Working In Uncertainty

Understanding uncertainty in interpersonal encounters


People – a riddle wrapped in an enigma, inside a mystery

Most of the uncertainty we deal with from day to day is uncertainty about people. We spend a lot of time in the company of others. For many, work and home life are a succession of interpersonal encounters, some trivial, some important, some relaxing but many stressful.

People are complicated, unpredictable, sometimes devious (even when they're trying to help), and acting from complex motives according to perceptions of situations even they cannot explain fully.

It is easy for misunderstandings to arise, particularly when dialogue is limited, and these can lead to big mistakes.

Fortunately, using simple techniques focused on the uncertainty involved it is possible to approach these situations with greater confidence - not confidence that we always know what will happen, just confidence that we will be less surprised and inconvenienced by whatever does happen.

On this web page I'll analyse some commonly occurring situations and show how to deal with the uncertainty.

Situation 1: Complex negotiations

Empirical studies of interpersonal encounters looking at uncertainty are few. One exception is a study of negotiation behaviour by Neil Rackham, the sales and negotiation psychologist. Rackham analysed the behaviour of 48 successful negotiators compared with another group of average negotiators across 102 negotiations. A number of behavioural differences were found, some of which shed light on the way the top negotiators deal with uncertainty.

Although the top negotiators spent about the same amount of time on planning as other negotiators they generated 5.1 options per issue, against the 2.6 options thought of by others. The top negotiators also planned to deal with issues independently so they could take them in any order. Some average negotiators would plan a sequence of issues and get stuck if the meeting did not go quite as they expected. The top negotiators tended to go with the agenda and issues raised by the other side to make sure they understood the thinking of the other side and worked with it.

Generally, Rackham favours negotiating within a range of potential outcomes rather than just setting an objective.

Situation 2: Boy meets girl

Imagine you are young and single (perhaps you don't have to imagine) and have just met someone at a party that you find very attractive, and who does not appear to be attached. This other person seems to have noticed you but the signals are unclear. Are they thinking ‘Mmmm yummy’ or ‘Uh oh’? Or maybe they've not thought anything at all.

You could find out more if you made some kind of move and got a response. But what will the response be? It might be encouraging, neutral, or disappointing (potentially even humiliating). You are uncertain.

If a relationship develops, this is just the first of many such occasions when you may probe for some kind of response that indicates a degree of interest or commitment. Each time, you must weigh the possibility of outright rejection against the gain of slightly cementing the relationship.

This pattern of incremental commitments is also common in sales meetings and other encounters involving persuasion.

If your move is, effectively, asking the other person for a big commitment the likelihood of rejection is higher. The other person will weigh you up based on the amount of evidence they have about you and how positive the indications are so far. Ask for too much before the evidence is there and you will be rejected, when perhaps you would have been accepted if you had waited longer for the evidence to build up.

Some strategies to deal with uncertainty in the ‘boy meets girl’ situation are:

  • Don't probe for a big increase in commitment - keep each increase small. Start very small indeed. This way you keep the risk of outright rejection low and a single refusal to commit is unlikely to be the end.

  • Probe for evidence that is clear rather than ambiguous. For example, agreement by the other person to do something requiring some kind of effort, freely given, is more encouraging than a smile, however warm it may be.

  • Keep each increase in commitment in line with the evidence that has passed since the last probe and the indications received.

  • Look for other evidence that the relationship might or might not work. There might be a religious or political difference, a hobby you don't like, or the person may turn out to be a smoker.

Situation 3: Reaching agreement

Many disagreements happen because people take different views about something that is not certain. Each side typically underestimates the uncertainty involved, and feels its own view is certainly right while the other view is clearly wrong.

For months, countries in the United Nations argued about whether to go to war on Saddam Hussein's regime in Iraq. Did he or did he not have ‘weapons of mass destruction’? Would he or would he not use them? What would be the consequences of allowing him to continue in power? Did he or did he not support or have links with terrorists? What could be achieved by military force? What could be achieved by inspection of weapons?

In truth few if any at the United Nations knew with certainty the answer to any of these questions, yet many felt they did and that some kind of action was needed.

On a less momentous scale, you might have an argument with a colleague over the best route for driving to a conference in another city because you have different expectations about traffic congestion on different routes. My wife and I have this kind of discussion often.

By understanding the role of uncertainty and recognising it in disagreements we can resolve them more quickly and devise robust plans.

Research by psychologists has tended to show that we have an overly narrow view of the future. However, the spread of views between 3 or 4 people tends to be more representative of the uncertainty we should take into account when making plans. For example, imagine four people each make optimistic and pessimistic estimates for sales of a new product by their company. The spread of each person's estimates will be too narrow. However, if we take the most optimistic estimate by any of the four, and the most pessimistic estimate by any of the four, the spread is likely to be more realistic. This is the spread the group should use for planning purposes.

A good way to resolve disagreements based on uncertainty rather than differing values is to suggest that the parties agree to differ on their interpretation of events and predictions for the future, and try to devise a plan of action that will cover all positions satisfactorily. The solution will probably involve some further investigation, monitoring, and risk-aware planning that keeps options open and allows for adjustment along the way.

Example: Outsourcing. Imagine Steve and Andy are colleagues responsible for out-sourcing IT support for a long established personnel database used by their company. They are trying to choose a company to out-source to and through a questionnaire and some initial meetings have narrowed it down to two companies. Steve favours Aardvark Computing while Andy is more inclined towards Bison Business Services. After debating the matter for over an hour they establish that the main difference between their views is that Steve thinks Aardvark have more Unix knowledge than Bison, and that their datacentre is better equipped. Andy thinks there is nothing between the companies in these areas, but that Bison have a better understanding of the specific requirements of the database. In fact their views are based on somewhat limited information and these differences have emerged only after the meetings.

Steve and Andy are deadlocked until their colleague Dave arrives to talk about something else. Dave says ‘It seems to me that you guys can't be sure about the differences you've been arguing over. Why don't you agree that the differences between your views are an indication of the uncertainty you still have and make a plan together to find out more. And isn't it time you started to think more about the stages of the out-sourcing project? Surely you don't want to just hand over to these people one weekend. Why not do it in stages so that your concerns can be dealt with gradually before each level of commitment is reached?’

‘Good idea’ say Steve and Andy together, in relief.

Situation 4: Delegating a task

Delegating to someone usually involves a number of uncertainties. Can they handle it? If they fail, how will they fail? How do they like to be managed? Did they understand the task?

One of the simplest and most reasonable approaches to this is Situational Leadership (see ‘Leadership and the One Minute Manager’ by Kenneth Blanchard, Patricia Zigarmi, and Drea Zigarmi). The idea is that the best way to delegate depends on the ability of the person being delegated to, on the task, and as their skill grows they should be treated differently.

At first the learner's competence is low but motivation tends to be high. At this stage the best leadership style is typically Directing, stating exactly what needs to be done, how, and when, and supervising closely. As the learner develops their competence improves a little but motivation tends to be variable. The leadership style to use now changes to a Coaching style, with directing and support. As the learner makes further progress their competence rises high, but motivation and self confidence may take time to catch up. The best leadership style is now to be Supportive, by listening, praising, and facilitating. Finally, the learner is both competent and motivated, and the job can be left to them to get on with.

Situational Leadership also involves dropping back down the leadership styles if something goes wrong, suggesting the learner has not progressed as far as was thought. It also involves goal setting, praising, and reprimands for attitude problems.

This approach does quite a good job of managing risk, simply by ensuring that supervision reduces only as skill increases.

It can also be helpful to think through and write down what you think you should monitor about the person, and then monitor those things.

Beyond that, one good indicator of whether a subordinate is heading for failure or not is their understanding of the limitations of their control and knowledge, and resulting uncertainty. If someone seems to have little understanding of their limitations and their response to any concerns you raise is ‘shouldn't be a problem’ then expect the worst. If someone wants more authority and freedom to act they can be offered it, on condition that they explain the uncertainty they face and how they will manage it, perhaps in writing, and provide updates as part of ongoing reporting.

If I had to choose between asking a subordinate to tell me how things were going and asking them about the latest position on areas of uncertainty I would choose the latter. Talking about areas of uncertainty is your chance as a manager to prevent things going wrong, which is more desirable than being called in to pick up the pieces afterwards.

Situation 5: The job interview

Job interviews are an example of a type of meeting where both sides tend to micro-analyse every detail of a short encounter in an attempt to make important judgments. In fact, face-to-face interviewing is far worse for removing uncertainty about the other party than we feel it must be. However, this doesn't make interviews any less important or stressful.

Many books and articles have been written about succeeding at job interviews, either as the job seeker or the recruiter. The number of detailed guidelines is astounding and far beyond the capacity of most people to remember and act upon.

Of course it is important to look the part and be polite, but fundamentally a job interview is an exchange of evidence to resolve uncertainty. It is helpful to think about two sets of uncertainties: (1) your own and (2) those of the other person.

Imagine you are the interviewer, with a job to fill. By the time the interview starts you will have seen biographical information and perhaps the results of psychometric tests. Your initial questions (Can they do the job? Will they fit in? Will they stay?) will have been partly answered but still there will be things you are unsure of, and perhaps some specific concerns you want to explore. List them, and consider what evidence might be most helpful. Therefore, what questions should you ask?

Also, consider how things look from the candidate's point of view. How likely is it that they will have a realistic understanding of the job to be filled? What worries and uncertainties might they have about progression or culture? Will they be worried about going to work for someone who turns out to be a terrible boss? What evidence would most quickly and surely settle those questions?

What if you are the candidate? Again, consider what your main uncertainties are and what evidence could reduce them best. Can you get that in an interview or would it be better to research the employer some other way? Also, consider the interviewer's perspective. Assume they have seen your CV, then speculate about what uncertainties and outright concerns they may have. What evidence can you provide that will do most to resolve those worries?

This is not a perfect prescription for winning at interviews because people don't always evaluate evidence rationally. For example, there is the Halo effect – a tendency to make an assessment on the basis of the initial evidence and try to make subsequent evidence fit the initial assessment. There is also a tendency to average evidence. Imagine there were three terrific reasons for hiring you, and seven helpful but much less important reasons. Is it more powerful to give just the three best reasons or give all ten reasons? Logically it is more helpful to give all ten. However, people have a tendency to take the average of the evidence so giving the extra seven weaker reasons could actually pull you down.

Situation 6: Disciplinary action

If someone who works for you seems to have done something bad that makes you angry and frustrated it is often unclear what should be done. In the heat of the moment our normal human bias towards assuming we know for certain can take over as we storm towards a reprimand or worse. Most often I've seen this happen when a manager tells a higher manager that a subordinate has done something wrong. The higher manager is often horrified, but really should be even less certain of the facts, being more removed from them.

Typically, we should not be as certain as we usually feel. Perhaps they didn't do what they seem to have done? Perhaps there were special reasons? Is the problem one of competence or attitude? If it is competence alone then a reprimand or disciplinary action is not appropriate. Could it be your fault?

Instead of acting on instinct, consider the uncertainty and, if necessary, find out more. Listen to the person concerned. Explain the implications of what they appear to have done and simply ask them to explain what actually happened. Pursue important points non-judgmentally. Ask if the person has learned anything from the experience, accepting that it may be they have nothing to learn from it.

If, after this, it is clear what has happened then act accordingly. If not, you may have to explain that the evidence is not conclusive. If it is clear that the person acted incorrectly, but not clear if that was because of a faulty attitude, one option is to say that you hope the person's attitude was not at fault but cannot establish the truth with certainty. You are fully supportive of those who sincerely want to contribute and, as the matter is important, you will be paying close attention to the progress of the corrective actions and to any repetition of the problem.

Situation 7: Proposing a plan in a meeting

Imagine you have a plan you would like to see adopted and intend to propose it in a meeting. Instinctively we know that others will expect us to be confident that the plan will work, and we get ready to look as positive and confident as we can. The plan is ours, success is identified with getting our plan accepted, and saying ‘I think it will work’ instead of ‘It will work’ is a mistake.

Or is it? Consider the harm done by this approach. Your reputation is now pinned to the success or failure of the plan. What if it is later sabotaged by a rival or simply doesn't work for any number of reasons? How many plans truly are certain to work? It is difficult for you to agree to, let alone suggest, sensible risk and uncertainty management actions because that would seem an admission of doubt. Lacking these actions your chances of failure and loss of reputation are higher.

An alternative is to present your plan and the theory underlying it as what it really is, a hypothesis. Give your reasons for thinking the hypothesis is true and your plan will work, but acknowledge specific doubts and concerns. Show how you could address them, in at least some cases. Invite others present to explore the hypothesis and plan together to see if it can be improved and if it is something that should be investigated further, or perhaps even decided immediately.

This way your reputation is built on your superb skills as a manager of risk and uncertainty, not on the specific success or failure of individual plans. Your plans will have a higher success rate because of your skills.

If you are a salesperson it is particularly helpful to work this way. Instead of starting a conversation by saying to your customer ‘I know you have a problem and I'm going to prove it to you’ you can say ‘I suspect there is a problem because I see the following evidence . . . Would you like to explore this further and see if perhaps there really is an issue and perhaps a way to address it?’ The latter is much more attractive to the customer and easier to say ‘Yes’ to.

Situation 8: Making a complaint

Most people find complaining stressful. Sometimes the person you complain to is courteous, interested, helpful, and fair. Sometimes they are not. I once complained politely about a bus service to an employee of the bus company who said hotly ‘That's ridiculous’. After a bit more customer service along these lines I may have said something impolite as I walked away, only to be pursued by his colleague, white with rage and clenching his fists, obviously looking for a physical fight. Fortunately, this ended without one as he accepted my explanation that I was insulting his friend, and not him.

Complaining is stressful because of the possibility that the situation will develop along unpleasant, and perhaps even violent lines. A sales assistant may refuse to give the refund you ask for, perhaps creating the unpleasant choice of how far to pursue the matter.

To reduce the stress involved, start by reducing to near zero the probability that a physical ‘fight or flight’ response will be needed. Not only is it extremely unlikely under any circumstances, but also by deciding in advance how you would diffuse a potentially dangerous situation you can virtually eliminate that danger. There are dangerous people in our society but few of them work in customer facing jobs.

Next, think about the chance of objections being raised. Think through how you could respond to various possible objections or offers. By this stage your stress should be low, provided all this forward planning didn't start you fantasizing about how it could go horribly wrong.

Situation 9: Communicating a reorganisation

In today's organisations we are continually sending e-mails to each other. Misunderstandings are common, as in all human communication. However, some communications have more impact and give rise to more confusion than others. There is more uncertainty as to the impact they will have.

Some of the most problematic are those e-mails from someone on high providing news of an ominous sounding ‘strategic review’ and resulting ‘reorganisation’. The main problems are that:

  • We tend to underestimate the psychological differences between ourselves and others. The greater the difference in status, age, wealth, and so on between the writer and the reader the more difficult it is for the writer to communicate clearly and without giving offence.

  • The subject matter – job changes and possibly job losses – means readers are on edge and straining to pick up every last clue as to what may be intended.

  • The ‘strategic review’ itself is often a highly political exercise in which much is left unclear and unresolved, while the need for diplomacy leads to bizarre buzz phrases coming into use, like ‘deconsolidation’ and ‘our bread and cake structure’.

I've seen some atrocious examples of communications from on high, but we can all make that mistake, and I'm sure I have. Even a difference of a few years in age can be crucial. For example, if you are married with children, in a management position, and highly educated it is very difficult to anticipate the thoughts of someone who is in their early twenties, unmarried, with no mortgage, and little education.

If you are the one who has to communicate the reorganisation, what can be done to manage the uncertainty of these communications? The techniques that work are to test your communications beforehand to make the initial announcement as effective as possible, and then to encourage questions, particularly about meaning.

The ideal announcement is short, precise, clear, and meaningful. It anticipates the main concerns of the reader. Think hard to get the right tone and use words precisely. Having done that the key is to test your communication by trying drafts on representative members of the intended audience and focusing on the detail of their reactions. Don't ask a peer to read your draft e-mail and be satisfied if their feedback is ‘Seems fine to me’.

You need between 1 and 3 trials, at least, depending on how important the communication is. Ask people to read the message then ask them to go through it with you, almost line by line, and ask them what they thought was meant and what went through their heads at the time. You will sometimes be astonished at how much people miss and mis-interpret even when the words are quite clear. Pay attention to every detail. If someone in a trial gets slightly confused by something but then resolves the confusion without help that is still a sign that redrafting is needed.

When encouraging questions in response to the initial announcements it can be helpful to encourage people to use their first questions to seek genuine clarification and further information. ‘Please think back over what you've heard and tell me if there's any part that wasn't completely clear to you, or that seemed to miss out something you really want to know about.’ If you don't provide this encouragement there is a danger of the entire question and answer session being swamped by ‘issue’ questions, in part fuelled by residual doubt and confusion.

These simple techniques work because they effectively reduce the chances of reactions that are not what you intended. Other attempts to improve communication that lack this element of uncertainty management tend to be ineffective. For example, it is often thought that a face-to-face communication is better. However, recorded audio and video messages are no different from e-mails if the message has not been tested. In fact they can be irritating to listeners, who are forced to write notes. Delegating the communication so that lower level management communicate the messages to smaller teams introduces a new problem of mis-communication between senior and lower levels of management, followed by mis-communication by lower levels of management.


Each of the situations described above illustrates a different way that uncertainty affects our dealings with people. Many of the situations show what happens in a range of similar situations. ‘Boy meets girl’ is similar to sales situations, for example.

By learning to understand and respond to uncertainty in interpersonal encounters we can live more successfully and less stressfully.

Made in England


Words © 2003 Matthew Leitch. First published 5 August 2003.