Alex Poulos, CMO at Crossbeam, joined Jesus Requena, Marketing + Growth at Hex, and Mike Sharkey, CEO and co-founder of Ortto, sat down for a thought-provoking discussion on the future of data-driven marketing, and how increasing privacy regulations and new technology such as AI might impact the way marketers use data in the coming years.

On the power of data and AI

Alex Poulos: Let's make some bold predictions, and then we can get together 12 months from now and prove ourselves wrong. In this new world, what are some of the applications or systems or processes that we're using now that we think will break because they rely on a different premise or a different concept?

Mike Sharkey: I think the biggest problem I see right now with any company - whether they’re starting out or scaling up or are quite established - is just the amount of data that's being collected. And then we still serve it up in traditional ways - here's a dashboard or here's the ability to run a query. There’s a lot of predictive stuff coming to the market, but what can you do with that data with AI that will be meaningful?

One of the best examples we recently experimented with was training a neural net on every subject line on every email that we had ever sent and the corresponding open rate - it was a lot of data, like billions of subject lines, and we would get the AI to suggest better subject lines to the user. This first experiment was okay - sometimes it was a bit crazy because it was trained for pure clickbait.

Then we started thinking, what if we taught the AI all about the data structure - so what subject lines correlate to open rate? That improved the model quite a bit, we could get the AI to predict the subject line open rate based on what it had seen before. And at first, we were like, this good output, really interesting data. Then we ran it against historical emails, and the predictions were within two or three basis points.

When you realize you can make these predictions, why would you do A/B tests? Why would you even test? So the thing that I will predict is that AI will have a profound impact and take over all these things. I don't think A/B testing will exist in three to four years. I just don't see how after seeing that with my own eyes, I don't see how it exists. I think that practical uses of AI will be the next big thing that we can't fathom today.

On the impact of data protection regulations

Jesus: I have another prediction, I think privacy regulations are going to stop us from doing all the things that we currently do. So whether data usage inside products or outside of them. So I do think in some areas, we might have to go back to the just basic principles and I think all these “spray and pray” things like trying to catch people from 10 impressions on Facebook, that's just gonna die.

So it's going to be about going back into where the people are, the communities, being authentic. I think marketers are going to have to come back to some of those principles. I think once you have the data then it's going to be okay - you’ll be able to drop the perfect email, do the perfect segmentation - but to engage them and bring them in is going to be about providing them with something of value.

Mike: The privacy one is going to be so unpredictable. Because it's government, you don't know what's going to happen. I think a lot of the privacy measures that are happening are good, even for someone that's deeply involved in tracking data. I think it's really good to have a lot of these measures in place because it can get really creepy. I think as a user, if it gets too crazy, you start to get a bit worried.

Like you were saying Jesus, if you’re helping the buyer by serving them up the right content, that kind of personalization, I don't think anyone has an allergic reaction to it. If anything that's a good use of where you can take data and make the user's experience way better. So yeah, it seems like there's a bit of give-and-take with privacy.

On the future of the CRM

Alex: Since I asked the question, let me share my point. And again, we'll be back in a year to prove that wrong. I think that there's going to be a profound change in how we think about CRMs. I think this concept that all your data lives in a CRM and all your data in leads for contacts or opportunities or for accounts, I think that model is going to break because your data lives in so many different places now. And because there are so many different systems of engagement we use, I think we're going to see this CRM be broken into three different layers.

One being, how do we access the data? The second one is how do we personalize and act on the data? And then the third one is, are the systems of engagement - like chatbots, or email or websites - where do you actually do the engagement? So I think we're gonna start seeing more forward-thinking on that front as well.

Jesus: I know some data platforms are doing data lakes where you don't need to query on SQL anymore. You have your raw data and you can query the data on a processor. Everything was very crazy.

Mike: I think that's totally true. Alex, I've been thinking the same for a long time, you don't describe data anymore as related objects in this archaic field way. It's got to end at some point. But it'll be interesting to see how it evolves. People have been predicting the death of CRM for as long as I've been alive.

This conversation has been edited for length and clarity. Listen to the whole conversation now.