Print publishing finally meets 21st century with predictive analytics

Australia’s oldest and largest independent magazine distributor, Gordon and Gotch, has taken up IBM’s predictive analytics tools to forecast sales and demand.

Gordon and Gotch’s decision to use predictive analytics comes at a time when the print media industry is seen to be falling behind in a digital age, with magazine circulations seeing some decline.

“Distribution is not just about efficient logistics, it’s about being smarter about the magazines that publishers print and where retailers sell them. We wanted to show our clients that we could use the goldmine of data we had to change the way distributors are perceived in the industry and optimise sales with more intelligent allocations,” commented Gordon and Gotch’s general manager, David Hogan.

As Google analytics cannot be applied to print magazines like it can websites for insight, Gordon and Gotch instead draws insights from its wealth of sales data. This, along with many years of recording how many magazines were delivered and returned to each retailer, was used to train a predictive model.

Some numbers around this include in excess of 130 million copies over 3,000 mass media and special interest publications distributed each year, with five products selling every second.

The model or predictive analytics tool forecasts how many copies of a magazine is likely to sell at a particular retail store so that it can optimise this for its publishers.

Sixty-five per cent of Gordon and Gotch’s titles/issues are automatically allocated to retail stores though the predictive analytics tool, freeing up the publisher to further analyse sales performance, measure marketing activity effectiveness and make recommendations to clients for future growth.

“We decided to build a solution that would integrate with modern analytics tools to provide a much more powerful and flexible way of forecasting sales and allocating copies,” said Gordon and Gotch’s head of IT and business systems, Poppy Stamateas.

“When we first launched the new allocations solution, we were hoping for at least 50 per cent of titles/issues to have their allocations automatically finalised. In fact, the actual result was 65 to 70 per cent.”