Back to Top

Yext Brain and the future of conversational AI: Q&A with CMO Jeff Rohrs

In late October, Yext announced Yext Brain, the next generation of their Digital Knowledge Management (DKM) platform.

For all its technological complexity, the goal of Yext is simple: give businesses a way to make sure all their most basic facts are accurate and up to date across the web. What are your hours? What’s your address? Are you closed on Veterans Day?

These “basic” facts often come with little fanfare in the world of marketing. But when it comes to local SEO, mobile and voice search, they’re game changers.

Yext is hedging their bet that businesses who can get their facts straight first — serving up accurate, timely, relevant information to consumers — will win top SERP spots. As mobile and voice push their ways to the forefront of consumer searches, the coveted “position zero” matters more than ever.

What is Yext Brain?

Yext Brain is an evolution of their existing DKM platform. It lets businesses create digital “entities” such as events, locations, products, professionals, and limited time offers. Each entity then has “relationships” with other entities. For example, items on a menu could contain information about type of dish, vegetarian-friendly, season, calories, etc, and then be linked to each other accordingly. Within a hospital network, physicians could be listed as entities with information about insurance, specialties, location, hours, etc, and likewise be connected accordingly.

For the digital marketer trying to keep thousands of minute details up to date across countless places on the web, Yext Brain lets them update the entity information once, and have that automatically populate across all the places that entity appears.

Alongside Brain, the company also released Think, a type of “answers engine” housed on a brand’s own website, that delivers accurate information per user queries, as well as Pages, which allows brands to create new pages for their website based on any entity stored in the Brain.

Ultimately, the goal of all three is to help consumers quickly find accurate answers to their questions, whether they’re looking on desktop, mobile, or voice, across Google, Alexa, Siri, Amazon, or a number of other apps and services.

CMO Jeff Rohrs

We sat down with CMO Jeff Rohrs to pick his *brain* about the product, the perspective behind it, and how it fits into the future of conversational AI.

Before joining Yext in 2015, Jeff spent 8+ years as VP of Marketing Insights at ExactTarget and then Salesforce following their $2.7 billion acquisition. He holds both a JD and a Masters of Mass Communication from Boston University. And might I say, his knowledge of Spanish soccer clubs was extraordinary.

Before we dive in, a quick note on the context

We were at the Yext Onward 2018 event here in New York at the end of October. As life would have it, the entire venue was evacuated earlier that morning due to an explosive device mailed to CNN, which shared the building complex (same day as the explosives sent to the Obamas, the Clintons, Soros, and a number of others).

Months of planning, and now thousands of attendees were standing on the sidewalk waiting for NYPD to clear the building for re-entering. We flocked around on the street and inside Starbucks, wondering what was going on. I took to twitter for information (like we do) and found what I need via a tweet from Jeff himself. A couple hours later, we were all allowed back in the building. Sessions were rescheduled, meetings moved — you can only imagine the frenzy.

So that was how Jeff, myself, and the PR rep found ourselves in a small supply room for this interview. No windows, about the size of four desks — and it didn’t matter at all. Jeff was remarkably composed and calm. He began the conversation commenting on how these were the types of situations that remind all of us what’s important, that in the end we were safe and could go home to our families and friends.

Needless to say, it was an uncommon chat on a wild day. Here are the highlights:

ClickZ: What was the vision behind Yext Brain?

Jeff Rohrs: Our whole vision is that the web has traditionally been a series of very flat, or, for lack of better term, dumb documents. They’ve got great information, but there’s no AI hooked in.

We’re relying on a paradigm where we want a third party to come in and index and make understanding from this information, and then we pull it back and serve it in the moments that matter.

I think that model has changed. It’s changed because of the rise of AI services, intelligent services, etc.

Now, brands need to get incredibly granular. They need to manage the very finite facts about their people, their location, their products, menus, practitioners, what have you. And they need to manage the connectivity of those to different attributes about that particular entity.

We like to say your brand has heart, we want to help it get a brain. That brain makes sure all that information gets out to your consumers in the moment that matters.

Intelligent services are digesting the finite facts, they’re not digesting documents. Brands will have to get very immersed in all the facts that matter to consumers.

CZ: Can you walk us through Brain, Think, and Pages?

JR: Brain is the evolution of our knowledge manager. The big thing there is of course your ability to map connectivity, so you can create the relational associations in different data sets, and create a lot of efficiency in how to update that and publish it.

So we’re solving for that conversational AI moment. Today’s conversational text-based chatbots and AIs become voice-enabled. Voice is just a UI. So it’s really what’s comfortable for the consumer, and making sure that you can serve up information at the most granular level that they want in the moment.

The Think piece is then how do you take the Brain and what’s happening on your website and make them able to answer deeper questions that consumers have. Think is going to help you understand what is being asked by consumers, so that you can then build that information out in the Brain and better serve those types of questions with answers. It makes your website more conversational, reduces friction, and reduces consumers leaving because they’re frustrated.

So now you’ve amassed all this data, how are you going to manage it on your website? That’s where we have our Pages product to build local pages. Done right, local pages in aggregate are getting more traffic than home pages on a website.

CZ: How do these fit into the broader play of intelligent services?

JR:  An intelligent service has a UI that the consumer chooses and the service dictates. So I want to use Siri on an iPhone, or I want to type in a search on Google on my laptop.

Then I’ve got an AI, some sort of intelligent algorithm or thing that’s going to determine what is relevant to me in that moment, using what it knows about me and what it knows about what I’m asking.

Then you have the knowledge layer, the knowledge graph that it’s built on. That knowledge is sourced from wherever that intelligent service feels is the best content.

And in the absence of direct information from the business which is the best source of truth, they’re going to source it as they have in the past from third party sources, from websites that could be out of date. We like to say you can’t control the UI and you can’t control the AI, but you can control the knowledge they know about you — and you have to manage it.

CZ: That’s an interesting perspective — feels opposite to what a lot of people are focusing on.

JR: It’s funny, because everything that’s old is new again. When I first got involved in the web in its nascent infancy in the mid 90s, we would go out as an agency and sell websites as, “oh, this is like digital brochure.” We’d analogize it to paper. It’s hilarious, but it largely remains true.

Well now we have all these intelligent services that are only getting more intelligent. However, if you deconstruct intelligence and say okay intelligence is based on knowledge, knowledge is the acquisition and use of facts.

And so our philosophy is, okay, instead of putting all this knowledge out there in long form, we now know that AI really wants facts, and to understand the connectivity and relationship between things.

It’s simple, but it’s a paradigm shift. Get your facts in order.

It’s different to what marketers are used to thinking about. They’re used to, “We have to update the website, and go do this, and this.” And we’re like “hold it,” because AI services today and in the future put so much value on these facts at the granular level. You need to have people and processes in place to update them regularly.

A lot of the legacy SEO world created that mentality that I can just update this once and it’s done. But the reality of the digital world is that if you don’t manage digital knowledge, it atrophies.

CZ: How do you influence SERPs?

JR: It’s a control-through-quality play. We have a track record with these integrations.

When we were on our lead up to going public last April, we did our IPO video and we interviewed our partner manager at Microsoft Bing. He shared with us that they call their program with us, “Yext on Top.” The reason they do is that the quality of the data coming from Yext means that they can trust and rely on it in all these instances and they default to it.

The services are always going to be in charge of what the output is. It’s their service, it’s their relationship with the consumer.

Our philosophy, however, is that the best source of truth is the business itself. Our mission is to help this world get to a place where consumers have perfect information everywhere.

We believe that if the business, who has the most economic incentive to make sure the information, the objective facts — and that’s key, we’re not talking about subjective facts — to make sure that is correct, then over time these services rely more on that. They don’t want to serve up something that’s wrong.

What we’re solving for is such a principle problem that dogs every brand, and that is more pronounced now in an era of voice search, when there’s only one answer. We’re no longer in an universe of ten blue links.

CZ: How are you preparing for the future of search, with visual, voice?

JR: Regardless of what the future holds, by getting the facts straight you’re future proofing your brand. Voice, visual search — all these pull from the same knowledge base the search engines have already accumulated.

Another important part of our philosophy: we don’t charge any of those intelligent services for any of the data that we provide. We’re looking to optimize attribution points. As new consumer data sources arise, we want to be there. We want perfect information in the hands of consumers.

When I was at ExactTarget, there was a lot of education around “subscribers rule, serve honor deliver.” Here, there’s an education around digital knowledge management is a thing, ongoing, why does it matter.

As marketers we’re so tuned to hit for the fences. We’ve neglected the foot traffic, the basic stuff. If at the moment of need, your address is wrong, things go sideways pretty quick.

Brain makes brands future ready in an AI world, because it makes their data structured in a way that’s digestible to those systems.

CZ: You keep saying “AI-ready,” can you elaborate on what that means to you?

JR: I use AI to be broad. I also talk about discovery rather than search. Search is a narrow, intentional, I am going to search for this. When you migrate over to a voice assistant world, am I searching if I’m asking Google for the hours of the Best Buy near me? No, I’m asking. It’s analogous, but it’s discovery.

The reason we focus on the AI ready piece of this is that there are an awful lot of services out there that do not crawl or index the web at all. They’re using information from third party sources or directly from a business through a platform like Yext. Those services need that information. If you’re just doing an SEO or search optimized approach, you’re missing out on probably the fastest growing piece of this puzzle.

Real time cuts two ways: in terms of the business itself managing to make sure information is accurate in real time. And then it’s the services themselves can they serve that up in real time.

I can change my store hours, but if a service doesn’t update but every 24 hours, it will take 24 hours to update. The consumer and the economics of those services are going to dictate how close to real time they get.

You can obviously assume the big ones — Google, Amazon, Apple, Microsoft — they’re going to to want to get as close to real time as possible. That really changes your mentality because you want to make the foundation — and really, digital knowledge management is foundational to marketing — broader than search. It’s about discovery.

AI ready is about getting the facts, the data, into digestible format so that anything that comes, regardless of the sophistication of AI, can take this information and deliver it in a way that’s meaningful to the consumer in the moment that matters.

The post Yext Brain and the future of conversational AI: Q&A with CMO Jeff Rohrs appeared first on ClickZ.

Reblogged 1 week ago from www.clickz.com

Comments

Write a comment

*