Incentive Structures
Incentive Structures
No one knows what it means, but it's provocative. It gets the people GOING!
Incentive Structures straight up might be the only thing that matters. ever.
Ok Cameron, that cat got me in the door. What’s the deal with incentive structures?
Glad you asked!
Incentive Structures are the systems that exist, explicitly or implicitly, to return data to participants of a system. Some return data is useful. Some is noise. Some structures move us (and others) closer to our goals. Some don’t. All the structures and return data matter.
Incentive Structures are what create feedback loops and since, in some sense, we are just Russian Nesting Dolls of micro-feedback loop driven Cyborg Yeast, these are the things that shape our decisions all day, every day, at a neurological level.
So let’s build some from First Principles
Give me those tasty Brain Chemicals, baby
I’m not going to re-explain how serotonin, dopamine, and how our brains react to stimuli. This is the internet and you’ve read papers before and watched TED talks and seen videos so I’m assuming you have a popsci understanding of brain rewards. I mean I definitely, 100% know how all of it works, for sure, but if you’re like a biologist and have some thoughts or whatever I’d probably be open to hearing your thoughts just to see how right you are I guess. Either way, this leads to our first First Principle:
People generally react consistently to stimuli they’re exposed to ie Good Stuff makes them feel Good and Bad Stuff makes them Feel Bad
Good and Bad Stuff are relative and cultural/context/people specific, but hey models and being wrong and usefulness and all that. Next up? Our optimistic Tony Robbins function.
People learn how the rules of their game work over time and adapt their behavior to optimize for their own personal multivariable utility function
If you disagree that people can learn, then man do we have some things to discuss in another post. Just like, read this and this and work on a team with other human beings for like 15 minutes and you’ll see what we’re talking about here.
Ok cool. Are we done with First Principles? Just two? Yea I think so.
The Good Stuff
Attention, Time, Status, and Money.
That’s the good stuff.
That’s it. We’re done here. We can all go home now.
Just kidding. But really these are the things that incentivize people. Whatever the environment, whatever the mechanism, it all rolls back up to one of these puppies.
But wait, what about sex? You forgot sex!!! Well, first imo sex is just Attention with extra steps. Second, and MUCH more importantly, we’re trying to get some practical frameworks going here and while I guess sex as a professional reward is kind of a thing or at least it was 10 years ago?? it’s basically off limits for anyone that wants to keep an up and to the right trajectory (lol) in polite society and man the ethics and agency discussion of that are weird. Hard pass.
Attention
This is the most fundamental incentive reward. Even babies know it’s awesome. Mom talking at you in a sweet voice? So tight. Gold stars in school? yes please. My name placed on public lists showing my achievements? oh yea. Getting yelled at when I get in trouble in middle school? not ideal, but still better than being ignored for sure. Flame wars in the Youtube comment section? I’m just trying to feel alive again
Attention is, at it’s core, acknowledgement from others that we exist.
We need this to survive, grow, and learn. Things that help us survive self-perpetuate and work together with other aspects of ourselves to grow into complex systems and I guess this is how we invented Instagram? uh oh
But you can leverage this need in a non-manipulative way. I mean all of social interaction is manipulation to a degree, but we’re talking about Attention for net positive ends (assuming you have a standard Western moral framework whatever the hell that means) like publicly praising your team for prosocial behaviors like taking a data science course to support the group’s next project and not like social engineering everyone you know into buying you donuts.
Assuming responsible intent, if you want more of something, frame it to people in a positive way and if you want less of something, frame it negatively. The attention makes people feel that root acknowledgement. Specificity is best because we’re pretty lazy. Vague “good” and “bad” things are hard to emulate. Specific behaviors are much more easily copied. And the trick here is that people love (even most of the shy ones) when they’re singled out for being awesome, but NO ONE can handle being negatively singled out.
Our lizard brains just shut down out of embarrassment and we can’t learn a damn thing during that traumatic experience. Don’t single people out for negative actions unless they’re leaving the company for jail or something. Everybody already knows who messed up anyway, so no need to make them feel worse in front of their friends.
Finally, rewarding with Attention can be super deep and deliberate and take a lot of work like awards ceremonies, but it doesn’t have to and frankly the step function GREAT feeling of singular events might actually be harmful to sustaining positive behaviors over time. The compounding benefits of consistent, smaller, but omnipresent approval through things like deliberately having a positive conversation with a team member about their current project or a weekly callout in an all hands meeting for a job well done or a spot compliment for someone taking the initiative to teach a less experienced teammate something new are incredible to watch over time. Everyone is just trying to make Mom proud so why not support that need?
Lots of words, let’s distill it down.
Praise in public, punish in private
Positive Attention begets proactive behaviors. Negative attention begets avoidance behaviors
Small, consistent Attention is more impactful in the long run than big, flashy Attention (but both are probably useful)
Time
Time is a flat circle and also an interesting thing we can’t get enough of
The nature of this one deserves a lot more thought than I’m going to give it (because we’re operators here, not academics), but if you’re interested then knock yourself out, buddy.
In the military, you can’t pay people more (woof) so if you want to “bonus” someone for a specific behavior you can really only give them extra time in their day. This literally means like letting people go home at 4pm instead of 5pm or sleep in a bit.
And MAN does this reward motivate people.
You can get some stuff REALLY clean REALLY fast if people know that it’s the last task of the day. Like REALLY fast. Turns out that a lot of work can be done faster than it would normally be done in a timeboxed day.
That’s why people complain about traditional work weeks. Work is a gas and expands to fill the size of the container we give it. So why not let people set their own time schedules? I get the impulse. I like it. I feel it. I think the dynamic tension here is with quality control.
I’ve seen firsthand what can happen when a time incentivize and a quality standard…interact.
Take your classic Armory. The place where guns go. You have to count and check you have the right guns. Every time. Because lost guns are bad. yea?
So how this typically works is everyone is assigned a weapon with a serial number etched into it and whenever you take the weapons in and out, everyone has to line up and you have to sign some paper. Too easy. It takes a long time and then the Armorer has to go back and count the guns and check the serial numbers again after everyone leaves and also every night and a few times a day sometimes too. Kind of a pain, but again lost guns are bad.
Enter technology. Barring a discussion around the combat impact of RFID tags lol, someone figured out we can put RFID tags or barcodes on the weapons and just scan them in! So much easier than matching numbers to paper because that’s what databases basically do! tight. And it worked well. but then one day as Arms Room Officer I realized that the Armorer seemed to not have that much going on? Even with the scanner, it shouldn’t have taken like 5 minutes to count the guns. So I asked about it.
And I swear on my life this Corporal pulled open a drawer to show me how he had put a bunch of the barcodes together so he could just scan them all at once without having to walk around and those were what was being logged into the weapon tracking system. I immediately made him (together with myself because I’m not an idiot) check all the weapons the old way (no lost guns phew) and I terminated that “clever” system which he had only just started that day thank god. I cannot effectively describe how bad this could have turned out for us if I didn’t catch this. Anyway, getting some time back is a great motivator.
tbh now it kind of sounds like the Crypto Oracle problem which is funny
Be careful with this one. It’s powerful and it will take some failed attempts to get a time reward structure right. Don’t let it burn you.
Status
Status = Power.
We are relative creatures. We compare ourselves to other people all the time. It’s almost all we do. I mean…
We desperately try to get things and wear things and exhibit behaviors that show our hard earned higher relative status levels compared to each other. It makes us feel good to be “better” than others and it is useful to us as we accrue power and influence as a result.
Some places codify earned status differences with visible symbols as a reward like Amazon and the military.
This makes some people feel good and some people feel bad. It also influences the dynamic of every status differential interaction you have. But also no differential is bad?
Reward wisely.
Money
just remember that all wealth is relative too
Quick Note on Precise vs General Incentives
Some incentives solve specific problems and others fill in the gaps where you can’t build a precise enough mechanism to do the actual thing you want people to do. You need both tools to work together.
Take my old ride, the M2A3 Bradley Fighting Vehicle.
It has an M242 Bushmaster Chain Gun as its primary weapon system. It’s pretty fun to shoot. The problem though is sometimes you fight armored vehicles and sometimes you don’t. You see, armored vehicles are a precise type of problem that requires the M791 Armor-piercing discarding Sabot with Tracer round of ammunition to penetrate thick armor. It shoots a piece of tungsten really really fast through a bad guy’s hull. v cool. but also not as useful for other stuff like trucks.
For everything else (the General case), we use the M792 High Explosive Incendiary with Tracer (HEI-T) round. This explodes on impact to produce an area of effect rather than a point impact to uh…complete the mission.
Different problems require different solutions.
Another, nicer way to think about all this is that performance bonuses provide precise incentives when they’re tied to the right metrics and equity provides general incentives for everything else.
We’re not so different you and I.
Quick Note on Short Term vs Long Term Incentives
Long Term incentives provide a reward in the future and according to nerds the value of money today is worth more than the value of money in the future. According to this nerd, the same is true for attention, status, and time as well.
Ergo with equivalently denominated short and long term incentives, people will inherently skew their behavior towards the short term reward.
So make thinking about the future more valuable than thinking about today (if you want people thinking about the future).
Also every short term incentive turns into the long term incentive with enough repetition. hmm.
What do I do with all of this?
Ok we get it. People respond to stimuli (not clever) and learn (still not clever) and you have four reward primitives in your pocket. But can we do something useful with this information maybe?
If you’re designing an Incentive Structure and want a process to follow - here you go:
Define the one Behavior you want more of
I want more interesting and serendipitous ideas from my teams
Decide the Good Stuff that makes the most sense for spurring this Behavior
Attention - Maybe not enough reward for something this hard?
Money - Probably too hard to properly calculate the right amount in the near term?
Status - Same problem
Time - ooh I like it. Seems relatively cheap, but better than a compliment
Build the one sentence mechanism
If you want, you can spend 20% of your time working on something you think will benefit the company that isn’t your normal function (oh hey Google)
Think of the worst person you’ve ever met. Now what would that person do with this new mechanism?
Maybe take that 20% time and just screw around? probably worse
Build a one sentence check to reasonably counteract that bad outcome
Whoever takes the 20% time option has to showcase a progress report of their work once a month to their reporting chain. Got em.
That’s basically all you can do. One by one. Then eventually you’ve built and nested and adjusted your feedback loops to form, storm, and norm a group enough to finally perform the stuff you want to happen. But also you gotta think about it as you go. and iterate (ugh). and fail and try again. There’s not too many shortcuts to learning about people. sorry not sorry
Look, some people think all of this is dumb. I agree with some of them, but definitely not this hilariously terrible HBR article.
Incentives can be short or long term, precise or general, paid out in attention/time/status/money, but fundamentally they just…do what they are structured to do (relative to the individual person they impact). Extrinsic motivation can literally move mountains so don’t ignore them. People react to them whether you want them to do so or not.
I mean, intrinsic motivation matters too, but a lot of the time intrinsic motivation is actually developed extrinsically through group cohesion events that form implicit social pressure to define your cultural values, intellectual and personal experiences that shape your internal locus of control, and exposure to ideas from others that ignite passion within you.
As a Leader, maybe think of intrinsic motivation within your team as the spackle that helps in the gaps between your explicit incentives. Critical sure, but you still need some bricks.
I hope this added value to your day.
Please share this with someone who might find this interesting!
If you have any thoughts or questions about this essay - Let’s Chat
To hear more from me, add me on Twitter or Farcaster,
and, of course, please subscribe to Wysr