We value your views. We will use them to add content to the e-newsletter and to take on board as suggestions for magazine features. To see your view posted here, please email the editor.

Archived views
Ronel Schoeman, Databases for retail uplift
Richard Roche, The times are a changing
Jeremy Saul, Holistic
Richard McCann, Friday's of Piccadilly
David Harrison, WSPS
Russell Loarridge, Firstwave Technologies
Alan Joen, Touchstone
John Whittingdale MP

Today's view

By Ronel Schoeman, business and analytical consultant, Database Group

The times are a changing...

Retailers can make databases pay, says Ronel Schoeman, business and analytical consultant, Database Group.

In the dotcom boom, many retailers became seriously concerned about the lean and mean e-retailers who were threatening to steal large swathes of business from under their noses. By the time of the dotcom crash, a number of leading retail brands were developing their own remote shopping channels in response to this perceived threat. One such was Iceland Home Shopping, which wisely set itself up in 2000 not as a dotcom pureplay, but as a multi-channel remote shopping venture, available through telephone, internet and digital interactive cable TV. 'Talking Food', Iceland's home-shopping catalogue now covers some 97% of the UK population.

The challenge:
Following its initial commercial success through to 2002, Iceland Home Shopping's next main objective was to improve both order volume and profitability from its catalogue mailing campaigns to registered customers. Up to this point, the catalogue had been mailed to customers who had recently bought through the service. The company now wished to find a means of understanding, and predicting, who would be stimulated into purchasing when they received the catalogue. In this way, Iceland Home Shopping aimed to systematically grow its return on investment from the catalogue mailing cycle. To achieve this objective, Iceland engaged The Database Group (DbG) to build a catalogue mailing scorecard to drive and refine its campaign selections.

Solution:
Initially, DbG devised a set of key questions relevant to Iceland's commercial objectives. These questions - essentially variations on "what are the characteristics of each spending group" - categorised customers by behaviour. Once classified in this way, DbG then tested all the main geo-demographic segmentation systems in order to see which one most sharply differentiated each group from the others. The best performing segmentation system was selected as the foundation - along with transactional information - to build a scorecard that could then be used to predict likely spending activity.

Iceland required a solution that would integrate with their existing systems, so DbG delivered an automated SAS model that was applied immediately and easily. Iceland run six campaign cycles per year and the scorecard is now used for each mailing selection. After each campaign cycle, the scorecard is refreshed by analysing actual response data and adjusting the scores accordingly. This adjustment is crucial as the customer base grows and develops - changing subtly in its characteristics as time goes on.

In short, Iceland are able to refine their selection process regularly, using actual response to increase the power of the predictive model.

Results:
The effectiveness of this initiative is best described through the results it is producing - a fourfold increase in response rate!

In summary, Iceland Home Shopping wanted to improve campaign productivity by moving from simple recency selections to an approach that more powerfully modelled people's propensity to spend. DbG provided the company with the commercial analysis and statistical modelling expertise to deliver a predictive scorecard solution that integrated easily within the Iceland IT environment. Mailing campaign effort is now more productively applied and wastage greatly reduced. Campaign return on investment has quadrupled. 'Talking Food' is now talking turkey!