously difficult, especially when it can include in-store comparisons, both in person and on mobile devices.
Helen Southgate, BSkyB online marketing controller, strategy and planning, who was formerly chairwoman of the Internet Advertising Bureau’s Affiliate Marketing Council, says, despite the wealth of data, understanding the effectiveness of online advertising will never be an exact science. “It’s like TV ads in that sense,” she says.
However, the trail left by online shoppers is significant. If a customer makes a purchase, retailers can obtain data, made anonymous, on what kind of sites the customer has visited during the preceding cookie period. This period varies by site operator but might typically enable the retailer to access data on which sites have been clicked on during the past 30 days and page impressions from the previous 48 hours.
By identifying, through analysis of thousands of transactions, how much each channel contributes to a sale, retailers can establish a tailored payment structure to reward search engine optimisation, affiliate marketing, display advertising and so on.