What regular work activity has the highest impact on the organization in the least amount of time and effort?
I haven’t done any science on this, but anecdotally it sure feels like recruiting, interviewing and mentoring are all huge impact-per-time compared to technical stuff like writing code. I can think of people I spent a few hours convincing to interview who ended up making multiple decades of contributions to Microsoft, for instance. Encouraging good hires and then helping grow each other’s skills are multipliers.
Why then, are so many companies so bad at helping their employees recruit their talented friends?
I don’t know! There’s got to be some perverse incentive somewhere that makes these processes broken. I was asked recently by a reader to share a “war story” on this subject. (I think I have told this story before but if I did I can’t find it, so here you go again.) Today’s story takes place about twenty years ago, and I sincerely hope things have improved at Microsoft in the intervening decades.
A friend of mine, let’s call them B, who I knew to be a talented software engineer with great technical PM skills was looking to change companies; I happened to know of a team in devdiv that needed someone with their exact skill set, and so I submitted an employee referral into the system. B got an interview and accepted an offer, and I was very happy right up until a few weeks later when I got a package in interoffice mail. The package contained:
- A single sticky-paper gold star, like a primary school teacher puts on a perfect quiz.
- A flimsily-built off-brand miniature lava lamp knockoff. (For my younger readers: a lava lamp is a novelty lamp in which the bulb is below a conical closed glass vessel containing oil and wax; the lamp melts the wax which then circulates in a sort of psychedelic convection pattern. There was a lava lamp fad in the late 1960s and early 1970s; I remember my parents had a blue lava lamp when I was a very small child. There was a (very brief!) resurgence in this fad in the late 1990s and it was common at Microsoft for lava lamps to be given out as funny prizes. I had three in my office at one point.)
- A poorly photocopied note. Obviously I do not recall the exact wording of the note but I can very easily give you the gist. It was something like:
“You’re a recruiting all-star! Thanks for your successful referral. Here’s a gold star to put on your office door to let everyone know that you’re an all-star! Please fill out a survey on your referral experience at [internal site]”
Sigh.
They invited me to give them feedback and so I did. Rather than filling out the survey, I sent the head of recruiting a personal email. That of course is lost to the mists of time, but again, I can certainly give you the idea. It went something like:
Dear X, thank you for the off-brand lava lamp and gold star. As requested, I’m providing feedback on my experience of referring B to the Z team.
I do not require either a reward or recognition for successfully referring a friend. The reward is that I now get to work with someone great who will help us achieve our division’s goals. But if you are going to reward that behaviour, you could do a lot better than a literal paper gold star, a photocopied form letter, and a five dollar junk store lamp.
You could, for instance, pay out a bonus for successful referrals. You know better than I do that sourcing talent is expensive. You would pay in excess of 10% of the salary of my new employee friend if they had been sourced by an outside talent agency. You could pay employees for referrals at, say 2% of salary. We’d feel genuinely appreciated and incentivized, and you’d pay a fraction of what it would normally cost. A cheap lamp and a sticky paper star makes it seem like the company cares very little that I put in this effort, and I know that’s not your intention.
You know what I would find even more motivating than a bonus? Instead of a photocopied impersonal form letter from recruiting, you could encourage the hiring manager to send me and my manager a personal email that says “Thank you so much Eric for helping recruit B to our team; we really needed someone with that skill set and I am looking forward to working with them. I will remember that you helped out our team in the next performance review cycle.” The documented goodwill of a hiring manager in devdiv is not something I can buy with money, but it is valuable to my career.
I was young and naive. I expected a reply to my thoughtful, well-intentioned, solicited critique; I got none. I also never got another gold star from recruiting, so maybe that message landed; I don’t know.
I’d be curious to know if any of my readers have received similar “awards” that send the opposite message as was intended. Leave a comment if you have!
Next time on FAIC: I have not forgotten that I said I’d blog more about Bean Machine, but before I do that I’d like to share some thoughts on a Python library I’ve been developing as part of that effort. I am a novice Python programmer and I’d love to get some feedback from the experts and novices among my readers alike.
Fabulous adventures in coding 
While I haven’t had an experience like that, I can certainly relate to how weird it is that companies don’t work to ensure that their employees know how to interview people. While I’m not on the “manager track” or anything, I’ve been in various “team lead” roles that have on occasion involved interviewing people and giving my feedback to the hiring manager (since I’d be working closely with the person who ended up in the position). But nowhere, big company or small, has ever given me any indication or help in how to interview, what kinds of questions to ask, or anything like that. I’m actually *really* surprised that nowhere has given me any guidance on questions to be sure *not* to ask for legal or company-policy reasons (it’s really easy to ask about things like family when trying to get to know somebody new, but of course it’s really easy for that information to be used in a discriminatory way).
And “of course” developers know about lava lamps, they’re a well known hardware random number generator source 🙂 https://www.cloudflare.com/learning/ssl/lava-lamp-encryption/
One thing that both Facebook and Microsoft did very well by me was training and feedback on interviewing techniques — and in particular, how to avoid both the appearance and the consequences of unconscious or conscious biases.
I was quite surprised when I worked at a much smaller company how their interviewing standards were at the same time reasonably well-calibrated on the technical side, but lawsuit-prone disasters on the “eliminating bias against protected groups” side by comparison. And when looking at questions about interviewing on web sites I often see advice on practices that I would not touch with a ten foot pole.
I had forgotten about the lava lamp seed generator; I’m glad to know it is still making entropy!
Reminds me of a story about the importance of how you incentivize people. There was a town in the Netherlands where they were going to store some nuclear waste, the town was actually proud to accept it overall as they saw it as a greater good kind of thing, but when the state offered to pay them some amount (I think a few thousand dollars) it caused the level of approval to go down because now people saw it as just a cheap pay-off. If you’re going to give compensation, it needs to be sufficiently high as to be seen as having value, or your probably better off not paying at all. Hence why good economists understand that humans almost never make truly rational decisions, and setting up economies that way results in disaster overall
100% agree. I vaguely recall a similar story from many years ago — I think it was on Raymond Chen’s blog — about a software developer who gets called on in retirement to come back to do some feature work for their old team; the team was short staffed or some such thing and at risk of the project getting cancelled if they missed a deadline. They pitch it to him as “come do us a favour for old times sake” and he says sure, I’ll do it for free as a favour for old times sake. “Well, legally we can’t do that, so we’ll pay you your old salary while you’re doing the work.” Nope. Developer refuses the job. If the failure to get the work done is an existential threat to the product, then you either ask for a favour or you pay emergency rates to a professional to fix it. Other professions double or triple their fees for rush jobs and emergencies.
The point of this anecdote, and your anecdote, and my original post can be summarized as: pricing a thing in terms of dollars sends a clear signal about what it is truly worth to you. Pricing a thing in terms of intangibles like “help out your old friends” or “help out your division” or “here’s a unique, personal memento that you can’t buy for money” feels very different when there is no money involved.
This reminds me of a story in Freakonomics about a kindergarten that tried to deal with parents who showed up late to pick up their kids by introducing a fine. The number of late pick-ups immediately increased, as parents could now do this with a guilt-free conscience.
I don’t know if your email had a direct impact, but I can tell you that Microsoft now pays a referral bonus to employees. I was with Microsoft for about 5 years in the mid-teens and received a bonus of (IIRC) $1.5k for referring a friend who was then hired. Not the 2% of salary you suggested, but significantly better than a gold star and a dollar-store lava lamp.
Well that’s good to know.
You illustrate a good point: for the bonus to be seen as a *bonus*, it’s got to have meaningful impact. $1.5K isn’t life changing for someone making an MSFT salary but it definitely falls into the “meaningfully out of the ordinary” bucket.
The “out of the ordinary” factor is real. In my many years at Microsoft I frequently pointed out to management that past experience had trained me to expect a specific bonus payment on a particular date, and because of that, it was *priced in* to my morale; the only way to give me a genuine morale boost on bonus announcement day was an *unexpectedly* large bonus.
That only happened to me once, and I am 100% sure that the amount of embarrassment and expensive product recalls I saved the company by finding that .NET security hole and broadcasting how we found it far and wide throughout devdiv was ENORMOUSLY more than the price of the MX5 I bought with the resulting bonus. Every other year that I got a normal bonus it was a non-event.
The year I got a below-normal bonus was my last at MSFT.
Summing up: a bonus has got to feel like a special, valuable prize that makes a meaningful difference. Otherwise it actually drives *dissatisfaction*, and it’s rather perverse to be spending money on increasing dissatisfaction.
In France, nowadays, it is common to see service companies giving a bonus to their employes which help to an actual recrutement of one of their friend, for instance.
The bonus is usually something like 300 to 500€ (out of Paris, where 500 would be the 25% of a beginer salary).
But out of sercice companies which hire lots of programers to place the in large companies, I never see such a thing.
I work for a large comagny, and would probably get a brief thank you is I would help to get a friend hired by the company.
Happy to see you back, Eric.
The biggest beef I have with my current company’s recruiting is how _slow_ it is sometimes. They’ll have someone come in for interviews, and then just sit and hem and haw and wring their hands about some small unimportant thing, or want to “see what else is out there” and then, surprise!, the candidate takes an offer somewhere else.
I have had the most difficult time convincing them that if you’re interested in a smart and talented person, you have to move _fast_ because you’re competing against every other company they’re applying to and if they really are a great candidate, they will have their choice of company.
I like Joel Spolsky’s articles on hiring developers where, after you’re done with the interviews everyone makes a firm “hire” or “no hire” statement, and if they are a “hire” you make them an offer right away. The sad thing is that one reason I work where I do is that they called me on the drive home from my interview to offer me the job. I was impressed with them already and that decisiveness sealed the deal for me. As the company has grown in size that’s another thing which has been lost.
Happy to be back!
Absolutely you have to move fast. When I was an intern back in the 1990s, Microsoft made it very clear that I would have a job on the VB team if I wanted one, which was very nice of them. At FB we go even farther and actually make offers even if the intern isn’t graduating for another four months.
Back when I was at MSFT, there was a nominal referral bonus (maybe $1k). Occasionally it would double or triple for folks with the right skills. I had a friend who was looking for work, so I called up a recruiter I knew (the one who had recruited me) and gave her his name.
We hired the guy. A little while later, I called my recruiter friend and asked about the bonus. She pointed out that I’d never gone to the “refer a friend” site and filled in the form. So much for that one.
Another time (another company). I referred a friend (who had been at MSFT at the same time as me) for a job with my new company. He was a perfect fit for the position they were looking for. It was a very quick hire. But, right as he was hired, someone quit and they decided he’d be a great fit. He literally told them that “yes, I have experience in that technology, but I really want to move out of it, that’s why I left my old company to come here”. Obviously, they ignored that plea. Not quite 3 months later he was gone. When I enquired about the referral bonus, I was shown the fine print that the new hire had to last three months.
I’m two for two.
Back in the day, when I ran an IT strategy practice for a top consulting firm, hiring was the most difficult and often expensive aspect of the work. Our process was roughly as follows:
1) External recruiters scour hundreds of resumes from outside sources with detailed requirements;
2) Internal Recruiting triage apparently well qualified external referrals as well as potential in-house transfers;
3) Senior Consultants (5-10) conduct target, preliminary screening to assess candidate resumes for educational, technical and experiential fit. Forwarding initial recommendations to Junior Managers for concurrence;
4) Junior Managers vet the candidates for in-depth phone screens by mixed team comprised of cross-disciplined Consultants, Senior Consultants, Junior Managers and Managers.
5) Managers accept hand-off of a list of vetted candidates, resumes, proposed phone interviews using subject matter matrix based up experience level, domain of expertise, depth- and breadth of experience;
6) Relevant Managers review proposals and work with staff and management to develop a list of follow-up questions and new behavioral- and experienced-based questions, and decide which critical metrics will be used to establish in-person interviews;
7) Managers work with Senior Management to approve/disapprove of borderline potential hires, returning rejected resumes to Internal Recruiting for follow-up with External Recruiters and internally sourced candidates;
8) Assessment and Interview teams work with management to conduct up to 5-6 formal interviews with an ultimate goal to screen out all but the top 20 candidates;
9) Well qualified candidates that pass the rigorous phone screens with a specific score by the A&I team are referred to HR for a thorough background screen, including initial references contacts;
10) Fully vetted (HR) candidates are scheduled for brief 30 minute meetings with hiring 2-3 managers, including the hiring manager;
11) Internal Recruiting receives a list of in-person interview Candidates (typically 20) and setup and coordinate the quarterly in-person interview days;
12) Quarterly, Senior Management (~two), key managers and 6-10 business domain/technology experts converge near a major airport and conduct all day interviews from Wednesday-Friday, with mismatched candidates being eliminated as early as possible in the in-person interviews;
13) Thursday evening, the team meets for information sharing, question raising, air clearing and re-prioritization of the rating matrix;
14) The eight-to-fifteen (typically) candidates meet with management for detailed questions and rating (we developed an Excel spreadsheet with ~20 hiring criteria with areas ranging from “ability to apply research properly, “intellectual curiosity” and “ability to formulate solutions outside of expertise”);
15) Two groups of final meetings are held to consolidate written evaluations, fit forms, concerns, etc. First, the technical and business teams provide their feedback to Management and Senior Management. Next, Senior Management and hiring managers discussed positions and offers;
16) The Managers had to accept responsibility for any candidate targeted for their team, including getting them “billable” at the level of hire. By the end of this process, any candidates not targeted for a “P and L” was either rejected outright or, if applicable, to other practices within the organization for hiring;
17) And, finally, Senior Management met with managers to finalize P&Ls, titles, office locations, potential start dates, etc.
If this process seems onerous, that is because it was. At the time, we estimated that it cost about $150,000 per consultant in hiring process and on-boarding. Further, consultant seniority, pay-level and bonus were all based-upon net revenue. Hiring a poor candidate could hurt everyone in the practice. The process worked very well; after recruiting some 200 candidates over a five-year period, only a handful (~1%) were deemed unsuitable and released before their first full year.
The distribution of the recruiting responsibility was an excellent way to groom everyone in the practice from Junior Consultants to Managers in hiring processes, tools and techniques. And of course, mentoring, internships and other “Human Capital Optimization” processes were a great asset in this regard.