LLenny's Podcast
← All frameworks
InnovationPaige Costello (Asana, Intercom, Intuit)

The Double Diamond Product Process

Alternate broad exploration and narrow selection across customer, problem, and solution

Difficulty
Advanced
Time to result
~weeks to results
Steps
4
Confidence
85%

A discovery process that forces teams out of opinion-driven decisions by repeatedly diverging (going broad) then converging (going narrow) across three questions: which customer, which problem, which solution. Product review artifacts are mapped to specific inflection points on the diamonds so decision quality stays high.

Origin

The Double Diamond is a design-process model from the British Design Council; Costello relates it to Lean Startup concepts and describes how Asana operationalizes it by mapping its review artifacts onto the diamond's inflection points.

Core principles

  • 01Alternating broad and narrow breaks opinion-driven framing and forces curiosity — both quantitative and qualitative.
  • 02It always starts with enough customer insight to know who you're solving for and what 'problem solved' means.
  • 03If you pick success metrics as 'usage of a feature,' you're teaching to the test rather than driving the real outcome.

How to run it

  1. 1

    Diverge then converge on the customer

    Go broad asking which customer you could solve for, then narrow to pick one specific target customer.

    Pro tip Match the depth of the kickoff research to the size of the problem and the ambiguity you're solving.

  2. 2

    Diverge then converge on the problem

    Go broad on all the problems that customer has, then narrow to the specific problem you'll tackle.

    Pro tip Define what it actually means for that problem to be solved, so success isn't 'they used the feature.'

    Watch out Choosing feature-usage as your success metric teaches to the test and doesn't drive the outcome.

  3. 3

    Diverge then converge on the solution

    Go broad on possible solutions, then narrow to the solution you'll start with.

  4. 4

    Map artifacts to inflection points

    Attach each review — kickoff, customer/direction selection, design concept review, full experience review/design crit, launch review — to its point on the diamonds so thinking is validated at each stage.

    Pro tip Run design crits in person and keep most spec reviews async, calling a meeting only when question volume warrants it.

In the wild

When Asana deliberately skipped the process for LLMs

For the recent leap in LLM capabilities, Asana staffed a team to prototype quickly and go straight to building instead of running the Double Diamond, applying hypotheses outside the usual norms.

They rapidly discovered some things were 'so much better than we imagined' and would never have been prioritized under normal estimates, while other promising ideas failed fast — validating that the rigorous process is right for typical work but not for high-uncertainty exploration.

Common mistakes

Optimizing feature-usage metrics as the goal

Picking 'used the feature' as success means you teach to the test and miss whether the customer's actual problem got solved.

Starting from solutions before customer insight

Without enough understanding of who you're solving for and what 'solved' means, the broad/narrow passes just formalize opinion.

Is it for you?

Best for

Product teams that want rigorous, customer-anchored discovery on typical roadmap problems

Not ideal for

High-uncertainty exploration (e.g. brand-new technology) where fast prototyping beats staged divergence — as Asana chose for LLMs

From the transcript

Double Diamond as it relates to going Broad and then going narrow and so you go broad when you ask like what customer should I…

13:00

if you pick your success metrics as using a feature that's it you're teaching to the test it's not actually driving the outcome

14:30

From the episode

How to ask the right questions, project confidence, and win over skeptics

Paige Costello (Asana, Intercom, Intuit)