20 Jun 2018

How do Expert Networks operate?

Photo of Max Friberg, CEO at Inex One
Max FribergCEO at Inex One

Expert Networks are matchmakers between organizations that seek expertise and the individuals that possess it. There are 5 different expert network operating models. Their work is similar to traditional headhunting, with the difference that Expert Networks recruit for 1-hour consultation sessions rather than for full-time jobs. Experts sometimes get recruited for surveys or meetings, but the 1-hour phone call is the dominant service.

The industry as we know it today emerged with Gerson Lehrman Group (GLG) in 1998 and boomed in the early 2000’s with the emergence of additional players (including Alphasights, Third Bridge, and Guidepoint). It then took a double beating: first from the drop in M&A activity of the 2008 financial crisis, and then from a series of insider-trading scandals in subsequent years. In the recovery years since, however, the industry has enjoyed healthy growth and profitability. These are the five main expert network operating models.

Expert Networks: “The standard model”

Expert Networks rely heavily on the smarts and flexibility of their junior employees, who go great lengths to find suitable experts for their customers. This requires hours of manual work, googling, and trawling career databases such as LinkedIn, Xing and Viadeo. In addition, Expert Networks build internal databases of previously used experts that might be used again in the future (although the value and legitimacy of such databases is in question). This business model is nevertheless proven successful, so let’s call it “the standard model”.

Expert Network operating models: The standard model

Expert Networks: “The standard model”

In recent years, four adjacent business models have emerged:

  • Machine-driven Expert Networks

  • Expert Q&A companies

  • DIY marketplaces

  • Crowd-funded expert calls

Please note: This article is based on research conducted in May-June 2018. It is meant to be comparative and informative. In case you find any details to be incorrect or outdated, please contact us.

Alternative 1: Machine-driven Expert Networks

Expert Network operating models: Machine-driven Expert Networks

Using our proprietary algorithm, we precisely identify the top experts [...]

- NewtonX

[...] experienced managers and sophisticated computerized systems to scour dozens of sources [...]"

- Chime Advisors

Our machine learning methods help our analysts provide more relevant experts.

- ProSapient

The machine-driven expert networks include Xperiti, NewtonX, ENG llc, CleverX, Techspert.io, Prosapient and Atheneum.

Machine-driven Expert Networks identified that the time Associates spend googling for experts is a bottleneck to their business. Their solution is to automate this work, or parts of it, to increase the “yield” per employee, i.e. any combination of (1) finding more experts (2) finding experts faster, and (3) finding more relevant experts per client request. We started out with this ambition at Previro (an expert network that we started in 2016). Alas, we soon ran into roadblocks. Scraping and applying statistical models (“machine learning”) to data on individuals sounds straightforward, but it’s trickier than you would think.

Scraping is difficult: (1) there is a multitude of data sources (e.g. career databases, company databases, telephone directories and professional forums), (2) you’ll need to update your scraping tools as UI:s change, and (3) you’re typically not allowed to scrape databases (see LinkedIn’s policy and a recent ruling). Anyways, let’s say you get hold of a comprehensive, detailed, and up-to-date dataset:

Machine Learning is the real challenge: making sense of the large datasets that you have – and hopefully continue to – collect. You need to figure out which data attributes correspond with expertise on a given topic (e.g. the niche hydraulic automotive component that your client is looking for). There might be a comprehensive technical solution to this, but it will not be simple.

Nevertheless, the Machine-driven Expert Networks are on to something very interesting. They could be very successful if they manage to create machine-generated gross lists, that are then fed to analysts for human vetting.

<u>Update Feb 2019</u>: Check out this deepdive in AI and expert networks.

Alternative 2: Expert Q&A companies

Expert Network operating models: Expert Q&A companies

The Expert Q&A companies such as Quora, Answers.com and AskWonder.com let you ask a question, and get a vetted reply from experts. The service is best for non-urgent, non-confidential research, and helps users get an oversight over what’s out there. Whereas Quora and Answers are mostly B2C, I know management consultants who use AskWonder early on in projects, as an alternative to googling themselves. The limitation is that you rarely get beyond what is publicly available on Google. The benefit is that you get all publicly available data collected and explained to you by someone who is typically used to industry terminology. This can save you considerable time, and at $75 a pop, it’s affordable for most organizations.

There is debate whether the staffing model is malfunctioning, but all in all, the service is cool and looks like it’s on a good track.

Alternative 3: DIY Marketplaces

Reach 100s of 1000s of experts with a single algorithmically matched Inquiry and expect 5-12 relevant Expert replies within a few days.

- Zintro

Expert Network operating models: DIY Marketplaces

DIY Marketplaces open up their internal databases. They let customers browse for experts and/or submit requests that experts bid on. The marketplace is at the core of the gig economy, writes NYU professor Arun Sundararajan, as it exists in many other industries (e.g. finding freelancers to build a website, or to cut your lawn).

The task to be performed is typically generic/non-specific and can be performed by many of the gig workers on the marketplace. Requests for expert calls are typically highly specific, and a handful of individuals could be significantly more relevant than others. Consequently, Expert Network DIY Marketplaces differ from other marketplaces: they apply algorithms to match client requests to experts (gig workers) on the marketplace. They typically do custom-recruiting of experts in addition, to keep the marketplaces updated and filled with highly specific experts.

DIY marketplaces include Zintro and Clarity.fm (specialized on helping startups). Danish company Mintebi started out as a DIY marketplace, but soon pivoted into the “standard model” and changed their name to Nordic Knowledge Partners.

Alternative 4: Crowd-funded expert calls

What if you want to speak to an expert, but you don’t want to pay the fees associated with an expert network? Perhaps your research is not confidential, and you don’t mind sharing the conference line with other researchers? Many “standard model” expert networks offer such open expert calls as a side business, but Slingshot Insights has specialized in it as their core service.

How will the market for Expert Networks develop?

The market for expert networks is the market for experts-on-demand. It stretches far beyond the “standard model” into alternative business models. Besides, there is a parallel and thriving market for legal experts (“Expert Witnesses”), with companies such as IMS Expert Services and The Expert Institute. Industry analysts and incumbent players expect it to continue growing. In my view, this is driven by simple supply and demand:

Demand is growing. Global value chains are ever more complex, and companies specialize into niche segments. If you’re looking to invest in a company that exports niche components around the globe, you probably want to research the local market conditions (competitive dynamics, regulatory trends etc.) in at least the largest regional markets (e.g. China, Germany, and the US). Speaking with local experts is a very efficient way to do this.

Supply is growing. More people make themselves available on online career directories, and people are increasingly open to “gigs”. This makes it easier for expert networks to identify and recruit global experts.

There is merit to all expert network operating models, and there is likely room for multiple players to thrive.

Expert networks complement other services

We will round off with a last image, showing what the Inex One team sees as a central trend in investment research. With increasing industry complexity, more sources of expertise are required to maintain a given level of industry insight.

Take Sweden as an example: in the 1950’s our main exports were iron ore, paper, and pulp, to Northern Europe. Industry analysis and investment research was mostly done in boardrooms of, well, middle-aged men. Since the 1980’s, rapid industrialization and growing global trade in subsequent years have been some of the factors behind the success of global management consulting firms. Management consulting is a complement to having full-time industrial advisors/employees, and the industry sees strong, sustained growth. In the last decade, experts-on-demand has grown rapidly. The industry is in turn a complement to what came before it, as it is used by both consultants and regular corporates.

Inex One helps you gather all your expert-on-demand-sources (i.e. your Expert Networks) in a single interface. We think it’s a logical continuation to the trend, and a great complement to other sources of expertise.

Click here to get started with Inex One.

Expert networks are a complementary service in investment research

Expert networks are a complementary service in investment research