#MarTechConf: 5 Expert Tips and Tricks on Architecting your MarTech Stack

“You have to be nimble,” said Alexei. “One of the biggest challenges is — and it’s not just for companies of our scale, it’s companies of all sizes — if you’re tied to monolithic systems…you are not agile by definition. We strive [to] deconstruct the monolith. You really have to find ways to work with many different vendors to build a stack.”

“The biggest issue is we have as marketers is that there’s information we’re gathering across channels, [but] we’re not getting a true view of who our customer is, what our customer is doing and I think, more importantly, what they are not doing in those spaces,” said Alexei. “So the challenge becomes how do we pull together all of these different data points across different avenues, across different channels, especially as channels emerge…how do you have a view of customers that makes the most sense, how are you able to pull in real and meaningful data and do actual, strategic execution against that data.”

“Customers are moving between different forms of identity,” added Vijay. “We may know a little bit about them, they may be anonymous or they may seem anonymous at some point in time, they might become known customers at other points in time, they might become mobile-addressable customers…so, managing that transition has to be part of the full circle view.”

“The [AI] layer was unachievable 5, 6, 7 years ago,” says Alexei, “but now it’s available. Sometimes [it’s] inherent in some platforms, or leasable, and it can be utilized. It used to be a scary thing to have AI make decisions for you, but it really does cut down the toil and make those intelligent activation decisions.”

“You can always go to a data scientist and say, go build me a model to do XYZ,” added Vijay. “But then, often times, they don’t have the full circle view of the customer so it takes them a month to stitch together data from disparate sources, a month to normalize the data, another month to extract features [all] before they even start building these models. And once they’re built out, they’re still not informing a customer experience.”

“The biggest change that’s come in our world is what was crazy expensive a few years [ago] — storing mass data to create these views — that cost has come down,” noted Alexei. “We used to be in the batch and blast world; it was all push and we had no way to understand what right screen, right time, right audience meant. Now we’re able to get that feedback through these systems. The ability to have all that all crunched data-wise and have data stored for us has been the big revelation in the space. Without that, we’re not going to have any…decisioning.

“It’s a lot of computational power, but it’s an exceptionally large piece of this puzzle to understand what decisioning means,” he added. “It means saving time — marketers can do more work, they can do more marketing, they can be more strategic. It’s a piece that I think is missing in the conversation’s stack with a lot of firms. It has to be part of that discussion.”

“The paradigm shift is customer-centric,” said Vijay. “The old way of doing things was buying an email system and a paid media system, and they would all have their own data and decisioning. What’s happened now is… architecture allows you to put your customer at the center.”

I love the phrase paradigm shift in this context,” Alexei added. “It’s both ways…much in the way social media marketing exploded on the scene as…a conversation, that’s the way to look at this type of exercise. It’s information we’re getting from a customer about their preferences…simply as they interact with our offerings. For us organizationally, it’s looking from the top down so we can make data-driven decisions.”

Ready to get started? Alexei’s advice is:

“If you want to start somewhere, start with having an opinion about where your customer data lives and how you’re looking at that customer data. I would caution to say it doesn’t mean a CRM. A CRM is storing objects that have happened; customer data, customer data platforms, customer data activation platforms are storing data you can work in real time. If I was to sit down and carve out a budget, that’s where I would look.”

This talk took place at this year’s MarTech Conference in San Jose. If you missed it, or are looking for a deeper dive into this expert advice, watch the full recording.