2 builders who never spoke to each other

Rank the roadmap on one list, not a scorecard

Put candidate work in a single ranked order instead of scoring and weighting it, because weights tie and a ranked list gives you a decision you can defend to whoever lost. Build the criteria from the reasons customers actually give for leaving, make someone state the size of change a project would cause and drop anything too small to see, and write down only the threats already happening.

The method

1. 01

   Before a project earns a slot, make someone write the size of change it will cause, and throw out anything whose effect would be too small to see in the numbers.

2. 02

   Put your criteria in a single ranked order rather than scoring and weighting them, because a ranked list gives you a decision you can defend to the person whose idea lost.

3. 03

   Build the ranked criteria out of the reasons customers actually give you for leaving, not out of generic goals like growing revenue.

4. 04

   Write down only threats that are already happening or are at least seventy percent likely, in the present tense and backed by data, and drop the hypothetical ones every company could list.

5. 05

   Allow one or two goals that are not numbers, written as a crisp sentence anyone can repeat, and treat arguments about whether you are hitting them as useful rather than a defect.

What it returned

• 10%

  His thresholds for counting as material include a 10% lift in revenue, a 20% fall in cancellations traced to one cause, 40% fewer support tickets on a topic, a 1% gain in gross margin, and a 50% rise in the share of people who get through a given interface flow.

• 20

  Two options with different strengths tie at 20 points each, and stay tied even when he rigs the weights, because the numbers are mostly noise and the units are not comparable.

• For a product sold to mid-sized restaurant chains, first-year cancellations were blamed on staff training being too hard, so the software went unused, while long-tenured customers left over price, which set the priority order.

• Real examples he gives: the market is shrinking, a rival's share is accelerating, or cancellation rates showing paying customers do not value the product. Lazy ones: a large company could copy us, or a funded startup could appear.

Sources

• ⭐️ You need a game plan – Jakob GreenfeldJakob Greenfeld · jakobgreenfeld.com · ⭐️ You need a game plan · 2022-03-08
• What makes a strategy greatJason Cohen · longform.asmartbear.com · Candid · 2023-08-13
• Binstack: Making a maximal multi-dimensional decisionJason Cohen · longform.asmartbear.com · Binary materiality · 2022-07-02
• Quarterly strategic planning using the fairytale structureJason Cohen · longform.asmartbear.com · Metrics & Indicators · 2023-11-26