We should not only ask, “Did we finish?”
We should also ask:
- What worked?
- What did not work?
- What slowed us down?
- What should we change next time?
- What needs a new owner or decision?
Learning should be part of the work.
Why this matters
Most teams measure projects with a single question: did we deliver? The answer is binary and reports cleanly to leadership. It also teaches the team almost nothing about whether the work was worth delivering or how the next attempt should be different.
A review that asks five questions instead of one produces something the binary version cannot: a small, specific change the team can test next quarter. Over four cycles of that practice, the team gets faster at the same kind of work without working harder. That is what “learning organization” actually looks like in a sentence.
What it looks like in practice
At the end of a quarter, the team that drafted a Quarterly Goal Sheet at the start sits down with the Look-Back Review and walks the questions. Not abstractly — against the actual goal they wrote 12 weeks ago:
- What worked? The Tuesday status pattern caught two issues that would have hit week five.
- What did not work? The “not doing” line held for six weeks; then a leadership request added scope without anyone noticing the original tradeoff.
- What slowed us down? Two handoffs to a shared service that averaged five days when they were supposed to average two.
- What should we change next time? One specific process change to test, with a named owner and a review date 30 days out.
- What needs a new owner or decision? The scope-creep conversation is a leadership decision; surface it before next quarter.
That conversation produces a specific improvement the next quarter can test. The conversation that just asks “did we finish?” produces a sense of completion and no learning.
What it looks like when violated
The most common failure: the review never happens. The quarter ends, the team moves to the next quarter, and the same problems recur invisibly. Look-backs lose to fresh work every time unless they are scheduled before the quarter ends.
A second common failure: the review does happen, but it slides into either a victory lap (“great job everyone”) or a blame search (“why did X happen?”). Neither produces a change to test. The discipline is the third option: name the specific system or behavior to try differently, name an owner, name a review date.
A third failure: a review produces a long list of “things to improve.” The team is energized for a week. None of the items get owners or dates. Two months later, nothing has changed. One small change with an owner beats ten ideas with none.
How to apply it this week
- For every project that closed in the last 60 days, ask: did we actually do a look-back, or did we just move on? If not, schedule a 30-minute one this month.
- When you run a look-back, end it with one named process change — one — with an owner and a review date. If the room cannot agree on one, the look-back is incomplete.
- Put your team’s next look-back on the calendar before the current sprint or quarter ends. Calendar discipline is the only thing that protects the practice from urgent work.