Home/Blog/Product Discovery/How to Close the Loop After Shipping, an…
Product Discovery

How to Close the Loop After Shipping, and Own It

How to close the loop after shipping: check each release against the outcome it predicted, make one person responsible and feed the result back into your goals.

Cover of the FIKR Space guide: How to Close the Loop After Shipping, and Own It

To close the loop after shipping, check every release against the outcome it was expected to produce, make that check somebody's named job, and feed the result back into your goals and your next round of discovery. Almost no team does this consistently, which is why most cannot say which of their releases worked.

Discovery is usually described as the work before building. The same track also carries the work after, and this post is about that second half.

How to close the loop after shipping: what the loop is

When something ships, discovery's work is not finished. The prediction attached to the work should be checked. Did the outcome move?

The guide describes the chain that this check completes. The goal said what should be different. Discovery said this piece of work would achieve it. Delivery built it. Now the measurement says whether the reasoning held.

Link in the chainWhat it contributes
The goalWhat should be different
DiscoveryThe reasoning that this route would get there
DeliveryThe thing that was built
The check after shippingEvidence of whether the reasoning held

Skip the last link and the other three become a story the team tells itself. The loop is what turns a sequence of releases into learning.

Why this step gets skipped

The guide is blunt that almost no team does this consistently. The reasons are easy to see.

  • Shipping feels like an ending, and the next item is already waiting.
  • Delivery has visible deadlines. A check after release has none.
  • Nobody was asked to do it, so everybody assumes someone else will.
  • The result might be unwelcome, and there is no pressure to go looking for it.

The cost of skipping is hard to see precisely because it is the absence of something. Without the loop, the team never finds out whether its judgement is any good, and repeats the same estimation errors indefinitely. The guide puts it vividly: it is the difference between ten years of experience and one year repeated ten times.

Free guide, 34 pages: The Dual-Track Playbook. Send me the guide.

Make it somebody's job

The guide's instruction is explicit: close the loop, and make it somebody's job. Work that belongs to everyone is, in practice, work that belongs to no one.

Concretely, that means someone owns the check, and it happens. The guide's checklist phrases it as two separate boxes, which is deliberate. Having an owner is not enough if the check never takes place, and the check will not take place reliably without an owner.

  1. Name the person who owns the check for each piece of shipped work.
  2. Make sure the outcome to be checked was written down before shipping.
  3. Look at the result when it is due, and record it.
  4. Pass it on to whoever owns the goal, and to the next discovery cycle.

The second step depends on an earlier habit. If the work arrived from discovery with a stated expected outcome, there is something specific to check. That is one reason the hand-off note includes the outcome, as explained in how to hand off discovery to delivery.

Connect it to the quarterly goals

The loop works best when it plugs into something larger. The guide says discovery should be attached to the quarter's objectives, so the track has a destination, and the check after shipping is where that attachment pays off.

Objectives say what should be different. The measurement after shipping says whether it is. Without a goal, a result is a number with no meaning. With one, it is a verdict on a specific piece of reasoning.

The guide also notes that an outcome-based objective is what permits discovery, because it commits to a result rather than to a predetermined feature. That helps here too. If the goal is an outcome, then a release that did not move it is a clear signal, and a cheaper route to the same outcome is a success and not a deviation from the plan.

If the outcome movedIf the outcome did not move
Record that the reasoning heldRecord that it did not, and why you now think so
Note which assumptions proved soundNote which assumptions proved wrong
Feed it into the objectiveFeed it into the objective and the next discovery cycle

Feed the result back into discovery

A check that only produces a number is half a loop. The guide's checklist asks that results feed back into the objective and into the next discovery cycle. A release that missed its outcome often contains the best question for the next round: which belief did we hold that turned out not to be true?

That question belongs at the front of the next cycle, ranked like any other. Whether it is the most dangerous assumption now is something to decide with the same method as before, described in how to test your riskiest assumption.

Write the result into the shared record as well, so that it survives the people who were present. See how to keep a discovery record for a format that costs almost nothing to maintain.

What the check can look like

The guide does not prescribe a format, and the check does not need one. It needs to answer one question for each piece of shipped work: did the outcome we predicted actually move? A short entry is enough.

Part of the entryWhat it says
The workWhat shipped
The predictionThe outcome it was expected to produce, as stated at the hand-off
The resultWhat actually happened to that outcome
The readingWhether the reasoning held, and what that means for the next cycle

Because the prediction was written before shipping, the result cannot be bent to fit afterwards. That is the same protection the written bar provides during discovery, and it is just as valuable here.

Where the loop connects to the rest

The closing check touches every earlier stage of the practice, which is why it matters so much to the whole.

  • The goals receive evidence about whether the chosen route worked.
  • Discovery receives a new set of questions, drawn from what turned out to be wrong.
  • The bar can be adjusted if assumptions that cleared it later proved false.
  • The record gains an entry that shows how the reasoning held up against reality.

It is also the clearest answer to the question of whether discovery is worth the capacity it takes. A team that can show which of its releases worked, and how its predictions have improved, has a far easier time defending the share described in how to split capacity between discovery and delivery.

FIKR OS, strategy meets execution: OKRs, roadmaps and kanbans in one system. See FIKR OS.

Common reasons the loop breaks

What goes wrongWhat to do
No outcome was stated before shippingRequire it at the hand-off, so there is something to check
Nobody owns the checkName one person for each piece of shipped work
Goals, work and measurement are kept apartKeep them within reach of each other
The result is unwelcome and quietly ignoredRecord it, and treat the miss as the best question for the next cycle
The loop grows heavyCut it back to a question and a short answer

Review the reasoning, not only the output

The retrospective is the natural home for the loop. The guide recommends asking about the quality of decisions rather than the volume of work.

  • Which assumptions turned out wrong?
  • Where were we overconfident?
  • Which tests actually changed a decision, and which were theatre?
  • How long did our discovery items really take, against what we expected?

This produces the calibration that makes the whole practice compound, and it takes half an hour a cycle. Teams that only hold retrospectives on delivery get better at delivering things nobody needed.

What a closed loop gives you

Without the loopWith the loop
No one knows which releases workedThe team knows which of last quarter's releases actually worked
Estimation errors repeatPredictions improve as the team compares them with reality
Judgement stays where it beganJudgement gets better with each cycle
Goals and work drift apartGoals, work and measurement stay within reach of each other

The last row matters because the loop only works when the goals, the work and the measurement are close together. If they live in three places maintained by three different people, the check is the first thing to be dropped.

Keep the loop light

As with the hand-off, the loop should not grow a bureaucracy of its own. A check after shipping is a question and a short answer, not a new process. If maintaining the loop starts to take more effort than the learning is worth, the guide's warning about stage gates applies: the process has become the point, and it will be quietly abandoned.

The loop also fits neatly into the monthly rhythm the guide recommends. At that check, alongside what discovery decided and what delivery shipped, ask which shipped work has now been measured.

Free guide, 34 pages: The Dual-Track Playbook. Send me the guide.

A checklist for closing the loop

  • Shipped work is checked against the outcome it predicted.
  • Someone owns that check, and it happens.
  • Results feed back into the objective and into the next discovery cycle.
  • We know which of last quarter's releases actually worked.
  • The retrospective asks which assumptions were wrong, not only what shipped.
  • We noted where we were overconfident.
  • Discovery estimates are compared to reality, so they improve.
FIKR OS, strategy meets execution: OKRs, roadmaps and kanbans in one system. See FIKR OS.

Frequently asked questions

Who should own the check after shipping?

The guide says to make it somebody's job, and does not prescribe a role. The key is that a named person owns it and that it actually happens, since an owner without a check, or a check without an owner, both fail.

What do we check shipped work against?

The outcome it was expected to produce. That is why the work should arrive from discovery with its expected outcome stated, so there is something concrete to compare with.

What if the outcome did not move?

Record it honestly, ask which assumptions were wrong, and feed the answer into the objective and the next discovery cycle. A miss that is understood is more valuable than a hit that is not.

Is this part of discovery or part of delivery?

It sits with the discovery track, which carries work both before and after building. But it connects the whole chain from goal to measurement, which is why the people doing both tracks should be the same.

The full method, with the checklists, is in The Dual-Track Playbook, a free 34-page playbook.

LG

Luis Goncalves

// Founder & CEO at FIKR Space

Three-time founder. Built and exited Evolution4All before this. Now building FIKR Space — the operating infrastructure underneath every innovation ecosystem (startups, accelerators, governments, investors). Lisbon-based, works global.

Claim your founding member spot.

Every founding member is hand-reviewed. Tell us who you are and how you want to co-create — we get back inside 48 hours.

Role
Job Title
Country
Cookies

We use cookies.

Essential ones keep the site working. The rest only run if you say so. No dark patterns, no pre-ticked boxes.Privacy policy

Co-create with us. It's about building it together.Co-Create With Us