The Rewarder – a better kind of bonus system
Conventional bonus or compensation-setting exercises are (and here I trust you will agree I am being rather understated) pretty much universally unsuccessful and disliked.
At LoveMachine, we use a bonus/profit-sharing system that is simple, extremely accurate, inspiring, and empowering. And we’d love to come and install it in your company – just ask us.
The idea is simple: Instead of a management-led exercise in reviewing and ranking team members, simply give everyone in the company the same amount of money, and then tell them to give it away to everyone else, in any way they think makes sense.
That’s madness, right? Actually, no. It turns out that if you are willing to take the plunge (which you can easily do by starting with a small amount of money assigned to the program), and if you do it the right way, this process is statistically optimal in valuing team-members. It can’t actually be done any better. By engaging every brain in the company in a fast collective evaluation of everyone else, you maximize the ‘computation’ allotted to the bonus process. No small group of managers, however well-meaning, can approach the accuracy of this process, because they can’t possibly know the details of how everyone contributed and supported each other.
Additionally, even if this process was no better at finding high performers (which it is), it would still be an enormous win because it removes management from the process of evaluating bonuses. This is powerful and inspiring. Each person receiving a bonus knows that the amount they receive is quite literally the collective opinion of the entire company. Their boss (or bosses boss) has nothing at all to do with it. It’s a very hard number to argue with. You have to accept it as truth. There is no complaining about management. Separating the management function of valuation and review from that of leadership and mentoring is one of the single most powerful cultural changes a company can make. Historically we have needed to have people watching over each other and trying to maximize the productive output of reluctant employees. This is no longer the case in most companies. Most jobs today are jobs that are taken by people actually wanting to do the job, with a passion for the success of the company, and a sense of ownership. In these companies the optimal strategy is to have leaders inspire, mentor and suggest strategy. It is highly sub-optimal to empower managers with the poorly-executed and often-abused task of valuation. The solution is to create great transparency, and then let everyone participate equally in value-setting.
Lest you worry that this won’t work or scale, we did it for a number of quarters at Linden Lab, typically distributing $500-1000 from each of 100-200 people. At LoveMachine, we use it to distribute a ‘bonus’ amount of compensation equal to about 40% of total comp. Combining the process with the job-based / transparent method we are using at LoveMachine gives us what we think is an almost perfect way of paying each other.
FAQ
Q: How do you keep people from cheating and just trading with their friends?
A: You eliminate this by empowering a randomly chosen set of employees to see ALL the distributions being made – who is giving how much to who else. If you know there are many eyes on the data, you will not cheat. As an additional nice feature, our system doesn’t let individual auditors actually see the details of their own amounts received. So there isn’t any risk that being an auditor will improperly influence other people’s behavior.
Q: What data to you make available to everyone at the end of the process?
A: How transparent to be about results will vary by what sort of company/culture you are in. Generally there are two choices. One choice is to show everyone a graph of the final results (how much the #1 person received, the #2 person, etc), but not the names of the people. This lets people see where they ranked, but not what others received. The other choice is to fully publish how much each person in the company actually received. You don’t let people see where there money came from, though – the process of giving needs to be anonymous to work.
Q: Do popular people get more that they deserve? What about highly visible jobs verus ‘behind-the-scenes’ roles?
A: Surprisingly, no. The opposite actually tends to happen – the rewarder process ‘discovers’ the unsung heros. Groups also tend to be rewarded in the right proportions – you will see the correct mix of groups in the top-10, for example.
Q: Do the execs or senior managers get overpayed?
A: Actually, they tend to get underpaid by the system, at least initially. Again – the rewarder tends to discover/reward more ‘hazard pay’ – people doing things that are great for the company but not glamorous – which is exactly what you want.
Q: How big does the company need to be to use it? Is there a size beyond which it is no longer effective?
A: We did the first LoveMachine rewarder ‘run’ with about 10 people, and it’s been run at Linden Lab with about 200+ eligible people. We don’t think there is any uppper or lower size limit.
Q: Do people tend to give to their closest friends/cliques?
A: Actually in practice people give ‘farther out’ in the network – to jobs different than theirs, for example. We think the reason is that people tend to feel like they have already valued/rewarded their immediate teammates day-to-day, and they elect to use the rewarder to reach farther out to recognize people they don’t know but that they feel are vital to the success of the company.
Q: How often to do you do it? How much time do people have to decide?
A: Quarterly or Monthly, depending on how rapidly projects are getting done. You should match the rewarder interval to the typical scale of projects, so everyone is able to value big deliverables when they look back over a period of time. As to the time to decide – it’s up to you. There isn’t value to a lot of deliberation, because your first ‘gut’ reaction on how to hand out money is the right one. You don’t want people to talk to each other or spend much time thinking – it breaks the system to do that. A typical approach is to give everyone 1 or 2 days to decide how to allocate their amounts.
Well it’s definitely madness–i wouldn’t have come up with it, but since someone else has and it’s already proved successful, i was trying to thin about *why* it works. I wonder if maybe it’s three things: (i) it sounds goofy, but it is way more fun to give than to receive–some kind of deep pleasure is triggered by giving a gift to a loved one or valued colleague that exceeds accumulating more stuff for ourselves by an outrageous ratio; (ii) i think the economic term for this one is “endowment effect” but i’m not sure; at the moment, i am paid about three times more than my co-workers one pay level below me. Let’s say that through the “Rewarder” scheme, i give one of my junior colleagues $5,000. That’s a nearly 10% boost to his annual gross, but it only cost me just under 3% of mine annual gross (plus they pay less taxes on that amount); and (iii) there’s a strong selection bias i suspect–a company that does things like Rewarder might attract the very best people–at the very least, i would bet that it discourages quite strongly the worst. The best people i think have no interest in keeping money given to them in a system in which they can distribute it to the star performers–plus…well i bet you the flow of those Rewarder dollars would be very interesting to diagram, because i can imagine that an excellent manager might give away all of his bonus but still walk out the door with the same amount (or more) because of “rewards” he/she received from others.
Doug – note that with the rewarder you aren’t directly giving away your own salary, rather you are given an equal share of an overall pool of bonus/compensation to distribute. I think I perhaps didn’t clearly explain… it’s extra money, not money coming out of your own pocket.
I love it. As I read it I just smiled more and more. It’s that kind of outside of the box thinking that makes a lot of sense once someone suggests it.
Now if only I worked for someone who gave bonuses.
I love it as well. I have found that you are very correct in stating the further-out approach to gift giving. While some teams may feel they have to contribute to the Alpha of the team or work group being anonymous allows the rewards to flow out to those that are actually achievers in a company. It also allows those rewards to flow out to those that may normally be over looked in the well being of the company but are much appreciated by their fellow employees. Good work I’m interested on how this deploys as I have a location in mind already.
“Separating the management function of valuation and review from that of leadership and mentoring is one of the single most powerful cultural changes a company can make.”
I love the insight and your solution. Democratizing the workplace is a worthy cause. Although I wonder if you’ve run into the issue yet of certain participants campaigning for more bonus share, or in a sense is that a part of the mindset?
I hope you’re writing a book. I would like to see some of your personal “code” out in the open.
[...] last post was about the rewarder, which removes (and optimizes) one of the main historical jobs of ‘Management’: [...]
lol a few of the observations many people submit are so silly, once in a while i think about if they seriously read the pieces and items before leaving a comment or whether they take a moment to look at the titles and submit the very first idea that one thinks of. anyhow, it is really relaxing to look over intelligent commentary from time to time instead of the very same, old post vomit that i typically discover on the internet
Howdy there,I obtain that your site is very informative and useful and we wonder if there can be a possibility of getting More content like this on your blog. If you willing to help us out, we would be willing to compensate you… Best wishes, Lashandra Kiss
Great idea. When I built my first consulting company, Strategos, back in 1993, we used this approach. My view: if 360-reviews are good, 360-compensation would be better. So at the end of each year, we’d give every employee 100 “beans” that they could allocated as they saw across all of their colleagues. Your share of the bonus pool would then be calculated based on the share of the total “bean pile” that you were rewarded by your colleagues. Although I was the chairman, my salary was entirely in the hands of my colleagues. I know it’s wacky, but it works.