Fixing the High-End Hiring Process

I had an interesting conversation with a serial startup CTO yesterday. He's had some experience in the recruiting-website world before - think Monster, or HotJobs, or CareerBuilder - and despite success at all tiers of the market, believes the system is still broken.

Together we arrived at the following categorization of the job-placement system:

(1) There are some jobs for which anyone who meets minimum requirements can ably perform the job duties. These are "square peg in a square hole" kind of jobs, usually which entail no performance expectations or bonus. Can they answer phones? Can they operate the machinery properly? In this category, a Human Resources department or a basic posting to major jobs sites can pretty well find people to fill these roles.

(2) Then there are jobs where performance, motivation and/or competence are at a premium. You need someone who won't just do a job, but will oversee something important. A salesperson, or a department head, or a researcher. These qualities in the employee will not be immediately apparent from a resume, but a more detailed background check, reference check, and interview structure can usually weed them out sooner or later. Looking at things like sales track record, or doing case interviews for consultants, or even internships, can do an acceptable job of filtering your job candidates. These people will usually already have pretty good jobs, so it can be hard to reach them and harder to entice them to jump.

(3) Finally, there are jobs for which the judgment of the employee, or the extreme motivation for them to go above and beyond the call of duty, will materially affect the outcome of the company. Founders and cofounders, executives, top sales or business development people, top strategists, partners at services firms. When you hire for these roles, the value you're getting - i.e. the results of your choice - won't be necessarily apparent for months or even years after the hire is made. Someone gets picked based on relationships or on faith, and the company will rise or fall to a significant degree based on how awesome they are. Finding them is usually a matter of an Executive Search recruiting firm, a network of high-end professionals such as the personal networks of angel, VC and private equity investors (if they're good), or identifying them through industry experts. And hiring them usually requires significant equity or compensation packages that reflect the magnitude of importance they have and the risk they would be taking by joining, for which they must be given sufficient upside.

For fans of marketing and economics, these 3 categories are pretty able illustrations of the differences between Search Goods, Experience Goods, and Credence Goods, respectively. Depending on how critical that employee is, hiring one may be like buying a head of lettuce (you know what you want and you go get it), like taking a tropical vacation (once you go, you know if it was worth it), or like choosing a surgeon for your knee (their skill may not be manifest for 10-15 years).

My friend's point was that the candidate-search process is pretty efficient at this point for category 1. It does a poor, high-friction, high-variance job of #2. And for #3, the market is absolutely broken. Consider tech startups, for whom the first few engineering or marketing hires are crucial, and the difference between a merely good and a great choice can be a productivity factor of 30. That's the reason Google and Facebook are so fanatically picky about its engineering hires. If they choose a Credence Good like a Search Good, their quality - and their image - will quickly degrade. We see this even with Google these days, from some snarky corners.

The reason this system is so broken is that no HR person, and no recruiter, can ever really discern the fine gradations of skill, judgment, mental agility, and most importantly motivation to go the extra mile, that are the difference between those good and great hires. Entrusted to a person who has never held a job anywhere close to the target (i.e. an HR person or recruiter) is the duty to flip through hundreds of resumes, pick some to phone screen, and pick a few of those to refer on. How many false negatives are there? How many false positives? It's not an HR guy's fault that he can't replace the sophistication of a hiring manager for Category 3 jobs. That's not his role; he's been hired to manage the process, take care of the basics (background checks, verifying facts, reducing noise), protect the company from liability, and leverage the hiring manager's time. Same is true of recruiters, except for the liability part.

We believe someone has yet to bridge this gap - the "high-end job search" market. TheLadders filters for high-salary (i.e. $100k+) jobs, but there's no real filter beyond that. IvyExec and other startup sites aim for high-quality jobs but end up being a somewhat-curated job posting board. For certain disciplines - say, technology - there are sites like Dice, but they suffer from a lack of quality filter. There are top recruiters who handle financial services front-office roles - see Glocap or Mercury Partners - but while they may do better, their process takes a long time, still has plenty of noise, and costs 30% of a year's salary or more for a successful placement. That's huge market friction!

So what are the characteristics of a better solution? Let me brainstorm here.

1. The site must filter only for top-quality people. People whose motivation to stay late, come up with new ideas, build consensus, get along with everyone, and build a team or sales pipeline or set of partnerships, is all unquestioned. Such a filter will always be very subjective and lots of people at the margins will try to game it.
2. The site must also filter only for jobs that will be interesting to top-quality people. Jobs whose level of responsibility or criticality gives sufficient challenge, opportunity and reward to be worthwhile considerations for these top workers.
3. The site cannot unnecessarily flood or spam its workers. Any outreach must be bespoke and conservative in the choices. Of course, if the workers are actively searching opportunities, that's fine, they're only spamming themselves.
4. Likewise, the site cannot allow workers to find opportunities and submit themselves willy-nilly without some quality filter in place. Signal to noise ratio for applicants going to a hiring manager must be very high.
5. This site would replace the role of an HR person or recruiter, at least at the filtering level. HR will still be necessary to make sure the person isn't faking a degree, hiding a felony conviction, or that the search process itself meets all the legal requirements to avoid liability for the company.

How would we construct such a site, and seed it with people? How would we ensure such filters? The goal here is to eliminate the prohibitive amount of human time spent on intermediaries in this process, to make it cheaper while still being very profitable for the site itself. I don't have a firm answer, but my ideas include:

• Personal testimonials must be much more important, and given with significant anecdotes. Something like the review approval system currently used for CULPA, Columbia University's professor-ratings website, where trained admins scan every submitted professor-review by hand and grade it on a 10-point scale. Surely that's a task you could give to a Mechanical Turk-style group of high-school-grade people.
• Anecdotes from a worker's professional history, personal hobbies (in the evening, do you watch TV, or code projects for Sourceforge or Github? Have you ever voluntarily stayed late at work because you had an idea you just had to work on?), etc.
• A User's ability to invite trusted connections into the network depend on how highly rated they are, and how many ratings they've received, in the form of recommendations from professional connections.
• You could seed much of this with LinkedIn. The problem is, LinkedIn isn't just for high-end professionals anymore, and the signal/noise isn't maintained as carefully as it should be. Evidence: the LinkedIn Open Network. People with 1000+ connections. So you'd have to be careful about what you pull in, but it would be the only source I'd use to suggest initial connections to people.
• You'd have to start by targeting very small verticals. Probably communities who all know each other and have a high regard for each other. Investors. Entrepreneurs. Top/famous engineers. Scientists and researchers in particular fields. You might need a landing page / curated site for each discipline or category you launch.
• Job postings would have to have qualifications or detailed, subtle screening questions as a pre-qualifier for submitting yourself as a candidate. Maybe a startup hiring a Chief Revenue Officer doesn't need the CRO from Amazon or eBay, but "have you ever managed salespeople before" is a worthwhile screen. People who answer falsely get marked down by the hiring manager for wasting their time, and may eventually get shut out of the community if they're too disruptive/greedy about applications.
• Startups, in particular, are always looking for top people. They might be the easiest target market to seed this with, because the cost/benefit to posting their needs on the next new job site is very favorable.
• You could also use professional associations - think the AMA, ABA, NVCA, ACG, etc - as marketing partners. Who knows what they would require for helping out, but if you can help those organizations become a central part of the job market within their industry (or niche), they'll probably understand the value that coems from that and from being a partner.

Monetizing this is simple: you charge big bucks, probably several grand, for a successful hire. A headhunter firm might get $10-20k as it is, even for employees in the "Experience Good" category. Charging $4k for facilitating a crucial hire will probably be a price a startup is happy to pay. The human overhead will be in the form of moderation, but there will still be a lot of work in that regard.

Could this work? Depends on the network of the people starting this, I suppose. But as a model, getting paid as a result of disintermediating human brokers is a very successful tech startup model. eBay disintermediates the garage sale; Craigslist, Amazon, even Paypal generate network effects from disintermediating brokers who otherwise add no value. The key is how to make the user experience better than the human-moderated thing, or close enough so that price matters. In terms of creating a community like this, Quora has done a very good job of being fanatical about quality for user submissions, and having the community enforce such standards. Maybe a startup team exists out there that can recreate such a community for high-end job hires.