A technique for starting new habits and maintaining motivation: Attack Doses

Methods for beginning new habits

Friday, January 02, 2015

I’ve always had a nagging feeling (since writing Building Sustainable Habits: Why We Make Excuses and Resist Habit Change) that sometimes building habits using small steps isn’t always the right way to go. There are people who are able to start a new habit, ramp up fast, build a self-reinforcing loop of motivation, and continue to execute over and over—without the need for small steps. So, which one is the right way to build a new habit? Using small steps, or using the fast-track approach of getting motivated and going hard on the new habit?

One argument is that most people won’t be able do the fast-track approach. We might hear about the cases where this was successful, but there are many where the fast-track approach didn’t work, and we don’t hear about those cases.

Another argument might be that we can combine both of them to match the different levels of motivation that we have, such as the high motivation we have in the beginning of a habit and the sometimes declining motivation we have in the middle of one. In this article, I’ll introduce a hybrid approach, which takes advantage of both high motivation and sustainable habit development, and I’ll outline some example habit development plans that incorporate the attack dose technique. While this is wholly empirical and strictly a theory, it may be a useful concept to incorporate into your habit development, especially as you implement new goals for 2015.

Motivation waves

BJ Fogg, the director of the Stanford Persuasive Tech Lab, and Tiny Habits, a habit development program, introduced the idea of motivation waves, stating that we have different levels of motivation at different points during the development of a habit.

Fogg argues that during these high motivation points, we have a temporarily higher ability to engage in habit-building tasks. We might have higher willpower during these points, which would allow us to follow through on doing the habit, which we can apply to dual-process habit development.

A recap of dual-process habit development theory: the goal of habit development is to close the gap between “what you want to do” and “what you actually do”. Let’s take exercise: you know that you should exercise to be healthy, maintain a good appearance, increase confidence, and other reasons of that sort. This is the conscious side, which has reasoned out the benefits. But—a lot of people don’t actually do it, because they feel like it’s annoying, painful, don’t have enough energy, and other reasons, which is their instinctual side thinking about why they don’t want to exercise. These two sides are often in opposition. The goal of sustainable habit development is to build instinctual evidence—such as seeing results, weight loss, better health, etc.—so that “how much you know you should exercise” matches “how much you want to exercise”.

When we start a new habit, we’re almost always in a high motivation state: either it’s a new year, or something just gets us Mad As Hell and we have enough energy to change it.

For me, that’s certainly getting up earlier, which has been a challenge for, I dunno, 22 years or so. In my daily writing exercise for today, I noted my frustration for not being able to develop the waking-up-early habit for so long. For me, the key problem with waking up late is going to sleep late. Right now, I’m in a high motivation state, and I have the rare opportunity to use the energy I have to create the new habit of getting up early.

The attack dose

I could engage in the sustainable habit model and build the habit using small steps. With 2am as my usual bedtime, I could probably make “going to bed before 1am” a good first step, and wake up at 9am, feeling moderately accomplished, building some instinctual evidence that “waking up early is a good idea”.

Alternatively, I could harness the high motivation wave that I’m on in the beginning of a habit, and use an attack dose to take advantage of that high motivation. The key question is: “what’s the most intense thing I’m willing to do right now to act on this habit?“—to which my answer is, go to sleep 4 hours before my average bedtime. If you’re a midnight sleeper, that means going to sleep at 8pm. (I’m a 2am sleeper, so an attack dose would be 10pm for me). I’ll also have a specific metric for success as usual, like how good I feel after waking up earlier.

I believe the attack dose will allow you to fast-track building instinctual evidence. My attack dose for waking up earlier, which is going to sleep at 10pm and waking up at 6am, would get me up 4 more hours earlier and—presumably—I would feel much better about that and it would allow me to taste a bit of how good it feels to wake up so early, which can be highly motivating. A normal ‘small steps’ approach, waking up at 9am, would make me feel good, but not as euophriously good as waking up at 6am. Since we have the initial motivation to do it, we should use that to build more motivation for us to continue building the habit.

Planned deceleration

But there’s a caveat: keeping this up will be extremely difficult. Both internal factors (such as fatigue and willpower depletion, especially at night) plus external factors (such as events in the evening and urgent things that prevent you from keeping this up) will throw this into chaos given enough time and the randomness of life. Make no mistake: during the attack dose period, the prospective habit is highly volatile and not at all stable nor sustainable. As a result, we need to predict that these things will happen, and plan to relax the habit intensity over time to more sustainable levels. Consider the following:

For example, I might start with an attack dose of going to sleep at 10pm for 3 nights, and then gradually decelerate to going to sleep at 11pm, and then 12am, and then 1am. I’ll then engage in sustainable habit development, making small steps to get back to the 10pm goal—perhaps taking a few weeks to get back to 10pm, but doing so sustainably, facing some challenges but being able to get through them, and building a strong foundation of instinctual evidence.

The reason we deliberately decelerate the habit is because random events will happen that will disrupt the high bar you’ve set for yourself with the attack dose. Habits, especially the early days, are highly vulnerable to disruption, and we need to build resilience to disruption, which is one of the goals of using small steps for sustainable habit development. Habit resilience is what can allow us to continue developing a habit even when we’re facing challenges, like missing a day, but if we don’t have the resilience, failing to make good on a habit one day may unravel the entire habit.

In mindfulness meditation and in Zen, perhaps the most important lesson is to reserve judgment and be kind to oneself when things go wrong. If, in the middle of meditation, you start thinking about bills or something that happened yesterday or things you’re about to do, and then catch yourself in this state of ‘monkey mind,’ the key is to not berate yourself for failing to keep focused. Rather, the better course of action is to recognize it, accept that it happened, and refocus without judgment.

Resilience is the acceptance of challenges and pressing on regardless. You could be strong-willed and stick to the attack dose and see challenges as just “one-time things” that you can bounce back from—but this depends strongly on willpower, which is unreliable. Instead, a more accessible goal is to lower the bar, so when challenges inevitably crop up, it’s more likely that we are resilient enough to continue the habit, instead of losing it altogether. In other words, if we expect that there will be challenges, and plan that into habit development, we may be more capable of continuing the habit despite challenges.

This is especially relevant for new year’s resolutions. Many people go hard on a new resolution, such as by going to the gym every morning, and expect to keep this up for the entirety of the habit. But then they miss a day, then another, and the habit is often lost after a few discouraging failures. Instead of thinking that we’ll keep up an ambitious habit, we have to expect that our motivation will wane, and things will come up that will disrupt the habit—we have to plan for them, lower the bar for the habit, expect disruptions instead of being able to continue going to the gym every day at 6am. Building resilience.

When we do so, the habit is easier to do, and we may be more easily motivated to do it, especially compared to the attack dose. When we have a task that you are easily motivated to do, it can create a safety net for when your motivation is lower—the task is still easy enough to do regardless.

Consider the following hypothetical diagram:

Here, we see that we match our high motivation with high habit intensity in the beginning (the attack dose). Then, we taper down, but our motivation is most likely still high. The reason is that by creating a gap between the (lower) intensity of the habit and our (higher) motivation level, we may be more resilient to drops in motivation that may happen over the course of time. Assuming that missing one day of a habit puts us at high risk of dropping the habit altogether, it seems to be essential to have this safety net.

The attack dose: pros and cons

My theory is that the attack dose:

  • Builds more instinctual evidence for the habit. After waking up at 6am and getting a lot done, I’ll be able to really see the benefit of waking up early, which will increase the instinctual evidence way more than just doing the small-steps plan, taking advantage of the higher motivation wave, which will theoretically maintain my motivation to continue the habit due to the stronger evidence for doing it.
  • May create self-reinforcing motivation. By getting big successes early on, your initial high motivation may be even more elevated, potentially boosting future habit performance. Seeing lots of results from attack-dose habit development may contribute a lot more to your motivation than seeing smaller gains from normal small-steps habit development. (This may become a motivation ‘multiplier’ of sorts, causing cascading effects.)
  • Keeps the instinctual evidence more accessible. I can probably think of a day a long time ago when I got up early, but that far-away memory doesn’t affect my decision-making much. Rather, immediate, near-term evidence (such as how great it was to wake up early) may be more effective, since you experienced the evidence recently. [We know that activating certain memories using words makes them more accessible, which in turn influences judgment (Forster & Liberman, 2007). It doesn’t seem too much a jump to hypothesize that experiential activation, that is, doing some action, increases accessibility and influences behavior.]
  • Shows the contrast between the goal and your initial state. Imagine going from a few super-productive days waking up at 6am and then scaling down to a schedule close to your previous one. Seeing the contrast between the two may be highly motivating for you to get back to waking up at 6am.
  • Can be combined with other techniques to increase success. We can use techniques that may seem too heavy-handed for small-steps habit development, but that allow us to hit the attack dose goals. For waking up early, we can have negative consequences, such as losing money, if we don’t wake up in time. For exercise, we can enlist a personal trainer to make sure that we get through our attack dose, which could be, say, a full 60 minutes of training. For meditation, a difficult habit to start, we can enroll in a meditation class that will guide us through. While these are not necessary for small-steps development, they can increase the potential that we succeed at achieving the attack dose to build motivation.

There are risks present, of course:

  • Deceleration may be demotivating. Someone going through the planned deceleration process, despite knowing that this is what they planned all along, may be demotivated from seeing their goals dwindle during that period.
  • The habit may be lost during deceleration. If someone went through hardship to achieve the attack dose, perhaps if the costs outweighed the benefits, they may lose the habit during deceleration and fail to engage in sustainable development.

These risks can be potentially sidestepped if we build enough evidence in the beginning, during the attack dose, for the importance of the habit, ideally creating enough momentum to allow the person to either go back to the attack dose or find a balance between sustainable levels for the habit and the attack dose levels. There should be more work on figuring out how to reduce these risks.

Example implementations


Attack dose during high motivation: Go to the gym for 60 minutes, 3 times a week

Planned deceleration: Scale back to going 15 minutes per day, 3 days a week, then escalating back to 1 hour

Building evidence for: How exercise makes you feel great about yourself and your health, makes you more confident, etc.

Attack dose benefits: Allows you to build early evidence and taste a bit of how great it feels to exercise, but also plans for future disruptions by scaling back and developing sustainably

Attack dose risks: Someone may not be physically able to exercise for 60 minutes; scaling back to 15 minutes may make someone lazier over time; someone may not see the results they want to during the attack dose and may be less motivated.

Co-techniques (simultaneous techniques that can increase adoption): External incentives, such as signing up for a class or personal trainer that will make sure that we get through our attack dose.


Attack dose during high motivation: Meditate for 20 minutes for 5 days

Planned deceleration: Scale back to meditating for 3 minutes per day, then gradually increasing to 30 minutes

Building evidence for: How beneficial meditation is, awareness of how busy our minds are, and the need to meditate to become more mindful

Attack dose benefits: Allows you to build early evidence and understand how useful meditation is; 3 minutes is often too little time to see the benefits of meditation apart from “wow, my mind is really chatty”

Attack dose risks: Starting out on a meditation habit is difficult because the first few sessions are frustrating, and the attack dose can potentially exacerbate this frustration, but this is highly dependent on the individual.

Co-techniques: Enrolling in a meditation class, or meditating with someone who meditates often, who can convince you that the difficulties and frustrations are normal.

Waking up early

Attack dose during high motivation: Wake up 4 hours before average wake-up time (e.g. waking up at 6am if you usually wake up at 10am)

Planned deceleration: Scale back to waking up one hour earlier, then work up to desired time

Building evidence for: How great it is to wake up early, feeling less guilty about wasting the day, better work, etc.

Attack dose benefits: Allows you to build early evidence and really see the benefits of waking up early, and it can also get you out of the cycle of waking up late, going to bed late, waking up late, etc.

Attack dose risks: Going to sleep earlier than normal may be difficult, even with high motivation (physical limitations).

Co-techniques: External incentives, such as paying a certain amount of money if you don’t get up at a certain time, can help increase initial adoption. For waking up early, one can combat the physical limitations by using e.g. melatonin to meet the attack dose goals. (These are only for achieving the attack dose, and I believe they are counterproductive in the sustainable development process for building a true internal incentive for the habit.)


When we start a new habit, we are most likely highly motivated to carry it out. It may be smart to take advantage of this high motivation wave and engage in high-intensity habits to build a lot of instinctual evidence for the habit, and then enter into planned deceleration, shifting gears to sustainable habit development. The net result may be higher motivation and much more instinctual evidence that the habit is worthwhile—evidence that can give us strong motivation to continue building the habit, and evidence that is so much more real because, well, you just proved it was.


Thank you to David Ngo, Arjun Balaji, and Conrad Barrett for reviewing and discussing drafts of this article.


Forster, J., & Liberman, N. (2007). Knowledge Activation. In A. W. Kruglanski & E. T. Higgins (Eds.), Social Psychology: Handbook of Basic Principles (2nd., pp. 201–231).

By Mark Bao

I write about behavioral science, personal growth, mental models, and strategy.