It wasn't a marketing pitch, it was interesting.

If you want to incenvitize some behavior you can offer an extrinsic reward. Certainty in an extrinsic reward causes overjustification.

An uncertain extrinsic reward seems to mute the overjustification effect increasing the intrinsic motivation within the process, while still affording some of the incentives that you're looking for.

When we are outside of the process we weight the certain reward more highly than the uncertain reward. But a certainty in the reward means we lose our intrinsic motivation within the process and replace it with direct motivation towards the goal, for instance you might stop doing something because it's not worth the 2 dollars you'll get. An uncertain reward lets us stay motivated within the process. For instance, something might not be worth the 1 or 2 dollars you will get, but you might be motivated to stick it out till the end just to see whether you'll get 1 or 2, or because it's fun, and you're not necessarily going to get 2 anyways.

When you can abstract a process down to "Do x, get y" your enjoyment of x can get caught up in your value of y. When it is "Do x, get y or get z" then your enjoyment of x might stand alone, because you can't relate it specifically to y or z.

If it's worth it to do x for y, and z is more, then you might be interested to see if you get z, so you be excited to do x to see if you will win z, and be satisfied with y if you don't.

If it's only worth it to do x for z, then you have to take into account that you might get y. If you might get y, and it's not worth it, then there's some cognitive dissonance, one way for you to resolve that is to say that you just enjoy x even if you do get the lesser prize.

Either way, you're not thinking about abstracting the whole process into a routine to get y. It's either y with occasional exciting bonuses, or it's doing x because it's enjoyable and sometimes getting fair compensation for your time.

Think of it this way. Think of an employee paid hourly, $20/hr. They can come to work for as long as they want, and they have a budget of $400/week. They work 20 hours a week, except on the odd occasion when they want to save a bit more or splurge.

Then consider an employee paid hourly $15/hr. They can also come to work for as long as they want and have a budget of $400/week. The difference is there's a 25% chance every hour that they will get an extra $20. They're going to feel differently, If they get $400 in their first 15 hours, they might be excited. They might keep working and make themselves a bit of extra cash, or they might go home early. If they don't make the bonus at all, they know that's a possibility and they will work the 27 hours they need to make ends meet.

On average both employees need to work the same amount to reach their budget. The second employee is more likely going to work more than 20 hours per week on average. They've already resigned to the fact that 27 hour weeks might happen, but that will rarely happen. They will justify that $15/hr is OK. The rest is gravy. They're lucky to get it. The job's not so bad at $15/hr. The bonus is nice but not something you have to live off of, 27 hours isn't a bad work week in the worst case.

Both employees have the same average wage, but I think the second is more likely to work more hours over the course of a year. If the budget of the first person were to shrink or grow, they would more directly adjust their schedule than the second employee would. If the second employee were to actually like their job, they would be more motivated to go to it just to see if they could get more bonus regardless of what was needed.

In some ways you ironically feel more in control of the latter scenario. I don't need to go into work this week (because I earned a lot of bonus) but I want to because I choose to see if I can earn more bonus, if not, the job's not so bad anyways.

On the other hand it's easier to sell someone on the idea of a job that's $20/hr than a job that's $15/hr and randomly $35/hr.