cellpadding=0 cellspacing=0> | | |
One might think that furniture shoppers have a finite need for ottomans, typing tables and children's bunk beds, but at Rhodes Furniture, half of its customers put cash on the table — literally — twice within a two-month period.
This was the biggest revelation coming out of Rhodes' move from a flat file to a relational data system. But the change also allowed Rhodes to rank its customers into deciles based on their tendency to purchase, opening up marketing opportunities and cost-savings strategies.
The ability to skim the top deciles allowed Rhodes to maintain sales while paring mailing quantities by 30%, according to marketing director Tena Barnes. The Atlanta-based firm had been sending out between 300,000 and 400,000 12-page color circulars, and 400,000 to 600,000 postcards, letters and other direct mail pieces every month.
After reducing mailer circulation last fall, the company had enough money to augment other channels, such as direct response television and radio, newspaper advertisements, and a direct mail component to its year-old new mover program. Rhodes sends 50,000 to 70,000 mail pieces monthly to new movers.
But these mailings, geared to prospects, are an exception to Rhodes' marketing strategy. Prospect drops make up less than 10% of its mail efforts, according to Barnes.
“Historically prospects have not been a well-performing segment,” Barnes says.
The biggest data-driven change was Rhodes' repeat purchase program.
“We contact people who haven't come back in a six-month period. They are earmarked to receive three mailings, with each offer getting a little richer. This repeat program was a no-brainer once we saw the data,” Barnes says.
So what makes the new model effective? Nineteen variables taken from the chain's transaction files.
BeNow, a database marketing services provider in Wakefield, MA, tested appending more than two dozen demographic variables from Donnelley Marketing to Rhodes' customer list, but ended up not using any of them in its final model.
“Not because the data weren't good, but because knowing if someone is 27 is not the same as knowing if they bought a sofa,” says Andy Cutler, BeNow's chief strategy officer.
BeNow also offered weekly updates from the sales floor to the data file, rather than the monthly information Rhodes had been receiving. The increased frequency meant that the furniture store's “thank you for your purchase” drops were no longer hitting customers' mailboxes at the tail end of the 60-day repurchase window.
|