To get the full picture of how their actions influence revenue, marketers must bring together online and offline channels as well as trade promotions and external factors.
Some of that data like clicks and open rates, is instantaneous. Finding out how many people saw your TV ad may take a little longer while, if you’re a distributor, sales data belongs to the retailer.
- Cover Story: Which-50 Investigates: Mastering Predictive Analytics Will Enable Predictive Engagement
- Read More: A peek into the future – Identifying the benefits of predictive engagement
To address those challenges, Lego Australia worked with Accenture two years ago on a 12-week pilot to collate data and build models that would provide insight into the impact of the toymaker’s marketing activity on sales.
The case study specifically looked at the connection between Lego’s above-the-line and below-the-line activity on sales, establishing the link between marketing and trade activities against product sales volume across product themes, retailers and seasons.
The ultimate goal was to understand the return on their marketing spend and evaluate how they could budget differently next time.
“Marketers are no longer just working on gut feel. They’ve got scientific data that they can make decisions on and actually improve the effectiveness of what they’re trying to do,” says Amit Bansal, Accenture’s Managing Director, Applied Intelligence Leader.
- Read more: Predictive engagement helping drive Australian Banks’ CX and marketing: Report
- Analytics is now table stakes but marketers are still struggling with the tools
- Businesses want data to predict the future, but struggle to act on the insights
Accenture collected, validated and analysed three years’ weekly data across different dimensions such as media, trade promotions, PR and Lego Club. The consultants also provided data templates so Lego could collect sales data from five of their retailers.
They also drew on the input from marketers and product managers on different seasons, as well as other factors such as trade campaigns and the release of movies like Star Wars which also impact sales activity.
“It’s one thing to build a model but then you need to also allow for seasonality and other all the other things that happen around a campaign,” says Bansal.
The results identified opportunities to rebalance marketing and make trade investments more cost-effective.
Broadly, in some themed products, sales were driven by trade promotions with parents buying Lego because it has been discounted. Other sales showed a more direct correlation with the launch of a movie and the hype around it.
“The key thing for them was using that information to better do their marketing plan for the next year,” Bansal said.
“Now they can actually do better campaign planning because the data is being collected, they’ve got models that run and they can get a better understanding for a certain demographic or a cluster of time how they should run the campaign.”
The project was designed to help plan activity more effectively, helping minimise the impact of the delay from offline activity.
“Whether it’s a trade promotion, discount or a TV campaign, there’s always a two-month lag before you start seeing the data come back in,” Bansal said.
“So that’s why we looked at three years to make sure we’re really seeing all the correlation.”
Bansal says he has observed a shift in the market, where the time between launching an offline marketing campaign and reporting back with the results is closing, putting pressure on marketers to optimise their campaigns across all channels more quickly.
“Obviously, if it’s an online ad you know by click, and that’s pretty immediate feedback, but it’s more of the offline media and it’s harder to get quickly, your newspapers, your billboards.”
That shift is being driven by agencies which are getting more data savvy and start-ups using AI to provide advertisers with information agencies are not able to provide. For example, startups developing services that take all the feed from radio or TV broadcast, and they can tell a brand when their ad ran and how long it ran for.
“So you’re going to start seeing the lag times come down dramatically. And then the clients will start reacting faster, because what you can do in the digital world, in social media and your websites, etc is now starting to translate it into the physical world.”