Photo licensed from iStockphotos

Photo licensed from iStockphotos

Reporter with Silicon Hills News

Twitter emerged with a splash at South by Southwest in 2007, unleashing an ever-growing barrage of tweets, posts, photos, and other social media interactions.

People Pattern, an Austin startup, is fittingly staging its coming out party at this year’s SXSW.

The company is developing software that mines the oceans of information embedded in social media to provide insight that companies can use to develop sharply focused digital marketing campaigns quickly and efficiently.

People Pattern applies data science to analyze indicators such as the words people use and the people they associate with – their patterns – to identify specific segments and sub-segments of customers. With that information companies can tailor messages for specific groups, develop content, generate sales leads and in media planning.

“We’re able to scale it, collapse the timelines and you’re able to take action directly,” Ken Cho, a People Pattern co-founder and chief executive, said of the market research process.

Cho, a founder and former chief executive of social marketing company Spredfast, brings the digital marketing expertise to People Pattern. The data science comes from co-founder Jason Baldridge, an associate professor of linguistics at the University of Texas who conducts research in computational linguistics.

They started the company in early 2013. Now it formally launches at SxSW with an open house at its Pedernales Street offices in East Austin and an invitation-only party at Franklin Barbecue.

People Pattern looks to catch a rising wave of corporate spending on digital marketing. The Gartner Group reported that while digital marketing is still a fraction of traditional marketing spending, more companies are transitioning to digital channels and increasing spending to address them.

“Historically, hundreds of billions of dollars have been spent annually on marketing, yet businesses had no real way to measure its true impact and return on investment,” said Bryan Stolle, a partner in the Mohr Davidow venture capital firm and an investor in People Pattern. “It was also very difficult to give attribution to which marketing program or dollar actually resulted in a sale.”

Using computational methods in marketing, called quantitative marketing, provides accurate feedback on what works and enables precise understanding of customers, Stolle said.

With a venture investment, he’s betting on Cho’s marketing and entrepreneurial acumen and Baldridge’s computational expertise. “I immediately saw them as a very powerful team that could build something game-changing in the quantitative marketing field,” he said.

Mohr Davidow invested $250,000 of seed money in early 2013 and came back at the end of that year with a $4.5 million Series A round.

In the past two years, the company has been developing its software, which involves natural language processing and analytics. And it’s putting together a track record serving a range of customers, including some Fortune 500 companies.

As an example, Cho and Baldridge recounted work People Pattern did for a large fast food company known for its hamburger restaurants arching across the world.

The company hired People Pattern to figure out switchers, a segment of customers who go back and forth between its restaurants and those of competitors.

Cho said that within eight days People Pattern pulled in 600,000 social media profiles, harvested relevant information from posts and tweets and other interactions, and analyzed it.

People Pattern found that the hamburger chain shared 25% of its switchers with a chain of sandwich shops. The switchers subbed in sandwiches for burgers because of health perceptions.

Cho said that the switchers should be targeted with messaging highlighting the hamburger chain’s healthier options.

Beyond that core task, Cho said People Pattern matched research about customer categories that the company had learned over 35 years of market research. People Pattern also uncovered a group of “breakfast only professionals” in Sacramento, which led to a market test of all day breakfast in the California capital.

People Pattern delivered validation and contributed to product development as well as fulfilled its assigned task, Cho said.

“All in eight days,” he said. “Their CEO was blown away.”

Cho can talk marketing, personas and market all day. Baldridge matches him step for step on the technology side. He researches computational linguistics, particularly natural language processing and machine learning. He’s on leave from UT Austin during the 2014-2015 academic year to work on People Pattern.

His boss at UT said Baldridge brings strong skills to People Pattern.

“Jason is doing fundamental work in machine learning and geo-temporal grounding of natural language, among other things,” said Anthony Woodbury, chair of the Department of Linguistics at UT-Austin. “There is therefore a very close connection between Jason’s research and teaching, and his work with People Pattern.”

Cho and Baldridge had met six years ago and Cho had kept track of Baldridge. (“He was stalking me,” Baldridge joked.) When Cho shared his ideas with Baldridge, the professor, as he’s called around People Pattern, was on board.

“This guy knows what he’s doing,” Baldridge remembers thinking. “He’s got a business problem that needs to be solved by Fortune 500 companies. Cool! That’s exactly what I was looking for.”

The program that Baldridge and the other data scientists at People Pattern have developed taps into social network databases to gather information, classify words, associate people in groups, and analyze the data.

With the databases and the tools of data science People Pattern can do in days what takes months for traditional marketing campaigns to do.

“We have the potential and have shown several times that within a week we can collect hundreds of thousands of profiles and run it all through this automated pipeline that then leads into something akin to personas,” Baldridge said.

Cho said other companies are working in a similar area but none possess People Pattern’s hard-core science approach.

Stolle, for one, is encouraged by the company’s progress. “The company is definitely tracking ahead of plan – always a good sign for long-term success,” he said.