We’ve seen the continued fragmentation of audiences and the ever-fierce battle for eyeballs and making sense of this changing landscape is challenging. We’re also seeing a ‘connected intelligence’ approach, where syndicated data solutions of all types are being directly integrated with advertisers’ first party data.
“Artificial intelligence will drive media targeting”
With AI, the accuracy of predictive modelling and the amount of data that can be processed will increase exponentially. It’s with this increased computing capability that AI will allow us to design highly tailored marketing campaigns that will be driven by the probability of an individual to convert, at a specific price point, through determined channels. When building any kind of model, not just propensity models, the effectiveness and accuracy of the model is dependent on the quality of the data going into it. AI will further help improve the modelling process by identifying errors within the datasets that feed these predictive models. All of this will lead to further marketing automation, which will establish the most effective time and place for delivering a specific type of message to the right consumer.
Programmatic systems (DMPs) will be able to quickly run through historical attitudinal and behavioral data to determine which ads perform best for different types of people throughout the consumer journey. Today we bucket individuals into predefined segments, but with the future use of AI, we will be able to continually optimise targeting. This will enable the real-time reassignment of individuals into new segments, based upon both their thoughts and feelings about a brand or product, as well as the behaviours they exhibit both online and offline.
“More sophisticated use of data, combined with analytics techniques and artificial intelligence (AI), will allow marketers to understand true ROI in real time”
Measuring and/or proving ROI is the #1 struggle for advertisers globally, with understanding omnichannel behaviour and optimising media investment close behind [Source: Kantar’s Getting Media Right, 2018]. The biggest dilemma is that there is still a strong divide in the way measurement is done for different investments. Traditional Market Mix Modelling and econometric models have the advantage of being able to give an overall perspective of marketing investment’s impact on sales. They were perfect when the media plans and marketing investments followed a fairly standard approach. In today’s ever-changing marketing environment, these models lack granularity and are not very sensitive to digital investments, so they are increasingly seen as non-actionable tools for optimisation. Multi-touch attribution approaches are detailed and provide great platforms for optimisation of digital spend, but are too campaign-specific and limited to digital conversion. What use is this for companies that still have 90% of their sales happening offline?
What will make a difference? Firstly, advancements in data management systems and analytics will help to solve the offline/online conversion dilemma, and also navigate the online and offline investments ecosystem. And all this in an always-on data ingestion environment.
Secondly, the increased use of AI/machine learning during the modeling process will expedite the process and allow greater sophistication in the number of modelling solutions used as well as real-time validation (i.e. what is actually happening in the market). Getting this right is important for demonstrating the ROI of marketing investments and will be a massive competitive advantage to help forecast ROI for our clients’ strategies.
The above article is based on an excerpt of key trends from our 12 Key Global Media Predictions for 2019.