The expert network industry was born in New York when two hedge fund analysts interviewed hundreds of physicians and published a healthcare report which no one wanted to buy but they had requested to connect with the best experts they had talked to. Don’t worry even if you have not heard of it before or still do not really get what it means — Most of us have experiences connecting our friends, colleagues, and neighbors with another for their expertise in a particular industry, market, or topic meaning asking someone for advice is a universal habit of human beings.
From the very beginning, leading players did not hesitate to call themselves as tech companies as they put a lot of effort into building expert databases and search algorithms. That said, the industry is being disrupted by new technologies absolutely including AI…. or at least some companies are claiming so.
While it is a fast-growing market that can be applied to various needs such as investment decisions to the feasibility study, one-time consultation to full-time placement, it has clear limitations from a management and investment perspective — labor heavy. You might say… what? Didn’t you just say it is a tech-driven industry? Yes… and no. You can start the business right away if you know anyone your client is willing to pay for their time and expertise but in order to systemize, you will need huge resources including building and managing a database. But still matching, the core of the business, heavily relies on humans through analyzing clients’ requests, extracting the keywords, searching and filtering experts. This is a problem from a management perspective because it is hard to standardize the service quality (we only focus on matching attributes of the service) and for investors, there is a big question about how to scale up the business. Some players have been very smart and succeeded in getting investment by promoting themselves as AI-powered, featured, etc.
While I am curious how these players really apply AI technologies to their business, I am somewhat suspicious that their “AI technologies” are still very primary improving other minor service attributes such as scripts and schedules rather than matching which is the key element of the service quality based on their service description or growth rates. However, there is a hope we can automate at least a certain part of the matching process, to begin with, based on knowledge graph from various additional sources in addition to a company’s proprietary expert database and AI recommender systems. Happy to share my findings moving forward.
