We’ve all heard how Artificial Intelligence (AI), the new talk of the town for marketers increasingly fired up and finding it difficult to contain their endless enthusiasm for the apparently boundless opportunities that AI portends for the business world, is nothing short of a social and technological revolution. But, like all grand inventions, the capacity to solve huge problems and create endless opportunities (and we mean across every imaginable sector!) comes with a dark side that we need to be aware of.
Artificial Intelligence has a key role to play in Marketing, in how it can personalize the customer experience – by using data science to analyze browsing history, preferences and user characteristics and from this, predict behavior and deliver exclusive, personalized experiences to exponentially increase engagement levels. Despite the many advantages to be reaped from this kind of one-to-one marketing which engages in a dynamic way with all touchpoints between a brand and its users (better targeting, optimized conversion rates, etc.) a priority must be to assure - as Professor Pietro Perona from the California Institute of Technology emphasized at a conference recently held in Padova - the quality of the stored data used by AI engines.
Here at Content Intelligence Network, where our business is Content Intelligence, meaning Artificial Intelligence applied to content from which to extract high quality data (First Party Data) on the interests the people viewing the content, we wanted to discuss the issue in detail with Federico Pedrocchi, a scientific journalist with more than 30 years of experience writing for publishers, newspapers, radio and web. Director of Triwù, the Innovation web TV channel, Federico knows a lot about Artificial Intelligence from his role as conductor of “Artificial Intelligence” on Amazon Audible and “IA-Intelligence in prima linea” on Radio24.
Q: Federico, how do you see Artificial Intelligence relating to Marketing? Where is it being used already and what challenges will it have to overcome in the future?
A: We already have working algorithms to analyze interesting proposals for users. Clearly, in order to make the best possible purchases, users would need to do an impractical amount of research requiring an inordinate amount of time, so help from AI might be useful. This brings with it, obviously, a problem that crops up in many other contexts: algorithms that are not influenced by advertising interests. This is the major challenge.
Q: In the Era of the Customer, why is data so important? Why is it not enough just to gather data, why does it also have to be activated to be useful to your business?
A: As I said before, with increased choice - we can now buy from anywhere in the world - buyers can’t be spending hours online deciding what product to choose. So, a means of selection would be a massive help. This is the domain of Big Data tools, but we have to understand that data collection does not mean filling a big, universal bin with millions of records then expecting the Big Data algorithm to pull out the information we need. This is a common mistake. The quality of the data must be a primary concern before it goes into the bin, otherwise the old adage about the science of databases applies: rubbish in, rubbish out.
Q: Do you think one development of applying Artificial Intelligence to content could be the ability to identify the needs and interests of users from the use they make of content? To stand out from competitors these days, is it important to offer personalized experiences?
A: In generating suggestions to guide selection, the great value added of AI is that it seizes the huge personalization opportunities that the web has brought us and allows us to move in a direction that would have been impossible in the past. This ability to make customized proposals is strategic. Clearly, I have to emphasize that this is not something which should be taken lightly. Users must be able to see that what is being suggested to them is a true reflection of their actual characteristics. If this doesn't happen, the resultant boomerang effect could be dangerous: if someone says they know me but I realize the suggestions I’m receiving are nothing short of general, well, I think that would be devastating, extremely irritating and repulsive for customers. Banal: our personalities are important to us.