"Keeping track of the rapid pace of evolution and trying to integrate all those tools is single-handedly probably one of the biggest challenges for CMOs."
M7: What was the most successful marketing campaign you have ever worked on? What made it so successful?
JP: In my previous organization, we led a re-brand of the company in its first-ever national advertising campaign. That meant we weren’t constrained by existing rules and assumptions about how things had to be done, so we were able to quite creatively and innovatively harness several digital channels in a way that many of our competitors hadn’t. In terms of revenue or asset flows, new customers, traffic to the website and just about every other metric, it was a tremendous lift. That’s probably the single-most effective marketing campaign I’ve worked on, and it was completely comprehensive. It had a brand component, a communications component, a community component. It had very specific direct targeting and campaigns along with a goal of raising basic awareness. We were targeting by social media profile and audience. I think it was successful largely because it was multi-faceted and end-to-end.
M7: At Lionbridge, how do you see AI evolving with digital media?
JP: For AI, we are still really in a hype-cycle. There is absolutely no doubt that AI and everything that falls under its purview will have a profound impact on the scale, data, and decisions we are able to process as marketers. The number of simulations and tests we will be able to run to optimize a campaign or project pre-launch, plus the ability for us to pivot real-time and use new insights, will be absolutely profound. But I think we are still very early in the journey. There’s more hype than substance right now. I’m not skeptical about AI’s eventual impact at all. But right now, the tools are not mature, and we are not mature in our understanding of AI, even regarding some of the most impactful ways to use it. So, I’m super optimistic over a 3-5-7-year period. But, I’m very skeptical over a 6-12-24-months period.
"AI and everything that falls under its purview will have a profound impact on the scale, data, and decisions we are able to process as marketers."
M7: How does the acquisition of Gengo strengthen Lionbridge’s position in machine learning?
JP: We are one of the world leaders in AI training data services. The fuel that makes AI work effectively is incredibly high volumes of high-quality data. And the problem is, as we know, most organizations have not done a good job being stewards of their data. Their data exists in lots of places, their data is not clean, it’s not annotated or mapped, there’s no organizing taxonomy around its structure. CDPs are still relatively early for a lot of folks, and so the catch is that you have lots of data, but that data may be largely unusable. Sort of like, there is a lot of oil in the ground, but just because the oil is in the ground doesn’t mean you have gas that works in your car or in a plane. A lot has to happen to refine it, to get it ready. The same is true with data “fuel.”
The good news is that we are one of the world leaders in helping companies refine their data so it’s usable in training their AI systems. Identifying and collecting data, and giving that data meaning, for example, is it metadata or actual data, is that an important variable or not an important variable, etc., is an integral step in the machine learning process. We help companies collect data, organize it, give it meaning so it can best feed our customers’ AI platforms.
We do this by activating our one million-strong SmartCrowd. The Gengo platform really helps us create a scalable system that will allow more companies to be able to improve the quality of their AI training data and, in turn, the quality of their AI.
M7: How do you stay updated on the latest trends in marketing technology?
JP: I read a lot. I get a couple of hundred emails a day. Part of that is because I sign up for lots of different sources of information. I think you have to consume a tremendous amount of information right now to pull in all the weak signals. And I probably dedicate more of my days than I would normally to looking at new capabilities just to keep me informed and up-to-date. This allows me to get a sense of when I think something is at a point where we can use it and harness something early for a strategic advantage.
"Most organizations have not done a good job being stewards of their data. Their data exists in lots of places, their data is not clean, it’s not annotated or mapped, there’s no organizing taxonomy around its structure."
M7: What are the top 3 trends that you foresee for 2019 going into 2020?
JP: We are going to hear more about AI. We will see more examples of it not working in the next 18 months than we will see examples of it working, but I think that’s a short-term phenomenon, and we will work out the issues.
I think there is absolutely no doubt that we are going to see an increase in the production and consumption of video and its import in the production cycle. People are consuming more and more videos, and even B2B buyers are much more heavily consuming videos as a way to absorb a lot of information in a pretty short period of time. In 2-3 minutes, you can convey a lot more, and frankly people can understand a lot more than if they spend 20 minutes reading a whitepaper or something of that nature.
And clearly, voice search is on a big rise. There are a couple of key reasons for that. If you think about the US workforce, half of it now only knows a universe where Google is the dominant information organizational paradigm. Pair that with the sheer explosion of speech applications and devices, whether in your home, in your car, or in your hand—today, every phone is speech-enabled. It’s clear that all the large players are investing a tremendous amount of effort into proving the quality of voice search, and an increasing number of users are speaking to a device or multiple devices to get information. That’s a real challenge for organizations and marketers, because the way people speak to their device and physically ask for information is not the same as how they type a query into a search engine. So, the whole way in which we optimize and organize for search results in a voice-search world is quite different than in a text-based or type-based universe.
M7: What is your superpower?
JP: I’m not really sure I have a superpower. But there’s something I think I do that seems to be more of a struggle for some others. I’m able to assemble a collection of weak signals to predict what is going to happen before it happens. I really believe all the information is there, but I seem to be able to use that information to make some good bets about what the future holds.