AI-Driven Expert Network Seeks to Disrupt Industry

NewtonX wants to shake up the expert network industry with new technology and new business models.  “We make access to expertise as transparent as possible, as precise as possible, as fast as possible and as inexpensive as possible,” said co-founder Germain Chastel in an interview.  “Experts should be available to as many people as possible.”

The firm relies on a proprietary knowledge graph (a structured dataset of experts that draws on a variety of sources, often unstructured) to automate searches for relevant experts.  [Kensho’s version of IBM’s Watson used a knowledge graph to find historical market reactions analogous to current events.]  “We are a search engine, not an expert network,” said Chastel.  “We draw on a much wider pool of expertise than an internal database.”

As a search engine, NewtonX draws on three types of data: 1) third-party real time data obtained via the web; 2) asynchronous third-party data primarily consisting of purchased industry listings; and 3) internal data on experts collected by NewtonX.  Web sources include not only LinkedIn but question-and-answer site Quora, Google Scholar and others that might identify experts and their contact information.  These are supplemented by purchased lists and the proprietary data on experts collected directly.

One-on-one consultations represent only 20% of the services used by clients.  The majority of usage is comprised of short custom surveys allowing users to pose key questions to a few relevant experts.    Surveys also tend to fall into consistent types of queries, such as competitive assessments or ‘voice of the customer’ surveys, which has allowed NewtonX to create standard templates for certain categories of customer inquiries.

Target clients are corporates, consulting firms and private equity firms.  “Our goal is for access to expertise to be mass market – this industry should not be dominated by hedge funds,” said Chastel.  Clients include the top three consulting firms.

Because many of NewtonX’s processes are automated, the firm is able to pass along cost savings to clients.  Whereas the average consultation fee for the expert network industry is around $1,200 per hour, NewtonX charges $800 per hour and it bills in quarter-hour increments.  Surveys are charged based on the time required for each expert to complete the survey.

The firm appears to have adopted industry practices for compliance such as annual certifications from experts, interactive compliance training for experts and a database of employer guidelines or restrictions on employee participation in expert networks.

NewtonX was co-founded in early 2017 by ex-McKinsey consultant Chastel, COO Sascha Eder — also a former McKinsey consultant — and CTO Anuja Ketan, previously CTO at Zillion, a healthcare marketplace.  In September 2018, the firm raised $3 million in seed funding from venture funds XFund and Third Prime Capital.

NewtonX grew from three employees and one client to 25 employees, seven- digit revenue, and more than 20 clients from 2017 until its financing in October 2018.  It says it has 50 staff currently, of which 40 are based in New York and the remainder in San Francisco, Europe and Asia.

Our Take

NewtonX is not alone in using automation to try to upend the expert network industry.   UK-based proSapient, totaling over sixty staff, is using AI to source experts as does UK-based, which recently received angel funding.  Incumbent AlphaSights, the second largest expert network, is recruiting data scientists to “build new products, improve search and discovery, and perform predictive analysis to improve new client conversions,” according to its job descriptions.

The growing use of artificial intelligence potentially introduces new risks to the already compliance-heavy expert network industry.  As expert networks collect more personal data via social media and other sources, they potentially run afoul of burgeoning privacy laws which are propagating quickly in the wake of Facebook scandals.  LinkedIn’s pending lawsuit seeking to curb access to its publicly available data may also pose challenges for AI-heavy expert networks, depending on how the Ninth Circuit rules.  Nevertheless, it is inevitable that the expert network industry will become more automated, with the early adopters like NewtonX hoping to gain lasting competitive advantages from the transformation.