You Probably Suck at Estimating (And What You Can Do About It)

Teresa Torres argues that product people routinely make poor estimates because they produce single-point predictions for uncertain outcomes like build time or conversion rates. Drawing on research summarized in the Heath brothers' book Decisive, she recommends estimating a range instead, identifying the low end, high end, and most likely middle value. To sharpen those bounds she proposes two exercises: the pre-mortem, which imagines a failed launch to surface the low end, and the pre-parade, which imagines a wildly successful one to surface the high end. The technique matters because it forces teams to draw on past experience and makes uncertainty visible before decisions are made.

01Key takeaways

• Replace single-point estimates with a range that includes a low end, a high end, and a most likely value.
• Ask what the shortest and longest past efforts of similar work took to ground the range in experience.
• Run a pre-mortem by imagining the feature failed and listing why, to sharpen the low end.
• Run a pre-parade by imagining the feature succeeded spectacularly and listing why, to sharpen the high end.
• Make estimation a routine habit, since product decisions rely on estimates constantly.

02Key sections

Why single-point estimates fail

Single estimates such as 'three days' or '3% conversion' hide the uncertainty behind them. The author notes that everyone is generally poor at predicting the future.

Estimating a range

Asking for the shortest and longest plausible outcomes, plus a best guess, pulls on different kinds of knowledge and yields more accurate estimates according to the cited research.

The pre-mortem

Imagining a feature has already failed and asking why helps teams tap past experience to predict the low end of the range.

The pre-parade

Imagining a feature was a major success and asking what went right helps teams estimate the high end of the range.

Putting it into practice

The author encourages readers to apply the range-based exercise whenever they estimate something, which will happen often.

03From the post

“This post is part of a series about making better product decisions. You probably estimate every day. How long will it take to build a given feature? What impact will it have on your customer base? How will it impact business outcomes? It's too bad you probably aren't”

04Frameworks mentioned

Summary and takeaways written by PM Atlas; quotes are short excerpts. © the original author.