Speech recognition helps retailer determine ROI

Each year, Australia’s largest independent car service and tyre retailer, Kmart Tyre & Auto Service (KTAS), spends millions in advertising and marketing.

However, it has only been recently that the retailer has finally been able to pinpoint the exact effectiveness of this spend, and the vehicle for enabling this has been speech recognition technology and its set of online analytical tools.

Tony Skallias, Kmart Tyre & Auto Service General Manager, Marketing, said, “About a year ago I was casting about for a replacement to our outdated telephone store locator system which relied on callers knowing the postcode of the store they wanted to be connected to.

“Given that we have over 250 stores across Australia, expecting customers to remember postcodes was a huge ask and in many instances simply didn’t work.”

It was recommended that Skallias consider a store locator from new-generation speech recognition vendor Inference Communications.

Unlike more traditional call locator systems, which ask for a postcode and route a caller to the relevant store, the Inference locator has in-built intelligence which enables callers to be connected without knowledge of postcodes, making it a more flexible and valuable business tool.

Intrinsic to the Inference store locator is a centralised business reporting system which automatically collects data about where calls are coming from and where they’ve been connected to, enabling users like KTAS to drill down to the exact nature of the calls and profile customers.

“What we got from Inference was more than just a clever store locator. We also scored a rigorous data collection system, capable of assisting us improve operational efficiency as well as streamline our marketing spend,” said Skallias.

Provided as an externally hosted service, the locator system has already done much to improve call traffic throughout the entire organisation.

Although the data collection system is yet to be rolled out at state level across the country, Skallias and his marketing team are already using the system to assist in streamlining their marketing spend.

“Because the system can aggregate where calls are coming from, we can now effectively measure channel performance - whether it’s online, white pages, catalogues, press, radio, TV, etc.

“This is particularly useful when we’re running an integrated marketing campaign where multiple communication channels are being used.

“Because each marketing channel is given its own phone number, whenever we run a marketing campaign and the phone begins ringing in response to the campaign, the speech recognition system automatically recognises which number has been called. Once the campaign is over we have a raw measure of where all the calls have come from - in essence, which of our marketing channels have proved most effective.”

Skallias says although the system has been in place for little less than a year, it is already helping KTAS improve its marketing ROI.

“Basically, we’re now better informed about which marketing channels work well and which don’t and why.”

Asked how this has changed KTAS’s marketing spend, he says it has dictated that the spend needs to be across multiple channels and weighted according to ROI.

“That said, we’ve also learned that while radio rarely results in an immediate call to action from listeners, it’s great for brand building and awareness and for supporting other marketing channels.”