How to Make Data Smart and Machines Creative
Nicolle PangisSeptember 13, 2016
Originally featured in Campaign as part of the road to DMEXCO series.
In an already complex industry, we tend to overcomplicate. We all know the running joke that we put 3-letter acronyms on everything (and anything.) But when it comes to big data and smart data – and all the technology in between – we’re right to acknowledge the complexity. The truth is that the proper application of big data is extremely difficult and throwing more and more bodies at the problem is not the solution. At Xaxis, we believe that machine learning offers the missing piece of our industry data puzzle.
The Massive Effort of Data
All companies have data – but they don’t necessarily use it in the most efficient way. Storing, extracting, and applying data in real-time is a massive effort, and if there’s a delay in time to market, an ad may no longer be applicable to a user. We still have some way to go to ensure audience data is optimally leveraged in the creative and the way dynamic elements of the creative are leveraged on campaigns. But, a focus on efficiently using our data to boost brand allegiance and eliminate waste must become the standard of operation.
Shifting Towards Machine Intelligence
Creative and technology can talk to each other better than they did a couple of years ago, and they continue to come together even more closely. Part of the solution is machine intelligence. The majority of the digital ecosystem still manually manages small components of campaigns, including repetitive optimisation tasks that can and should be managed in real-time by machines.
At Xaxis we have our own machine learning technology called Co-Pilot, which helps us establish the best means of trading. In less than seven months, Co-Pilot has run over 5000 campaigns, and boosted client viewability KPIs by 30% in the US. It also generated a 50% lift in trader efficiency. This frees up our operational teams to concentrate more on campaign strategy – where their talents truly lie.
Continued investment in data science feeds the possibility of machine learning becoming more prolific. Artificial Intelligence will never replace the capabilities of the operational teams and the data scientists – a balance between AI and human intelligence must be maintained. But, data science roles will become increasingly vital to organisations to inform machine learning.
This is not a simple evolution for our industry. It is one that will take time and will inherently change job functions, workflows, and how creatives are managed in digital. These three legs of the stool: the components of the creative; machine learning/artificial intelligence, and the smart application of big data will create the next generation of the digital ecosystem. By nature, these types of changes will create an important, but complicated transition into the next evolution of our industry.