Social data represents the largest accessible database in our history. But without historical context, the insights obtained can be misleading and the trends distorted.
Everything that happens online is in a historical context.
An opinion can be influenced by seasonality, world events, the cultural climate or even a meme become viral. Without fully understanding the historical context of a post, analysts will find it difficult to derive real conclusions from the data gathered.
Imagine that you are in charge of a marketing campaign. You discover that your last campaign generated two million online entries, 60% of which are positive. This seems like good news, but is it really?
Looking at the historical data, you may find that this figure is actually 15% lower than your brand’s average results, that influencers were more committed to previous campaigns and that it was paid advertising that Rather than engaging content.
Without historical context, you have to consider different assumptions. These assumptions ultimately give rise to deceptive trends, which distort the results and can harm your insights.
It’s inevitable. If your data is used to make recommendations, predictions or actions, you need historical data to validate your findings. Here are three reasons why you need historical data to effectively analyze social media:
Getting insights fast
The advantages of historical data Seasonality: the example of Unilever and Ben & Jerry’s
A Brandwatch customer, one of the most well known consumer goods producers, relies on historical data. Why ? Because the teams responsible for consumer studies and industry professionals want reliable insights. In particular, they use historical data to study seasonality.
While important cyclical events, such as holidays or seasons, can produce predictable results, determining the effects of seasonality is not always so simple. Indeed, most brands adapt the trend in their own way.
Take, for example, the ice industry. Sales increase in summer and fall in winter. The reason is obvious, even a child could understand this phenomenon thanks to its implacable logic: when it is hot, consumers want to refresh themselves by eating ice cream.
It would be easy to infer that this applies to all types of ice and that your brand is following the trend. But Unilever put this hypothesis to the test.
The group’s teams went back several years and collected all the references where consumers expressed their desire to buy a Ben & Jerry’s ice cream.
Surprisingly, these data did not increase in summer and did not drop in winter as might be expected. On the contrary, Ben & Jerry’s mentions increased continuously throughout the year, regardless of season.
The usual seasonal fluctuation had no effect on the brand.
The team analyzed these historical data and compared them to the meteorological data. The seasons did not affect conversations, but what about the weather?
Their findings contradicted all previous assumptions about ice consumption.
Over the years studied, a rainy weather was systematically accompanied by a soaring number of people mentioning the desire to buy a bin of Jer & Jerry’s ice.
The rain, not the sun, made them want to buy ice.
This information was totally inconsistent with what was suggested by the industry data (and wisdom). Historical data from social networks revealed that Ben & Jerry’s and other competing brands did not see their ratings particularly rising during the summer
This insight was reliable. This trend has been confirmed and validated by years of historical references which prove it to be consistent. This phenomenon was repeated again and again, every time it rained.
The historical data gave Unilever an insight that prompted the brand to change its online advertising marketing strategies. Instead of increasing its marketing budget in summer – like its competitors – the group launched online advertisements when the weather forecast predicted rain. Thus, Unilever has optimized its marketing budget thanks to an intelligent analysis of the behavior of online consumers and a genuine knowledge of the market.
To find useful insights, you need historical data. But a simple glimpse like that can offer Google Trends is not enough. You need a scalable tool that you can divide and segment in order to provide answers to the crucial questions that arise every day the big brands.
You will also need data that spans three years. A year or two is not enough for a reliable verification of the seasonality and durability of a trend. After all, a mention may be an accident, two a coincidence, but three similar statements undoubtedly represent a true trend.
Online discussions, social networks, blogs, forums or consumer review sites generate a large number of on-line entries that allow the accumulation of very large amounts of historical data.
Paying for online mentions is not the most economical solution if you want to evaluate different campaigns, product launches, or even the reputation of your brand from a historical point of view. Demonstrating flexibility for unlimited historical data allows brands to quickly make adjustments and, if necessary, modify their marketing budgets in real time, correct a campaign or target a more relevant audience.
Getting insights fast
And finally, you will want quick results. Research is a repetitive process, a constant back and forth between the data and your assumptions. You will have to constantly repeat the process, constantly modifying and refining your research to find the right answers. A changing tool that will take hours to install will not work.
Brands who want to analyze social media effectively need historical data for many other reasons. But the first three that come to mind are seasonality, flexibility and speed. Using powerful technology to harvest historical data, brands can not only extract important insights from consumers faster, but they can also answer questions that arise in an organic way.