Feedforward: feedback without the defensiveness

Feedforward shifts attention from a past mistake to useful behavior for next time.

  • Focus on the next attempt
  • Suggest behavior, not blame
  • Make progress observable

Start a reflection about:

Feedforward in 3 steps

Feedforward plans the next attempt instead of judging the past. Set the outcome, choose an observable behavior, and define a success signal.

  1. Outcome

    Define what should change next time and how you will recognize success.

  2. Suggestions

    Gather observable options without judging the past.

  3. Next attempt

    Choose one option, use an existing strength, and set a review point.

When this method helps

Use Feedforward when you want to support development and plan the next attempt.

Key benefits

  • Focuses the conversation on behavior that can change
  • Makes advice specific and testable
  • Can reduce defensiveness without hiding accountability

Limits and when to pause

Feedforward must not hide harm, accountability, or recurring patterns. Past facts still matter when they define the problem.

Feedforward examples

The three steps stay the same whether you coach one person, bring two colleagues together, or run a round in the team. Each example shows the steps separately and then how you could open the conversation.

Suggestions for one person

Help someone handle a recurring moment better next time.

Illustrative scenario

Outcome
When Sarah raises a risk in sprint planning, Tom often jumps in with solutions. Next time, Sarah should be able to explain her risk fully before the team discusses solutions. It worked if the team confirms the risk list before estimating.
Suggestions
The team lead suggests a sequence to Tom: let Sarah finish, ask one clarifying question, then sum up the risk in one sentence. She does not go back over earlier plannings.
Next attempt
Tom tries it in Tuesday's planning. He builds on a strength: he structures technical topics quickly, so the one-sentence summary suits him. They look at the result together right after the meeting.
How you could open
I have an idea for Tuesday's planning. When Sarah raises a risk, could you let her finish, ask one clarifying question, and then sum it up in one sentence? You structure technical topics quickly, so the summary plays to your strength. If the team confirms the risk list before we estimate, it worked. Shall we look at it right after the meeting?

Two colleagues after friction

Let two peers give each other suggestions instead of reopening who was at fault.

Illustrative scenario

Outcome
In the last release, an API change reached the frontend two days before the code freeze. The team lead has already talked that through with Lena and Marco. Goal for the next release: API changes reach the frontend at least three working days before the freeze. It worked if nobody has to rework code in the last three days.
Suggestions
Each gives the other two suggestions. Marco to Lena: post every API change in the shared channel on the day it is merged, and mark breaking changes in the title. Lena to Marco: comment on the draft spec in Thursday's sync, and name the fields the frontend needs early. Neither debates the other's suggestions.
Next attempt
Each picks one. Lena builds on her thorough API docs and links them in every post. Marco comments on the spec on Thursday. The freeze is on the 20th. On the 21st, both meet with the team lead and check whether any change came later than three working days before it.
How you could open
We've already covered what went wrong with the last release. Today, each of you gives the other two suggestions for the next one. Just listen and say thanks, without discussing them. Then each of you picks the one you want to try.

A feedforward round in the team

Let everyone ask colleagues for suggestions on a goal they chose themselves.

Illustrative scenario

Outcome
In the retrospective, each of the six team members names one thing to do better in the next sprint. Jana picks her demo in the sprint review: it should fit in five minutes and put the question she needs answered in the first minute. It worked if the stakeholders answer her question in the review itself.
Suggestions
Two colleagues give her two suggestions each: open with the question, show one screen instead of five, rehearse once with a timer, and keep technical details for follow-up questions. Jana listens, takes notes, and says thanks.
Next attempt
She picks two of the four: open with the question and rehearse with a timer. She builds on her clear written summaries by writing out the first minute word for word. At the next retrospective, two weeks later, she reports what happened.
How you could open
Let's try something new today. Everyone names one thing they want to do better in the next sprint. Then two colleagues give you two suggestions each for next time. No comments on the past and no debate about the suggestions: just say thanks, and pick afterwards what you want to try.

Each opening looks only at the next attempt. If something caused harm or keeps happening, address it first in a separate conversation, for example with SBI.

Critique or suggestion?

Critique looks back and invites the person to defend what happened. A suggestion names something to try in a specific upcoming moment.

  • Critique of the past“You always cut Sarah off in planning.”

    Suggestion for next time“In Tuesday's planning, could you let Sarah finish and then sum up her risk in one sentence?”

  • Critique of the past“Your API change last time came far too late.”

    Suggestion for next time“For the next release, could you post each API change in the channel on the day you merge it?”

  • Critique of the past“Your demo was much too long.”

    Suggestion for next time“Try opening with the question you need answered, and show one screen instead of five.”

The person can take a suggestion, adapt it, or leave it without losing face. Words like “always” or “last time” pull the conversation back to the past, even when the rest of the sentence looks ahead.

Template: suggest the next attempt

Copy the template and replace each bracket with your own case.

Next time [upcoming situation], I'd like [outcome in one positive sentence]. One idea: [suggestion you can observe]. You're good at [strength], which should help. Let's check on [date] whether [success signal].

Check before you say it

  • Outcome: Does it describe what should happen, not what went wrong?
  • Suggestions: Could someone watch the person do it, and can the person turn it down?
  • Next attempt: Is there a date and a signal you will both be able to see?

Questions Reflacto asks

Marshall Goldsmith's original exercise is a brief exchange: one person names a change they want to make and asks others for suggestions for the future. Two questions follow that exercise: what success looks like, and specific suggestions. Reflacto added the other four: the situation, the current pattern, strengths to build on, and how to track progress.

Facilitating peer feedback

Help colleagues give each other suggestions for the next attempt.

  1. 1What is the specific peer feedback situation you need to facilitate?Added by Reflacto
  2. 2What current behavior or pattern needs to be addressed?Added by Reflacto
  3. 3What would success look like in the future?Feedforward part
  4. 4What specific, actionable suggestions can help them get there?Feedforward part
  5. 5What existing strengths can they build on?Added by Reflacto
  6. 6How will progress be measured and tracked?Added by Reflacto

After the last question, Reflacto shows a plan for the conversation with concrete next steps.

Questions about using feedforward

Prepare the suggestions

What's the difference between feedback and feedforward?

Feedback discusses what already happened. Feedforward plans the next attempt with suggestions for the future.

You could say
“Next time, could you let Sarah finish before you ask questions?”

If harm or accountability must be addressed, feedforward is not enough. Critique and suggestions side by side

How do I actually phrase it?

Name one upcoming situation and one behavior to try there, and offer it as an option.

You could say
“In the next planning, let the estimate finish before asking questions, so risks surface earlier.”

One suggestion the person can picture works better than a list of five. Template

With colleagues and your manager

How do I facilitate feedforward between two colleagues?

Deal with the past first. Then each person gives the other two suggestions for the next attempt, and the one receiving them listens and says thanks.

You could say
“Just listen and say thanks. You'll pick what to try afterwards.”

The no-discussion rule comes from Goldsmith's exercise. It keeps the exchange from turning into a debate about who was right, and the person receiving the suggestions chooses what to try. Example: two colleagues after friction

Can I give feedforward to my manager?

Yes. Phrase it as a request for how you work together from now on.

You could say
“For the next roadmap review, it would help me to see the draft two days before.”

Feedforward works well upward because it does not judge what your manager did. If you need to address a specific past behavior, SBI fits better.

After the conversation

How do we know whether it worked?

Agree on one signal others can see and a date to look at it together.

You could say
“If the team confirms the risk list before estimating on Tuesday, it worked.”

A signal like “fewer interruptions” is hard to check. Name the moment and what you would see in it. Example: suggestions for one person

What if the next attempt doesn't work?

Look at it together, keep what helped, and choose another suggestion for the next attempt.

You could say
“The summary worked, the clarifying question didn't. What do you want to try next week?”

If the same problem keeps coming back after several attempts, it is no longer a matter for feedforward. Address the pattern directly. Limitations

About the method

Is this Marshall Goldsmith's feedforward?

Yes. Marshall Goldsmith introduced feedforward in a 2002 Leader to Leader article: replace critique of the past with suggestions for the future.

His exercise is a brief exchange of suggestions. Two of Reflacto's six questions follow it. The situation, the current pattern, strengths, and progress measures are Reflacto's additions. You use the current pattern only to prepare, not as critique in the conversation. Origin and sources

When is feedforward the wrong choice?

When consequences, repeated problems, or documentation matter. Address performance or accountability first.

You could say
“Before we talk about next time, I need to talk about what happened with the customer data.”

Feedforward can follow that conversation, but it must not replace it. When to use it

Origin and Reflacto's adaptation

Marshall Goldsmith introduced Feedforward in a 2002 Leader to Leader article and later expanded it into a book of the same name.

Also known as: Feedforward, feed-forward coaching

What Reflacto changed

From the original model
From Goldsmith’s exercise, a brief exchange of suggestions: the change the person wants to make, and specific suggestions for the next attempt instead of critique of the past.
Added by Reflacto
Reflacto adds the situation, the current pattern, strengths to build on and a way to track progress. These additions are not part of the original exercise.

See every question Reflacto asks

Primary sources and further reading

Editorially reviewed by Daniel Schmeiß

Last substantive review: September 29, 2026

Compare all methods