Obtaining answers to challenging, qualitative queries about products from users can be a costly endeavor, consuming both valuable time and financial resources. This was the firsthand experience of Aaron Cannon, a former strategist at Deloitte, responsible for overseeing research projects for the company’s clients. Despite dedicating hundreds of hours to a client’s project, Cannon and his team found themselves compelled to invest even more time and resources into arranging and facilitating interviews with the client’s customers.
Cannon conveyed his perspective on this matter in an email interview with TechCrunch, explaining, “Enterprise decision makers expect faster and faster results from insights teams. Researchers are grappling with this daily, particularly in the wake of significant layoffs in 2022. The most significant risk to the industry at present is the increasing pace of decision-making, which hinders the ability of insights teams to keep up. This is why researchers need the tools to expedite and enhance their work.”
In response to this challenge, Cannon collaborated with Michael Hess, a former colleague from their time at Untapped, a talent recruitment startup, to establish Outset. As a Y Combinator-backed company, Outset autonomously conducts and synthesizes interviews.
Outset leverages GPT-4, OpenAI’s leading text-generating AI model, to lead interviews with participants involved in research studies. Here’s how it works: Outset users create a survey and share the survey link with potential respondents. Outset, powered by GPT-4, follows up with these respondents to seek clarification, ask probing questions, and establish a “conversational rapport” to elicit more in-depth responses.
For each question, GPT-4 generates overarching themes, tallies response counts, and highlights quotes to “unearth the narrative,” as described by Cannon. He further explained, “Currently, a substantial portion of the process for collecting and analyzing qualitative data is manual. In that sense, we are competing with the late nights I used to spend reading transcripts and scheduling interviews as a consultant. We believe that Outset will expand the research market, making user insights faster and more accessible to various teams within the business.”
Outset is still in its early stages, but despite the imperfections and limitations of GPT-4, it has already achieved notable success with a well-known brand: WeightWatchers. According to Cannon, WeightWatchers managed to conduct and synthesize more than 100 interviews within a 24-hour timeframe. These results are now being utilized to propose a new framework for user segmentation at WeightWatchers.
Cannon also mentioned that they are presently collaborating with 15 enterprise insights teams from companies like Opendoor and other major consumer-oriented organizations to assist them in making more informed and expedited user-centered decisions than ever before.
Outset recently secured $3.8 million in funding, primarily led by Adverb Ventures, with contributions from Weekend Fund and Jack Altman, the brother of Sam Altman. The company intends to increase its workforce from four full-time employees to six by the end of the year. Cannon expressed, “We just completed our seed funding round, and our team is relatively small, so we are maintaining a conservative spending rate. Despite the broader economic slowdown, there is a growing demand for AI-powered tools in everyday knowledge work, providing us with another favorable trend. With our seed funding, low expenditure, and accelerating tailwinds, we are well-prepared to face any challenges that may arise.”
