Anthony Hess Client Strategist , MiQ
Kaltrina Brahimi Global Analytics Product Manager, MiQ
Anthony and Kaltrina started off the workshop with asking two simple questions. One, how many of us in the audience worked for businesses that utilised data to drive their decisions. Approximately, 80% of the people from the audience raised their hands in response to this question. Two, is the brands using this data to maximise the potential of what they did. Only one person responded to this question and even then, she said maybe. This set the premise for why a workshop that helped brands connect their data, was essential.
This lead to a hypothesis that the speakers presented. They believed that brands were wasting millions of pounds because of disconnected data and through the talk, they wanted to achieve three key objectives, first, how to unlock your businesses data potential. Second, how to make sense of the data and finally, third, how you can turn what the data is saying into something actionable. To explain further, they talked about Netflix’s million dollar algorithm. Netflix posed a challenge to the data science community and promised a million dollars for an algorithm that would improve the recommendations that the platform would give to it users. It took the winner of the challenge three long years to come up with the winning algorithm but it prompted another question, what if Netflix had decided to look at the data in a more creative way. Instead of looking at data from the preexisting set, what if they had decided to look at data from the local cinema chains or non-traditional partnerships such as linear TV? This example explained why sitting on a large potion of data wasn’t enough to bridge the gap. You need to extract value from the data.
‘The answers to your business challenges have been hiding in your data, we just haven’t been looking at it in the right ways’
From a marketers point of view, there are various data dimensions available. Behavioural data is the most common and is the one thats used everyday to make decisions in the business. However, this is just the first layer of any data strategy. In order to engage with an audience on a different level or to get a full picture, marketers need to look at macro data dimension they have available to them. This allows them to answer questions like whats the emotional sentiment the audience has towards their brand right now or how is something like the weather affecting how they interact with the brand. It is essential to have variety in terms of data dimensions and having a strategic point of view while looking at the data.
Storytelling is a buzzword around the industry at the moment and speaking in relation to data, data can tell you a story about your customers that you might not be aware of, but enabling data connections effectively, enables you to tell better stories and engage with your customers.
Citing various examples, Hess and Kaltrina convinced the audience of the importance of data connectivity. Now came the question, how do you actually get data to connect?
Intensity and complexity of data are the challenges to any data scientist. The answer to these challenges are Big data and distributed computing. Three core elements to any big data set that need to be looked at while building your data dimensions and data strategy. First, volume, that is the sheer density of the data the clients have. Second, velocity at which the data arrives to you and last, variety. In summary, a business in order to utilise data should think about three things. How can I automate the way I collect data? Query automation and optimisation, that allows a business to ask questions and get answers from its data in real time. And lastly, distributed computing that allows a brand to be dynamic and flexible.
Using data in this form allows a brand to not only have incremental growth, it changes the very nature of a business and how they’re actually able to empower their planners.
Coming back to the Netflix algorithm challenge, the winning algorithm was never implemented. Instead, Netflix decided to go from a 5-star rating method to a thumbs up or thumbs down method and started looking at consumption data instead of looking at reviews. They shifted their data point and this idea of identifying the correct metric is essential when it comes to data connectivity.
Three parting recommendations given by them were first, start off with the right objectives. Second, think big and different and third, focus on the outcomes.
For more content like this, be sure to check out AWLearn, Advertising Week’s year-round, video-on-demand continuing education program, which combines the best moments and the brightest minds from the global stages of Advertising Week.