Appreciative InquiryAppreciative Inquiry (AI) is an attractive approach to change for many reasons; it generates results fast, it engages people, it liberates creativity, it moves away from ‘blame culture’, and – not least – it feels good to focus on the positive.

Even so, AI requires a considerable investment of time and resources, and perhaps a shift of attitude on the part of senior management, if it is to succeed. For simple problems it can be overkill, and a more traditional ‘problem-focused’ approach (diagnose the problem, find the cause, design a solution, and implement it) may be all you need.

So when should you choose Appreciative Inquiry as your problem-solving approach? When one or more of these circumstances are present:

Complex problems
Traditional problem-solving works well in simple systems, or for problems there is a single, easily-identifiable cause. Where there are multiple causes, or the system is complex, it becomes more productive to identify what is already working, and build on it.

The problem keeps coming back
Recurrent problems suggest that the fixes already tried are not addressing root causes, or that some causes have been missed. Even if all the causes are identified, the interactions between them may be too complex to predict that any given fix could guarantee a lasting resolution.

Attempted fixes make things worse
In complex systems, the fix designed to solve a problem in one area can cause worse problems to emerge later, or in other areas. Systemic effects and time-lags can lead to unforeseen consequences. For example, when New Labour came to power in the UK in 1997, one of their campaign pledges was to shorten hospital waiting lists. They set targets for the times that patients should have to wait for operations. These targets were met – but unfortunately, by focusing on the targets, hospital managements took their eye off other untargeted factors that were at least as important, such as keeping the wards clean. The result was a surge in ‘hospital superbugs’ such as MRSA and E. coli. In some cases the focus on targets even distracted some hospitals from keeping their patients alive (as in the Stafford Hospital scandal that came to light in the late 2000’s).

No clear diagnosis or course of action emerges
By their nature, complex problems are harder to diagnose. Where no single cause can be identified, this suggests that a different approach is needed. Similarly, where every proposed fix seems to be outweighed by potential downsides, it’s worth turning attention away from the problem and looking for the places where the problem isn’t happening. These will be where the seeds of solutions are already starting to grow.


Something else to watch out for
Given that we have a tendency as human beings to not look beyond the first answer to the question “What is causing this?”, we need to bear in mind the danger of jumping to ‘premature solutions’ and missing other contributing causes. So whenever you think you have identified ‘the’ cause of a problem, it’s worth asking “What else has to be present for this problem to exist? What other contributing causes are there?”

If the problems your organisation is facing meet more than one of these conditions, a facilitated Appreciative Inquiry approach may be what you need.

When to use Appreciative Inquiry – and when not to
Tagged on:         

7 thoughts on “When to use Appreciative Inquiry – and when not to

  • So sorry but I do not understand this do-do not use AI ‘logic’ at all. A lot of words but being a CEO or manager I would not understand one word of this …. whatever you wanna call it.
    John Lodder

    1. Thanks for taking the time to comment John. Perhaps you could be more specific about which bits either you or your hypothetical CEO or manager don’t or wouldn’t understand, and I’ll try to explain it better?

      I’ve checked the article for obscure or jargon words and I don’t think there are any. The article is aimed at people who have at least heard of AI (and there’s a link to a ‘What is AI?’ article for those who haven’t).

      All it’s trying to say is – there are some types of problems (simple ones with a single cause) where it may be less effort to use traditional problem-solving methods, and to suggest that AI may be a better solution for more complex and intractable problems.

      What do you think, readers? Was this article hard to understand?

      1. I confess that I know very little about Appreciative Inquiry, nor am I a CEO or manager, however I thought I understood (and agreed with) the points you were making.

        I would add two thoughts about the article that might help.

        1. “(MidStaffs)” without a link or explanation might be incomprehensible to anyone who doesn’t live in the UK.
        2. The last paragraph “Something else to watch for” didn’t seem to me to fit into the catagory of when to use AI.

        That’s my sixpenneth (2 cents worth)

        1. Thanks Andy
          You’re right of course about the Mid-Staffs thing – that’s what happens when I compose a post offline, leave a note to self in the text to add a link, and then forget to add a link! Now corrected.

          I’ve also added a horizontal line to make it clearer that the ‘something else to watch for’ is not in the list of indicators for using AI.

          1. I think you’ve spelt it out quite succintly, Andy.

            In CEO/COO-speak:

            For simple workflows, straightforward training interventions may work just fine (e,g, negotiation training and the A.I. approach may be over-complex.

            Where a problem or circumstance that causes unease – unrecognised or not – spans an organisation even though the basic infrastructure is sound with recognisable strengths. Maybe a number of of the organisation’s management may not be aware of all the consequences of their departments work elswhere within the organisation beyond their respective silo, even when communication is reckoned to be ‘good’ judged by internal standards.

            There may be a hidden paradigm within an otherwise functioning organisation that holds back progress. Leverage can be obtained by the judicial use of A.I. to implement a paradigm shift. in order to move the organisation forward.

      2. Dear Andy, sorry for the delay in answering.
        imho AI (as any strength-based approach) works better than a traditional problem-solving approach in ANY situation, this is the fundamental change in leadership we need today and what
        AI-practitioners stand for. Based on that, my main points are:
        1-‘AI requires a considerable investment of time and resources’: This depends on what level you plan the intervention (strategic/tactical/operational) and or the size (whole organisation, department) and then HOW you organise the intervention (whole internal system, external stakeholders and/or…In general you avoid all the ‘resistances’ which makes AI faster and more sustainable.

        2-‘For simple problems it can be overkill, and…’: What is a simple problem? And why is AI coaching/change e.g. more overkill than any
        other form of coaching/change? Here I lose the AI perspective (theory&practice) completely.

        3-So when should you choose Appreciative Inquiry as your problem-solving approach ? AI is by definition not a problem solving approach, contrary!

        4-‘Where there are multiple causes, or the system is complex, it becomes more productive to identify what is already working, and
        build on it.’: I think I understand what you mean but… at least this requires a bit of AI knowledge and still I doubt.

        5-The 1997 examples are clear for UK readers, not for me; this is very minor of course.

        6-‘…places where the problem isn’t happening. These will be where the seeds of solutions are already starting to grow.’: This is very, very questionable from OD point of view and especially when mgmt. started with traditional problem solving! A benchmark might e.g. be a better option than AI.

        7- ‘Given that we have a tendency as human beings…’: This is an
        Assumption that I am not able to follow. But…has nothing to do with AI I presume.
        I hope this clarifies my previous respons, if not I am happy to give it another try.

  • For what it’s worth I understood the article. You constructed the need, laid out when to use your A.I. solution, and when not to use it. Very clear.

    Good writing Andy,
    Cheers
    @georgehealthadv:disqus

Comment on this post