Data: platforms and grassroots
The data challenges to marketing teams continue to multiply, and are exacerbated when teams archaically and inefficiently report in a fragmented way, using spreadsheets and presentation decks. Reporting deficiencies can cripple a process that mandates urgency and actionability. To resolve this, Core invested three years in building a dynamic, device-neutral platform that is designed to bring all marketing and business data into one place, dramatically reducing manual reporting. It provides marketers with a clear view on campaign performance, while also providing the opportunity for more in-depth analytics. We call this platform Core 161 after the so-called “golden ratio” in mathematics.
Core 161 provides instant reports on the performance of online and offline media campaigns across search, online display, social, online video, television and radio. It breaks down campaign achievement by channel against all agreed metrics. In addition, marketers can review digital creative across campaigns and easily compare performance to previous years.
Core 161 enables essential business performance metrics, such as sales volume and revenue, to be fused with marketing activity, providing a single shared view across teams. In 2019, the platform will also provide clients with the opportunity to incorporate brand tracking data and run competitor expenditure analysis across all media.
With investment in online media approaching 50% of all advertising spend1, it is essential that marketers adopt a holistic approach to attribution that considers the granular contribution of online touchpoints, combined with the broader view of the marketing mix provided through econometric modelling. Therefore, this year, Core 161 will expand to support a measurement framework that combines both of the above attribution approaches with a budget optimiser to enable more effective planning. This will give marketers a deep understanding of the true impact of marketing investment, while improving decision-making and, ultimately, increasing the profitability of marketing.
A lot of the work carried out by marketing analytics specialists goes into assessing the channel-driven media impact on sales or how brand-building campaigns perform versus short-term activation. But that is a limited approach to drive greater efficiency and effectiveness. Analytics also provides the opportunity to work closely with other departments and show how marketing can use business intelligence to drive smarter ways of testing and investing.
Two specific areas that demand exploration are:
1. The role of weather in affecting sales performance.
2. The creation of local, town-level marketing communications strategies.
Weather effects on sales
Weather plays a role in how consumers shop in several ways; it affects their mood, impacts purchase decisions and can influence how much they are willing to spend. Yet, outside the obvious categories such as ice cream or soup, very few marketing communications briefs talk about weather, but its impacts are far-reaching. For example, a 1995 UK study found that as air temperature fell from 15°C to 10°C, sales of fresh fruit increased by 12%, whereas sales of cooked chicken fell by 12%2. While we can all speculate as to the reasons for this impact, the point is that a normally un-measured variable had a profound effect on a sales outcome. Ignoring weather as a variable may lead to mis-attribution and bad strategy.
Using advanced modelling techniques and an AI-powered recommender system, we can explore the impact that weather has on revenue and develop strategies to improve sales by region.
This heatmap shows the output of a project we conducted in 2018. It depicts the impact of rainfall on sales for one of our clients. The darker green regions experienced the greatest negative impact, while the white counties did not see a significant negative effect of rainfall.
The study found that it wasn’t the total amount of rainfall in general that mattered; as shown in the image (counter-intuitively) the ‘Sunny South East’ was found to suffer the strongest effect from rainfall. The hypothesis behind this finding is that, for this specific product, rainfall in a historically drier county has a greater negative impact on sales than the same level of rainfall in a typically ‘wet county’.
We can use this information to develop precision creative executions, at county level that are triggered by live weather information; the copy and tone of voice would vary depending on the historical level of rainfall in that region.
Identifying the top 50 towns in Ireland most likely to grow
Our next example is a hyper-local project to identify towns with the greatest growth potential for a specific product. Using a combination of geo-local analysis, coupled with demographic clustering techniques, we identified the top 50 towns from a longlist of 846 in the Republic of Ireland.
We analysed factors that were known to have an influence on local sales performance. These were: population make-up, the number of competitors available in each town and the presence of main amenities and services (i.e. train stations, hospitals, universities, shopping centres and supermarkets) within a 2.5 km radius. Once the list was established, we designed local media campaigns to drive greater penetration in these areas.
These examples of ‘grassroots’ analytics demonstrate some of the less obvious ways to work with data scientists to improve the performance of your marketing investment. However, before you invest in hyper-local analytics, you must work with an analyst to clearly define the business problem or opportunity and scope out the probability of analytics being able to find the incremental knowledge necessary to deliver meaningful business effects.
1. Core estimates (2019)
2. The weather sensitivity of the UK food retail and distribution industry, Agnew & Thorne (1995)