What Will AI, Big Data and Data Science Bring to Marketing in 2019?
Today, there are more data points about consumers, both collectively and individually, than ever before. More importantly, machine learning (ML) and artificial intelligence (AI) capabilities are becoming more sophisticated, increasing marketers’ abilities to scale to activate that data in new ways.
Many marketers know that these technologies will change how they do their job, but it’s still unclear exactly what will change and how fast. To get a glimpse of what the road ahead looks like, we asked a few of our data scientists to share what they think the biggest trends will be for the near future.
Scientists will improve their ability to teach computers to accurately comprehend language.
We will see significant advances this year in the way we model and process text. There is currently a heated debate going on within natural language processing (NLP) circles around the brittle nature of some of the most successful models in the field and how very simple manipulations of text can cause them to break down in embarrassing ways. Currently, the best models are only capable of memorizing elaborate patterns, and we are nowhere close to a model that can actually comprehend the text. Therefore, I expect a lot of research in this area, and I am optimistic that we will begin to read about ways that "common sense" can be introduced and incorporated into our models.
- Michael Morgan, Lead Data Scientist, Nielsen
The importance of quality, clean data will grow.
I think we’ll see an increased focus on finding innovative ways to supplement and improve training data for “neural networks” (e.g., computer systems that are modeled on the human brain and nervous system. These types of models can only deliver results that are as good as the data that is put in. Therefore, I believe we will see a greater focus on efficiently increasing the quality and quantity of these datasets.
- Jessica Brinson, Senior Data Scientist, Nielsen
Machine learning and artificial intelligence will become key differentiators for companies looking to unlock growth.
More sophisticated algorithms, stronger data science talent and an increasing volume of data will enable companies to use ML and AI in 2019 as key differentiators and unlock more value than ever before.
Five years from now, the conversation around data science and AI will advance toward how to develop complex problem-solving skills from the development of tools and algorithms. Additionally, algorithms and software will become more user friendly and democratized for the non-data science population.
- Avi Jain, Regional Data Science Client Lead, Nielsen
The ability to collect online data will become more challenging and fragmented.
As more and more internet browsers start blocking third-party cookies, data management platform (DMP) vendors will begin to face challenges when collecting online data. Marketers will find individual DMPs will not be able to offer the same amount of data with the same level of accuracy as they did before. This will greatly affect marketers’ ability to target and segment their audiences accurately. In the meantime, DMP vendors will look for solutions that will enable them to compensate for this loss of data.
- Pengfei Yi, Director, Data Science, Nielsen
Successful companies will significantly invest in data science-backed solutions
Digitization is disrupting the traditional marketing approach, creating a more dynamic and fragmented landscape. As a result, consumer packaged goods (CPG) companies in particular are seeing added pressure to perform with greater profitability. CPG marketers are placing higher importance on real-time solutions, relevance and speed over robust and ideal design. They’re also looking for agile and innovative approaches to provide a total market view, and companies will increasingly turn to data science to address and solve these challenges.
The amount of digital data that consumers generate is expected to double every year going forward, and companies that want to navigate this digital deluge will invest in digitized techniques to capture and measure data. They’ll also design solutions using AI, ML and neural networks. Companies that incorporate data science algorithms will find increased scalability, efficiency and will be able to advance beyond the conventional stream of measurement.
- Neerja Joshi, Director, Data Science, Nielsen