Ship faster with blunt metrics
A practical loop for using a small set of honest numbers to make weekly product decisions.
Speed in product isn’t about typing faster. It’s about deciding faster. And you decide faster when your numbers are blunt.
The weekly loop
Here’s the loop we run, and the one Slab is built to support:
- Monday — read the break. Open the dashboard. Find the single biggest drop-off from last week. Don’t read anything else yet.
- Tuesday — form one bet. Write a one-sentence hypothesis about why the break happened. One. Not a backlog.
- Through the week — ship the smallest test. Build the smallest thing that could move that number.
- Next Monday — check the same number. Did it move? Keep or kill. Repeat.
Why “blunt” wins
A blunt metric forces a decision. A nuanced metric invites a meeting. When the number is “37% of signups never create a project”, nobody argues about methodology — they argue about the fix. That’s the conversation you want.
A warning
Blunt doesn’t mean dumb. The danger of a single number is optimizing it into the ground. That’s why the loop pairs every number with a why. The number tells you where to look. The hypothesis keeps you honest.
Ship the small test. Read the same number. That’s the whole job.