By Nick Sleeth
From everything we hear and read today, it sounds like and artificial intelligence (AI) and machine learning (ML) are going to make data more insightful and actionable.
But the reality is that AI and ML are very complex and expensive technologies to leverage, if you include all the costs for the technology, processing power and labour. They also require very large amounts of data to be effective. In a recent Boston Consulting Group survey of over 3,000 companies, 85% believed they could gain a competitive advantage from AI, but only 15% of organizations were using AI extensively1.
As a result, AI and ML are mostly used in a “black box” approach, meaning AI and ML technologies are imbedded inside other solutions, like digital journey engines, dynamic content selectors or digital ad placement to enhance their effectiveness. Accessing AI and ML through the black boxes shields us from their complexities and high direct costs while giving us the benefits for our businesses.
Integrating data sources
Marketers have always tried to leverage data and insights to increase the performance of the next campaign with some degree of success. Up to now, we have mainly used a single source of data based on simple spreadsheets.
The reality is that there are many data sources. Using only one source with column filtering of a simple mailing or contact list is not enough to give us the insight into our customers and how to reach them with the right message.
For example, to accomplish a much more targeted approach to direct marketing, we would bring together 1st party data, usually from corporate systems or CRM, 2nd party data from competitive or market information and 3rd party data, such as prospects’ demographic or psychographic information. As you can imagine, the challenge is bringing it all together to create prospect insights, but it would allow for a real targeted approach to people who are the potential of becoming our customers.
If we agree on the need to integrate more data sources to build better insights, then where does your business start?
- Start small with identifying data and insight objectives driven by business needs. Don’t start with spending months building a complete data dictionary with access to all your data sources as this will be a wasted effort; and
- Definitely do not start with AI and ML solutions directly as both of those approaches require a massive amount of information and knowledge to be useful. Leave this to the experts.
Instead, look at sources of data you have, or could easily acquire, to be used to help your organization make decisions. For example, adding local demographic information to a customer list give insights into why and how those people purchase your product.
You may want to acquire a data analytics or management tool to bring your multiple sources of data together and allow you to run simple reporting to learn and grow. Is there a chance that you throw away this tool in a year? Absolutely, but that is the benefit of Software-as-a-Service (SaaS)-based tools. Easy in and easy out so when you are done with it, you can graduate to a more complex tool.
Using the black box
As you use AL and ML-embedded tools that leverage these complex technologies based on this black box approach, you need to make sure you understand how they work and how they get results. The providers of these tools should explain in a good level of detail how they work and give you the documentation on how these technologies come up with results. As you rely on the results from these tools to make decisions and build a strategy, you will be asked by people in your organization to explain how and why, so you need the underlying knowledge.
There is no doubt that AI and ML will be part of our marketing technology stack in the future. They will most likely be used in a black box approach to keep it simple particularly if we want better insight into our customers.
What matters today is to start thinking and developing better insights about our customers based on multiple sources of data to better target them and which sources of data we have to use to get there. AI and ML technologies will only be the enablers, and the strategy and tactics still need to come from marketers.
Nick Sleeth is a member of the Analytics and Insights Executive Council for the Canadian Marketing Association (www.the-cma.org). He earned a Bachelor of Science in Computer Science and Statistics from the University of Western Ontario and has held senior marketing and sales leadership roles at Myplanet Digital, and Cision among others.
1 Boston Consulting Group. “Most Companies Have Big Gaps Between AI Ambition and Execution, MIT Sloan Management Review and Boston Consulting Group Study”, press release, September 6, 2017.