Accountability Partner Contract: A One-Page Template

Ten clauses on one page make an accountability partner contract for two people with one goal each. They fix what counts as done, when you work and report, what a miss costs, and when the deal expires. Below you can copy them, see them filled in for studying and for work, and read the research behind each.

Copy this one-page template

Fill in the brackets together, in one sitting. The questions worth talking through first are in 40 questions for every check-in; this page is where the answers end up.

1. Partners: [Name] and [Name]. We see each other [where and how often].
2. Goals: [Name] finishes [what, as a number] by [date]; [Name], [what, as a number] by [date].
3. Work sessions: Each of us works on [days] from [time] to [time], at [place or online room]. A session that can't happen moves to [backup time], not to "later".
4. Done means: [the number or the file], shown by [a photo, a link or a dated screenshot].
5. Reports: Every [day and time], by [text, call or video]. A report gives one number against clause 2 and the time of the next session.
6. A missing report: counts as a missed session, unless it was a planned skip.
7. Planned skips: Each of us may skip [how many] sessions per [period], announced by [deadline].
8. Stakes (optional): A missed goal costs [amount or task], paid to [whom] within [days].
9. What we share: [numbers only, screenshots, grades]. Nothing else gets asked for.
10. Term: This agreement ends on [date, four weeks from today]. Going on takes a new page, signed again.

Signed: [Name], [date] · [Name], [date]

Here is the same page filled in for two common cases:

What to count for twelve kinds of goals, from a thesis to a job search, is in these worked pairings.

Why each clause of an accountability partner contract is there

Writing it yourselves (clauses 1, 2 and 10)

Commitment grows when you make the choice by doing something. Delia Cioffi and Randy Garner asked undergraduates to volunteer for a university committee or an AIDS awareness education project (Personality and Social Psychology Bulletin, 1996). Some volunteered by marking two statements that said yes; others by leaving blank two statements that said no. The decision was the same, but the active volunteers were more committed to it. The difference lasted six weeks and showed up in what they did.

So each partner writes their own goal in their own words, and renewing in clause 10 takes a new signature rather than happening by default.

Don't expect the signature itself to do more than that. A well-known 2012 study claimed that signing at the top of a form made people more honest. PNAS retracted it in 2021, after other researchers presented evidence questioning its data. A year earlier, a team that included the original authors had already failed to replicate it. Across five conceptual replications with 4,559 people and a preregistered direct one with 1,235, signing first had no effect on honest reporting (PNAS, 2020).

Times and places (clause 3)

A plan with a time works better than encouragement. Before the 2008 US presidential election, David Nickerson and Todd Rogers ran a field experiment with 287,228 people (Psychological Science, 2010). Calls that helped people form a voting plan raised turnout by 4.1 percentage points among those reached. A standard encouragement call and a call asking people to predict whether they'd vote had no significant effect.

The plan mattered most in households with only one eligible voter, where it raised turnout by 9.1 points. The authors' data suggest that people who live with other voters may already form such plans on their own. Clause 3 is the written plan, and the backup time covers the session most likely to slip.

A partner who sees you (clauses 1 and 5)

Someone close to your daily life makes a better witness. Emily Breza and Arun Chandrasekhar randomly assigned "monitors" to savers in 30 Indian villages: fellow villagers who were kept informed of each saver's progress (Econometrica, 2019). An average monitor raised savings by 35%. A monitor one step closer in the village's social network raised them by a further 16%, and 63% of monitors told other people how the saver was doing. The effect was still there a year after the program ended.

For a pair, that means a classmate or coworker you see most days, rather than a stranger you'll never run into, if you can choose. Clause 1 asks the question outright.

Planned skips (clause 7)

A plan with permission to stray can outlast a rigid one. Rita Coelho do Vale, Rik Pieters and Marcel Zeelenberg tested this in two controlled dieting experiments, one simulated and one real (Journal of Consumer Psychology, 2016). People whose plan included planned deviations kept their motivation better and felt better than those following a straight, rigid plan. A third study with people working toward other goals pointed the same way. The key word is planned: a skip announced in advance counts as part of the plan, while an unannounced one counts as a miss under clause 6.

Stakes that bite, or none (clause 8)

A stake works when losing it would hurt. Anett John offered low-income savers a commitment account in which each person chose a savings plan and a penalty for falling short (Management Science, 2020). Savings rose on average, but 55% of clients defaulted and paid the penalty. Her leading explanation is that the penalties people chose were too low to change what they did.

So clause 8 is optional, and an accountability partner contract works without it. If you add one, pick a stake you'd go out of your way to avoid paying. Apps that hold the money, and the trials behind them, are compared in our accountability app guide.

What written contracts alone have shown

On its own, a contract has thin evidence behind it; the clauses above carry the parts that have been tested. For contracts themselves, the clearest evidence is a Cochrane review of agreements between patients and health workers. Xavier Bosch-Capblanch and colleagues found 30 randomized trials with 4,691 participants, from addiction to blood pressure (Cochrane Database of Systematic Reviews, 2007). Each trial measured several outcomes:

A trial can appear in more than one row. The median group had just 21 people, and every trial fell short on at least three of eight quality standards. Where adherence was measured over longer periods, no effect showed up. The reviewers concluded that the evidence was too thin to recommend contracts routinely.

That is the case for clause 10. A four-week term is short, and renewing it means both of you decide again, on paper.

Putting a room in clause 3

Clause 3 needs a place, and it can be online. We make StudyDuo, so here's how its rooms fit the template:

• Study Hour starts at one fixed moment each day, shown in each partner's own time zone, so two people in different cities can share it. It's an hour with no breaks, and you can come in from 10 minutes before the start. Afterward, the app asks whether you got your task done, which gives clause 5 its number.
• Focus Room never closes. Each of you types the task, picks 30, 60, 90 or 120 minutes, and works on camera, mics off.
• A 1:1 is a backup for clause 3. If your partner can't make a session, post one on the board for now or any day in the coming week, and whoever takes it works with you on video. It needs a verified photo and is for ages 16 and up.

Follow each other in the app, and you each get a notification when the other starts studying: a report before the report. Rooms aren't recorded, so clause 4 still needs its own proof. More on pairing up is on the accountability partner page. Print the page above, or paste it into a shared document, and you have an accountability partner contract by the end of today.