When to reward the goal, not the rate

Opportunity Habit will judge a reward on one of two things. Your action rate, which is the share of the chances you took. Or goals met, which is the share of the habits in scope that hit their goal for the period.

The first guide argues for the rate, and I stand by it. This one is about the second bar, because the honest answer to “which should I use?” is not “always the rate”. It is “the rate first, and then the goal”.

Why there is a stage where the goal is the better bar

The case for rewarding the process is that a result is partly luck, so paying yourself for it teaches you nothing about what to repeat. That argument is strongest when the behaviour is new and unreliable. It gets weaker every week that the behaviour holds.

Once you take the chance roughly as often as you mean to, the interesting question stops being “did I act?” and becomes “did acting that often actually get me there?” A rate bar can no longer tell you that. A goal bar can.

Zimmerman & Kitsantas (1997), Developmental phases in self-regulation: Shifting from process goals to outcome goals, Journal of Educational Psychology 89(1), 29–36. Ninety high-school girls learned to throw darts. Those given goals about the method beat those given goals about the score, on skill, on self-efficacy and on interest in the game. The group that did best of all began on process goals and moved to outcome goals later in practice.

That is the whole ordering, from one study, in one sentence. Not process goals for ever. Process goals, then outcome goals. The group that only ever used outcome goals came last, and the group that only ever used process goals was beaten by the group that moved.

Staying on the process bar for ever is the second-best strategy, not the best one.

But moving too early has a measured cost

If the goal bar is the better instrument later, the obvious question is why not start there. There is an answer, and it is not a matter of taste.

Winters & Latham (1996), The effect of learning versus outcome goals on a simple versus a complex task, Group & Organization Management 21(2), 236–250. On a simple scheduling task, an outcome goal beat urging people to do their best. On a complex one it did not. A learning goal — a target for the number of strategies to find — beat both the outcome goal and do-your-best, and produced more effective strategies.

Seijts & Latham (2001), The effect of distal learning, outcome, and proximal goals on a moderately complex task, Journal of Organizational Behavior 22(3), 291–307. On a task where people had to acquire knowledge before they could do it well, a specific difficult outcome goal produced worse performance than simply telling them to do their best. A specific difficult learning goal beat both.

That is the sharp version of the warning. On something you have not yet worked out how to do, a demanding outcome target did not merely fail to help. It did worse than no target at all, because people scrambled at strategies instead of learning one.

The outcome goal is not a harder version of the process goal. On an unmastered task it is a different and worse instrument.

I should be straight about how much weight this bears. It is a small cluster of laboratory studies, most of them from one research group, and I could not find a meta-analysis of this exact comparison. Take it as a good reason to be patient with the goal bar, not as a law.

The reward research is on this bar’s side too

There is a well-known finding that rewards undermine motivation, and it is usually quoted as a reason not to do any of this.

Deci, Koestner & Ryan (1999), A meta-analytic review of experiments examining the effects of extrinsic rewards on intrinsic motivation, Psychological Bulletin 125(6), 627–668. Across 128 experiments, rewards paid for simply engaging with a task, or for finishing it, showed the largest undermining of intrinsic motivation (d = −0.40 and −0.36). Rewards paid for meeting a standard of performance showed less (d = −0.28), and did not significantly reduce self-reported interest. The differences between contingency types were not statistically reliable on the behavioural measure.

Read that carefully and the reward type that comes off best is the one this guide is about: a reward for clearing a standard you were aiming at. It was the least damaging of the three, and it was the only one that left self-reported interest alone.

I would not lean hard on that, because the authors are clear that the differences between the three types were not statistically reliable on the behavioural measure. But it is worth knowing that the paper most often cited to say “never reward yourself” is least damning about exactly this arrangement.

What the goals bar actually measures

This is where people get caught, because the bar behaves like two different features depending on how you scope it.

Across several habits, it is a percentage. Scope the reward to all your habits and the bar asks what share of them met their goal this period. Four habits, three of them there, 75%. Set the bar at 75 and you earn it. This is the version that behaves the way you expect a percentage to behave, and the version where a bad week on one habit can be carried by a good week on another.

On a single habit, it is a yes or no. Scope it to one habit and the share can only ever be 0% or 100%. There is no 60% of one habit. Every threshold from 1 to 100 would behave identically, so the app stops pretending and states the bargain instead: you met the goal this period, or you did not.

That is worth understanding before you set one up, because a single-habit goal reward is a much harder bargain than it looks. There is no partial credit anywhere in it.

Where to put the bar

For the multi-habit version, the same rule applies as for the rate bar: look at what actually happened before you decide.

The reward editor shows the last eight closed periods against whatever bar you are dialling, and tells you how many would have earned it. Something you would have taken four or five times out of eight is a bar worth having. Eight out of eight is decoration. Nought out of eight is a rule you have written to disappoint yourself with.

Two habits is the setting to be careful with. The only available answers are 0%, 50% and 100%, so a bar of 75 is secretly a bar of 100, and a bar of 60 is secretly a bar of 100 as well. Check what the numbers can actually be before you pick one.

And once you are on this bar, set it high enough to mean something. That is the one place where the oldest finding in this whole literature applies directly.

Locke & Latham (2002), Building a practically useful theory of goal setting and task motivation: A 35-year odyssey, American Psychologist 57(9), 705–717. Across thirty-five years of studies, specific and difficult goals produced higher performance than urging people to do their best (effect sizes .42 to .80), and performance rose with goal difficulty in a straight line (.52 to .82). It levelled off only at the limits of ability, or when commitment to a very hard goal lapsed. Commitment matters most exactly where the goal is hardest.

So a soft goal bar wastes the instrument. But note the second half of that finding, because it is the reason this guide keeps saying to check the last eight periods: the line stops going up the moment you stop believing in the goal. A bar you have privately given up on has no effect at all, however impressive it looks.

The case for setting it to zero

The bar goes down to zero, and zero is allowed on purpose. It means any finished period earns the reward.

That sounds like a loophole and it is occasionally the right answer. If you are restarting something after a bad stretch, the thing you are trying to rebuild is not performance, it is turning up at all. A gentle deal you actually collect on is worth more than a demanding one you quietly stop believing in.

It is also the honest setting for a first period with a brand-new habit, where you have no idea yet what a normal week looks like. Take the easy reward once, find out what your numbers really are, and set a real bar for the period after.

What a goal reward is not

A goal reward is still not a reward for a result out in the world, and the distinction matters.

“Three sessions a week” is a goal about what you did. “Lose ten pounds” is a result that depends on a great deal besides you. The app can judge the first, and deliberately cannot judge the second, because a bar you can fail while doing everything right is the bar this whole product exists to avoid.

So use the goals bar for goals you set about your own behaviour. It is an outcome bar in the sense that it measures whether the behaviour added up to the target. It is not an outcome bar in the sense of paying you for how the world responded.

And one caution that applies specifically here: if the habit is one where the chances are not yours to control, a fixed goal was a guess in the first place, and rewarding it inherits the guess. Those habits should stay on the rate bar for longer than you think.

How to know it is time to move

Three signs, in the order they usually show up.

• Your rate bar has stopped moving. You clear it most periods without thinking about it. It has done its job and is now telling you something you already knew.
• You are hitting the goal anyway. When the rate is reliable, goal completion tends to follow, and the goal figure becomes the one with news in it.
• The question in your head has changed. You have stopped asking whether you are acting and started asking whether acting that often is enough. That is the moment the goal bar has something to say and the rate bar does not.

None of this is one-way. If the behaviour falls apart, move back. The rate bar is the one that rebuilds it, because it is the one that pays you for the thing you can do today.

What to do with this

If you are new, ignore this article for a month and set a rate reward instead.

If you have a habit that has been steady for a while, set one goal reward on it, scoped to that habit, for a period you can feel — a week or a month, not a year. Accept that it is a yes-or-no bargain. Then leave it alone and let the period close.

The useful part is not the prize. It is the first period where you clear the rate comfortably and still miss the goal, because that is the period that tells you the behaviour is fine and the target was wrong.