Trust, Safety and Data: Wing Invests in Spectrum Labs’ $10 Million Series A

Peter Wagner | September 24, 2020

Today Spectrum Labs announced it’s $10 million Series A. Wing is excited to be an investor.

Our investment in Spectrum Labs is part of our Modern Enterprise thesis which drives everything we do at Wing. The Modern Enterprise is an agile workplace, built on data and powered by AI. It runs on a new generation of business applications that stand on the ground truth of data and steer business processes with AI. Early examples are already visible in Sales (Gong, SetSail), Engineering (Jellyfish), Employee Service (Espressive) and other critical business functions. I started writing about “Data First, AI-Powered” business applications in early 2016, and began investing in them with the Gong Series A-1 in 2017. Our thesis asserts that all enterprise application categories will eventually be disrupted by this archetype; and that net new applications will be created that couldn’t have been conceived any other way.

Spectrum Labs is one of these net new applications, empowering Trust and Safety teams. Modern Enterprises are building online communities and establishing direct digital connections with their customers as part of a sweeping transformation of the customer experience. But these initiatives can be polluted by toxic behavior and turn negative. Trust and Safety organizations are charged with safeguarding the experience, the community and the brand.

Social media, gaming and dating companies are in the vanguard of the Trust and Safety movement and have deployed a combination of technology and vast human moderation resources. But the scope and scale of the problem rapidly become overwhelming. There are petabytes of user-generated content to process, a mushrooming range of harmful behaviors to respond to, and a world of languages to support. Some of the largest internet companies spend billions each year on human moderators and internal tools, and still fail to meet their own expectations and those of their users. It has become a topic of existential importance for these companies, and a focus of industry collaborations as well as potential regulation.

Fortunately, this is a problem where AI can help. Spectrum Labs’ has built a platform that allows Trust and Safety teams to identify a broad range of negative behaviors in an equally broad array of languages. The core technology allows the rapid creation of new behavior classifiers and support of additional languages in a highly efficient manner, providing customers an ever-expanding menu of capabilities without consuming their own scarce engineering resources. The underlying technology is delivered in a product that includes essential workflow, reporting and analytics. Trust and Safety teams can now dramatically improve their effectiveness while keeping pace with the growth of the communities they protect.

In the world of Data First applications, algorithms are great but data wins. Access to more authentic training data leads to greater model efficacy. Working closely with customers in early adopter segments like social, gaming and dating offers first movers like Spectrum Labs the opportunity to build a data advantage and formidable entry barrier. Meanwhile, traditional enterprises late to the digital sphere will not have much user-generated content of their own, especially for the less common target behaviors. Vendors that hope to serve them will need to bring data as part of their solution.

We spent a lot of time thinking about addressable market for Spectrum Labs and came to the conclusion that Trust and Safety would be important well beyond the large early adopter segments of gaming, dating and social media. In positive market development scenarios, any enterprise investing in digital customer experience and community will need to build a Trust and Safety capability, including most consumer brands. There have already been examples of digital initiatives from major brands that had to be pulled back due to harassment, hate speech, bullying and the like. Trust and Safety is a necessary enabler of the digital transformation initiatives already underway in consumer-facing companies.

The Spectrum Labs team has deep experience building data-driven enterprise technology businesses. Founders Justin Davis and Josh Newman were key leaders at Krux Digital, a pioneer of marketing data technology that was acquired by Salesforce. Much of the rest of the Spectrum Labs team was also part of that seminal group. In fact, Spectrum Labs itself is the first company to emerge from super{set}, the startup studio founded by Krux co-founders Tom Chavez and Vivek Vaidya. My prior firm (Accel) invested in both Krux and Rapt (Tom and Vivek’s first startup, acquired by Microsoft) while I was a partner there. Both companies were applying the Data First, AI-Powered architecture to enterprise software long before it was obvious! It is a pleasure to be working with these two world-class innovators once again.

Tom Chavez and Vivek Vaidya of super{set}, Krux and Rapt

Wing has significant experience with voice and text-based AI in the enterprise. Our portfolio companies Gong and SetSail use it to help sales organizations raise their game. Deepgram delivers advanced voice recognition based on deep learning. All these companies are finding tremendous demand for their products and have opened our eyes to the power of this type of applied AI in the enterprise market.

Wing is thrilled to be working with founders Justin and Josh, old friends Tom and Vivek and new friend / co-investor Alison Lange Engel – a rare assembly of talent around a compelling Modern Enterprise opportunity. Stay tuned!

Share this Article