In association with Skyrise Intelligence.
What is perishable data, and how prevalent has it become over the past 18 months?
At its core, perishable data is data that loses value over time, and should be acted on swiftly to gain the most value. Having worked in analytics and data for almost twenty years, I’ve seen an explosion in what was initially labelled ‘big data.’
Using high quality data to determine tactics and strategies is nothing new, but the availability of datasets to describe complex subjects such as population behaviour, has grown exponentially. Never before has the need for accurate and up-to-date insight been more important
How big of an issue has it become for the ad industry, and what challenges does it pose to marketers?
A good marketing plan typically draws on a range of data sources to learn as much as possible about a target audience. This could range from mobile location data, which might be based on activity during the previous week, right through to census data that might be anything up to a decade old.
Previously, a strategic plan based on the latest, best-quality data available could be reasonably expected to remain valid for at least six months to a year. The pandemic has changed that, as it has led to changes in behaviours. Some are obvious, like how we work, communicate, and shop. Less obvious are the subtle behavioural changes among different demographics on things like propensity to buy, the sort of content they are consuming, or how and when they are consuming it.
The shelf-life of marketing decisions has shortened significantly. If advertisers make media plans and target customers based on data that's even a few months old, then their ad spend will end up being wasted due to data inaccuracies.
How have marketers adapted and what else can they do to ensure that they make full use of their data?
Progressive brands and marketers are keen to understand who their audience is today, rather than who they were 6 months ago. They want to understand how they are interacting with different digital channels and inventory - are they watching videos or listening to podcasts? They want to understand where the audience is, not simply in terms of where they live, but also how they move around - how many are going out? How many are commuting? How many are visiting public places? Where are the eyeballs I need to reach?
To get that level of dynamic detail, marketers need access to custom audiences built on aggregated statistical analysis patterns.
What strategies or techniques are available to help businesses to decipher the perishability of data and action it accordingly?
Marketing teams using new models based on old data are likely to provide inaccurate results. To hone their insights, they will need to take a broad approach to data collection and application. Marketers who are gathering data on behavioural trends and location-based insights will find themselves at a competitive advantage, as we head towards a privacy-first web. Overlaying third-party analytics on their customers and competitors can then be used to compare and contrast and, ultimately, complement their existing in-house customer data.
Mobile network data is one such dynamic data source that can be incorporated by marketing teams. With that insight, marketers can identify increases in demand, as well as where newly-acquired customers are coming from. They can analyse further, for example, by looking at which existing customers on their own database have increased spend, which ones have lapsed, and – more intuitively – they can analyse market share, to identify which competitors those lapsed customers are now using.
With an old, historic data model, a marketer in fashion retail we work with could only tell how many customers it was gaining or losing. By incorporating Skyrise Intelligence, we were able to work with the company's agency to show changes in their competitors’ web traffic alongside the overall market opportunity. The analysis showed that a significant number of customers gained during lockdown came from more expensive boutique style retailers, while those they were losing were going to cheaper, fast-fashion alternatives. On the basis of this information, the marketing team changed acquisition and retention campaigns. They used higher-end offerings in adverts and emails to the luxury customer segments who were transitioning from small specialists while pushing bargain-related messages and merchandise to the value-oriented customers who were at greater risk of churn.
What are the benefits of avoiding perishable data?
In simple terms, avoiding perishable data gives marketers more accurate and relevant targeting opportunities. Today, advertisers need continually refreshed data from a variety of sources and at a far more detailed level. Perishable data can be a powerful source of intelligence, as long as it is updated rapidly. It is difficult for companies to get that level of insight if they are relying on their own internally derived customer data. Most modelling tools they are using were not built to handle vast volumes of data which changes dynamically.
Rather than avoiding perishable data, marketers should be aware of the provenance of the data sources and how regularly they are updated. All data is perishable to a degree and highly perishable data can be a powerful source of actionable intelligence, as long as it is updated regularly, analysed rapidly and implemented immediately.
How will changes to the way data can be collected and processed affect marketers’ ability to leverage perishable data fully?
ATS London 2021 will take place on 3rd-4th November at Lee Valley VeloPark. Tickets and further details are available via the ATS London 2021 events hub.