Spiketrap has raised $3 million for its AI platform that helps brands understand how they are engaging with consumers on social media. The San Francisco-based company processes tons of text-based social data — like Twitter posts — to discern how consumers view a brand or a game in real time. When a big event happens, such as the launch of a new game like Nintendo’s Animal Crossing: New Horizons, Spiketrap can collect the prevailing sentiment about the game.

“This is really exciting for us to raise money with a fantastic syndicate of people,” said Spiketrap CEO Kieran Fitzpatrick in an interview with GamesBeat. “We continue to grow our core product with a focus on understanding language, specifically the language of the internet.” The company is also expanding from PC and console games to gathering data on mobile games.

Spiketrap claims that 90% of all data on the internet is unstructured, and it maintains that brands that neglect this data are making major decisions on a fraction of the total information. Spiketrap’s AI-powered platform enables companies and brands to understand what people are saying about them as they’re saying it, reducing the time it takes to get key insights and change messaging as needed.

The company’s initial focus has been on video games and what people are saying about them.

“Over time, we’ll round out the customer base and expand into new markets like movies, entertainment, and ancillary entertainment,” Fitzpatrick said. “Gaming was the most difficult market.”

What Spiketrap is learning

Above: Spiketrap can drill down on just about any game to get sentiment.

Image Credit: Spiketrap

Fitzpatrick said the company’s proprietary technology is able to ingest massive amounts of data in real time and contextually understand what is being discussed while eliminating spam and noise. For example, a FIFA soccer scandal was unfolding at the same time as one of the recent video game shows, and a cursory analysis would have concluded that FIFA was “winning,” or dominating the conversation online, at the video game show. In fact, most of the references were to things other than the video game itself, Fitzpatrick said.

“It was very obvious that [the cursory analysis] was incorrect,” Fitzpatrick said. “We take a multiple vector approach.”

Spiketrap uses natural language processing to understand terms like those used in the game industry and figure out whether the sentiment is positive or negative.

Sometimes, that sentiment is hard to discern if there are vocal parties. In one case, a game box cover featured two Black women, which upset a tiny group of gamers with racist reactions. But the majority of people on the internet had no issue with the cover, and Spiketrap convinced others not to worry about the magnitude of the negative comments.

“That helped restore my faith in humanity,” Fitzpatrick said.

In a separate example, The Last of Us Part II generated a strong negative reaction, as some players didn’t like the path of the story or the diverse characters in the cast of the PlayStation 4 game from Naughty Dog and Sony. But Fitzpatrick said he didn’t see much negativity in overall sentiment.

“There is a chunk of conversation that’s negative,” he said. “But then there are other positive conversations happening.”

Spiketrap pumps more than 400 sources of data into its AI engine, which parses the data, said chief business officer Mike Owen, who joined the company last year. That range of input helps it with accuracy, he said.

“We are able to ingest massive amounts of data at scale in real time, strip out the noise, and produce coherent conversations,” Owen said in an interview with GamesBeat.

Moving forward

Above: Spiketrap can show what’s trending or not.

Image Credit: Spiketrap

The best thing about this platform is that users don’t have to be search geniuses, Fitzpatrick said. You don’t have to formulate complex queries or use boolean algebra to search through the data, as Spiketrap surfaces the insights for you.

Spiketrap’s first customers have been video game publishers, developers, and streaming platforms, such as Ubisoft, Bethesda, 2K, and numerous equity research companies focused on the media and entertainment sectors. Now Spiketrap is ready to expand to other industries, Fitzpatrick said. The company plans to use the new funds to double down on its mission of helping brands make decisions based on consumer sentiment.

“You need to know if you’re winning the hearts of your audience, and then [understand] how to have a genuine conversation with them based on what folks are saying,” Owen said.

Fitzpatrick and Adam Sessler started the company in 2016, and it has 15 employees. [Update: Other founders included Virgilio Pigliucci (CTO) and Andrea Vattani (chief scientist).] Sessler, a well-known video game personality, left the company in March of this year, though Fitzpatrick declined to comment on the reasons for his departure.

Susa Ventures led the recent funding round, with participation from 645 Ventures, Pathbreaker Ventures, Oceans Ventures, WndrCo, SV Angel, Anorak Ventures, Harry Stebbings of 20VC, and some strategic angels.

Sign up for Funding Weekly to start your week with VB's top funding stories.