Appreciative Inquiry

by Andy Smith

Selling AI within the organisation

Appreciative Inquiry is so different from conventional methods of organisational change that some people will find it hard to accept. Your obstacle to implementation may be the boardroom sceptic who needs to be convinced of the benefits of AI before releasing a budget, or it may be the person in an AI workshop who finds it hard to let go of the habit of always looking for the flaws in any new idea. Whoever they are, you need to get them on side.

Tips for dealing with challengers

Here are some tips that you may find useful in getting AI accepted by your organisation or clients.

Establish credibility

Make sure that people are aware of examples of where AI has already been used successfully elsewhere in the organisation, or relevant examples of similar organisations that have used AI.

Anticipate possible objections and pre-frame them out

Put yourself in the shoes of the people you will be telling about AI. Notice what objections they may raise, and make sure that your presentation of AI addresses these objections before they are raised.


“Now you may be thinking that this approach is about rose-coloured spectacles and ignoring problems – in fact it’s a more effective way of heading off problems before they even occur, because...”

Appeal to ‘away from’ motivation

Many people in business are risk-averse and motivated by solving problems, rather than by possibilities and benefits. AI enthusiasts, by contrast, tend to be very ‘towards’ – so the possibilities for miscommunication are obvious.

To get ‘away from’ motivated people on side, explore the problems they face (the more you do this, the more it will remind them of the seriousness of the problems), and make sure they realise how much worse these problems could get if they don’t adopt AI, and the pitfalls of using traditional ‘deficit-focused’ methods.

Dealing with ‘macho’ people

One of the characteristics of a ‘macho’ person, according to influence specialist Shelle Rose Charvet in her article The Macho Test, is that they like to act as if they already know everything there is to know.

If you are dealing with a person like this, you need to ask yourself this question about your presentation or document:

Is it anywhere stated or implied that there is something they don’t already know, or that I am telling them what to do?

Note that the use of any kind of jargon will come across to them as implying that there is something they don’t know. Unfortunately a lot of the terminology commonly used by AI practitioners could fall into this category, so make sure you keep it simple and focus on the practicalities rather than the theory.

How will you measure success?

Some implementations of Appreciative Inquiry have been bold leaps in the dark, going against the organisation’s status quo. For example, Nutrimental SA in Brazil closed down their manufacturing plant for a four-day summit bringing together all 700 of their employees. In that instance, the drive for using AI came from its founder and CEO, Rodrigo Loures.

Where you don’t have this level of power, you will probably need to persuade sponsors to commit money and time to the new and unfamiliar AI process. Your progress may be scrutinised every step of the way. So if potential sponsors insist on it, how do you show that the AI process is bringing benefits?

  • Get specific on the benefits that the AI programme is expected to bring. If the original change agenda grew from recognised problems, where exactly do the symptoms of these problems show up, and how much are they costing? For example, how much might poor customer care be costing in lost sales?
  • Ask yourself what benefits you might expect from the AI process. If customer care improved to the desired levels, how much would be saved?
  • Think creatively about other knock-on and systemic benefits that might result, so that any unexpected benefits that do happen are noted. For example, improved customer care might also indirectly affect the level of referrals from happy customers.

It may be that the organisation is already measuring many of the possible indicators of success, so that tracking them would take no extra effort.