Introducing Brandwatch Consumer Research
By Giles PalmerSep 17
We all know the adage “bigger is better,” but today we want to give credit to the old saying “less is more”.
In statistics, samples are used to represent a large set of data with more efficiency than combing through every piece of information. Sampling is widely used in music, manufacturing, and science to gather insights from large data sets quickly and accurately.
And now, from tomorrow, sampling is available in Brandwatch Analytics with Sample Queries, a new feature within the new Query Builder.
Sample Queries opens up Brandwatch Analytics for more use in large scale, real-time tracking and research and creates opportunities for new types of analysis that you may not have been able to do before. You can now get representative samples of both real-time and historical Query data without having to see every mention.
So you can build more Queries to find key trends and themes without missing important insights. What’s not to love?
– This too does sampling, but not quite like Brandwatch Analytics!
Sample Queries mean that users of Brandwatch can now explore broader topics and types of analysis that might not have been open to them before due to large volumes. Not having to pull back a total volume means that not only do you need to worry about mentions limits, but also means a smaller, easier to manage dataset.
Here’s just a few examples of how Sample Queries might come in handy:
In market research, Sample Queries allow research of very broad topics for influencer and topic analysis, such as tracking conversation about ‘haircare’ or the iPhone 6.
For example, a broad Query about iPhone 6 is estimated at over 3 million mentions per month. Take that down to 1% and we get a much more manageable 30k or so mentions.
Still enough to accurately analyse conversation topics, influencers and trends about the iPhone 6, without having to pull back every mention (your chosen sample size will depend on your needs and mentions limits, but as we calculate the sample in a statistically accurate way, your sample will always be representative of your total data set).
“ The sample query feature was great to have when doing topic analyses and event monitoring. Even with a partial view, we were able to get a comprehensive understanding of topics such as The World Cup and the NFL while still being able to project the total number of conversations.”
– Jennifer Loeb, Assistant Social Media Planner, ZenithOptimedia Group
Sample Queries also open up broad historical analysis. You can look far into the past for year on year analysis without having to manage huge amounts of data, such as seasonal products mentions, holiday shopping sentiment, and political campaign trends.
For example, analysing the Superbowl and comparing it year-on-year to analyse trends, topics or even comparing categories (such as different ads) can be achieved just as effectively with just a sample rather than retrieving the entire dataset.
Share of Voice or large brand comparisons
Share of Voice tracking is also easier with Sample Queries, especially for larger brands.
With a handful of Sample Queries, you can easily track Share of Voice of multiple large-volume brands and competitors where overall volumes aren’t important – just use the same % for each Query for an accurate comparison.
This won’t of course replace traditional brand tracking, where you’ll want every mention, but it can help compare larger brands or over longer date ranges for broader analysis.
How They Work
Sample Queries are easy to set up – you can just choose the percentage of results you want when setting up your Query.
We sample the data in a statistically accurate way, returning every nth tweet or webpage that matches your Query, in the order we find them (and the same goes for historical data, but backwards through our archive).
With Sample Queries, less is more. Now you know.