The surprising power of feedback as a conversation

December 16th, 2008  |  Published in Performance  |  4 Comments

Funny old thing connections – in the past couple of weeks I’ve been part of several conversations with clients that have each identified a new management technique that seems to work in all sorts of situations; whether making a difference to health and safety, improving performance or highlighting potential areas for innovation. This new technique is called spending time talking with people.

Then I come across this by Johnnie Moore and as a result I’m in the midst of reading Everything’s an Offer by Rob Poynton. In the book Rob describes, and invites us to work on, the three practices of improvisation: let go, notice more and use everything – and it seems to me that these practices form a superb set of principles for spending time talking with people when you are looking for learning and change as an outcome. As in most conversations in business.

Now contrast these principles with what goes on in performance situations where an element of feedback is involved – where you might be looking to create opportunities for a bit of learning and change for instance. Sifting through google searches for ‘performance feedback’, ‘feedback skills’ and ‘feedback process’ there seem to be three main stages of feedback conversation, with the later varieties attempting to make up for the deficiencies of their predecessor:


Comments is your basic, unadorned, feedback. Comments allow you to make available to others the full benefit of your wisdom. In appropriate and managed doses this is as good a method as any of passing on the lessons you have derived from your experience. The risk with framing feedback as comments however is that passing on my wisdom can become a reflex rather than a choice. Like the time when I was surprised that a team of area trainers I was ‘working with’ weren’t that thrilled to be on the receiving end of the insightful 38 actions I had decided they needed to work on. Not much change happened as a result of that interaction – especially in my own head.  The routine for comment boils down to me saying ‘You did this, this and this but if you had done this instead it would’ve been better’. Your response to this meaningless drivel is to nod enthusiastically, smile and say ‘Thanks a lot. That was really useful’ which is the signal for us to move on and forget the conversation.


Clearly you can’t have people like me running around the country lobbing out lists of 38 things to do differently – what’s needed are guidelines for handling the interaction. So it should be quite easy to search google for ‘feedback skills’ and come across this sort of list:

  • Give positive feedback when it’s due; don’t just point out mistakes.
  • Focus on the task or behaviour, not on the person: ‘This page of the leaflet is not as clear as it could be’ rather than ‘you’ve made a real mess of this page.’
  • Avoid personal, judgemental comments: ‘It makes it difficult for us all if you are late for meetings’ rather than ‘You’re hopeless – you’re always late for meetings.’
  • Make the comment as soon as it is needed, rather than days later: ‘I’m not sure this is going to work’ rather than ‘I thought at the time that that wasn’t going to work.’
  • Be specific: identify precisely what has gone well and make specific suggestions for improvement: ‘The introduction and conclusion of the report were good and you covered all the main points. It would be helpful to give each section a heading’, rather than ‘Can’t you set out your reports more clearly?’
  • Avoid commenting on areas the other person can do nothing about, e.g. their physical appearance.

Now these all seem like common sense but still the suspicion lingered that not enough was changing as a result of performance conversations – maybe we need recipes rather than just guidelines.


This takes us into the flamboyant territory of the feedback sandwich – also known as the shit sandwich and, I’m reliably informed, the evaluation hamburger!! Recipes attempt to deal with the phenomenon of people not responding well to being told to amend their behaviour by larding the bad news with a couple of layers of praise. Here is an outline of the sandwich/hamburger manoeuvre (faithfully copied from the internet):

A feedback sandwich is a linguistic construction designed to cause the person being given feedback to accept the feedback. If you tell a person only what they do wrong, they will become demoralised and switch off. Likewise, if you personalise the feedback to who they are (their identity) rather than what they are doing that needs changing (their behaviour) then they will not be motivated to change. How do you do it? Think of three things the person does well, in the context of the task, behaviour or situation you are wanting them to change, and the one behaviour that you want them to change. Finish by praising them at the identity level.

So I might say to Kevin: “I like the creative approach, energy and determination you have as a consultant Kevin, the one thing you could change would be the attention to detail which currently is not sufficient to produce the overall quality on this project that I know you are capable of achieving.” And Kevin might well choose to punch me.

That seems to be pretty much it as far as guidance for feedback goes and all three versions do a lovely job of keeping the giver of the feedback firmly in control.

Now there’s another choice: to let go of all of the above and rely instead on our ability to participate in a two way dance of learners and change by putting the three disciplines into practice – let go, notice more and use anything. By moving away from the need to have the answer before we begin I can free myself up to enter into a conversation with you that rests on joint exploration and discovery rather than some bizarre shuffle to a pre-determined goal. The advantage of this approach is that it taps into people’s potential for creating change for themselves, and as Don Tapscott says “In the innovation economy human imagination is the main source of value”. The downside is the loss of certainty and control – there’s no way of knowing where the conversation will go.


  1. Robert Poynton says:

    December 19th, 2008 at 10:40 (#)

    In my view action, learning and change are very closely related. On stage, you won’t get action unless you have learning (often on the part of the audience) or change (one character being changed by another) which is why the improv practises are so relevant to anyone interested in creating change or learning.

  2. Jon says:

    December 20th, 2008 at 08:50 (#)

    Probably makes sense to assume that organisational life in general is a stage then.

  3. Benjamin says:

    December 21st, 2008 at 10:10 (#)

    I like the guidelines list. Matches with current behavioural psychology pretty well. Another technique is to use attribution error: Talk about when you carried out the behaviour you are trying to have them change, and have them reflect back to you.

  4. Jon says:

    January 20th, 2009 at 09:32 (#)

    I think the guidelines are ok but, like most things, they have what Dave Snowden would call bounded applicability. So in contexts where the correct answer is known or knowable and you conclude that the other participant is unlikely to work out a way forward for themselves a ‘guidelines’ strategy might work.
    Can you elaborate on the attribution error technique? Sounds as if it could be a bit more conversational

Leave a Response

September 2021
« Oct