Omnicom’s New Chatbot Arms Its Network of Agencies With More Data and Analytic Tools

Workers can better access deep troves of info

The agency's chatbot AUBI can handle complex data sets. Annalect
Headshot of Lauren Johnson

Omnicom’s newest data scientist hire is a chatbot—kind of.

To arm its network of agencies with more data and analytic tools, Omnicom-owned Annalect is launching an internal chatbot called Annalect Utility Bot Interface (or AUBI) that spits out deep troves of data faster than the agency’s team of data scientists can handle. The idea is to make Annalect’s data more readily available to anyone within Omnicom—from creatives, media buyers and strategists—that works with Annalect.

“Chatbot technology has arrived as a form of artificial intelligence to facilitate automated exchange between a user and an entity,” said Slavi Samardzija, global CEO at Annalect. “We said how about we build a chatbot to bridge this gap between a super powerful data scientist and an end user—a strategist, a creative director, a planner, a buyer.”

Here’s how it works: Once staffers log into Annalect’s platform, they can use the AUBI app. From there, AUBI prompts workers to answer a question about what they’re looking for. At launch, AUBI can perform a handful of tasks that previously required a lengthy process of manually coding data. So, a creative at BBDO or media buyer for OMD, for example, could use the tool to dig up data that crunches which websites consumers use.

“There is so much attention spent on the application of AI with consumers,” said Doug Rozen, chief digital officer for OMD. “As important as that is to the future of marketing, we see greater more immediate needs to use AI as an agency. AUBI is an example of how agencies can use AI to work at finding the insights that matter.”

For example, Annalect database includes stats called “clickstream journey” that tracks which websites someone goes to before and after they buy something. A car brand, for instance, could see that a 34-year-old in Dallas shops at in the morning, goes to TripAdvisor in the early evening and watches Hulu at night.

“To use this historically, you had to contact a data scientist,” said Anna Nicanorova, director of Annalect Labs. “We’ve created this choose-your-own adventure where the bot is trying to guide you to what’s available and what’s not.”

AUBI can also pull up data on topics like frequency, channel activation and audience indexing.

“Everyone talks about data-driven marketing, but that really comes down to a series of questions: Who is my audience? What are the insights about that audience? What’s the optimal mix? How should we measure? What’s your content strategy?” Samardzija explained. “Each one of those questions sits on one or more data sets, a bunch of lines of code and algorithms to deliver that. Any of these utilities can take an hour to explain—now we have a bot that sits on top of it.”

To launch AUBI at Cannes, Annalect is demoing how the technology works by answering questions like “Where can I get food in Cannes?” In addition to recommending a restaurant, AUBI also pulls in a map.

Dan Palen, Annalect’s CMO, equated the data science today to the early days of web design, when agencies and marketers didn’t know how to build their own websites.

“We’re getting to a point where data is so readily available, technology is decently easy to build, people are starting to over-engineer everything and at the end of the day, we’re trying to get people who understand the clients and brands best access the data so that they can make decisions in a timely manner,” Palen said. “As we get more data sources and as we build more analytics it will get smarter.”

@laurenjohnson Lauren Johnson is a senior technology editor for Adweek, where she specializes in covering mobile, social platforms and emerging tech.