Working In Uncertainty

Writing about flexible plans

by Matthew Leitch, first published 26 October 2005.


The conflict that causes our difficulties

There's a reason why it is often hard to write a medium to long term plan when it is to be shared widely with others, including discussions of the future such as is required by the new Operating and Financial Review (OFR) requirements in the UK. That reason is long established in our business culture and shows up again in the apparent conflict between the OFR's requirements for statements about the future and the FSA's rules on forward looking statements that encourage companies to avoid them.

So what is the reason? When we sit down to write about our plans two conflicting sets of ideas battle for control of our writing. On the one hand we have phrases like ‘vision’, ‘clear plan’, ‘hard choices’, and ‘clear direction’ pushing us towards setting out a precise statement of where we want to be in the future, and the steps we will take to get there. In short: one path to one destination. But, on the other hand, we instinctively know that life is full of surprises and if we make statements like that today it won't be long before we have to announce we have changed our ideas.

In day to day management we tend to act naturally, responding to new information and evolving our plans. But when we have to write our plans down for a large audience suddenly intellectual ideas about management come into play and the balance swings towards the one-path, one-destination ideology. Even an organization that internally is highly adaptive can feel this pressure when it comes to explain itself to others.

If you find yourself in this dilemma my suggestion is to end the conflict now. Setting down a strategic plan with one path and one destination is a mug's game; it is simplistic, unrealistic, and a recipe for confusion and lost credibility. There are plenty of ways to write about flexible plans and still project all the competence, decisiveness, and leadership one could want, satisfying demands for information and clarity, without taking a crazy gamble on the future.

In this article I will suggest a four part approach that aims to avoid both premature commitments and the frustration readers feel when presented with vague statements or boiler plate disclaimers.

An introduction to the four elements to include

It doesn't make much sense to set out a strategy as if the future is known, and then have a separate section called something like ‘Risks and Uncertainties’ where uncertainty suddenly comes into view. Since uncertainty is fundamental to the formulation of any strategy it is easiest to explain the uncertainties and the strategy at the same time, and the exposition seems more reasoned as a result.

Four helpful areas to write about are as follows:

  • The environment and inevitable uncertainties: This is where we talk about the ways the environment drives the organization, and about those uncertainties about the future that are inevitable for the organization, no matter how competent and hard working its leaders. Many of those environmental factors and uncertainties will affect every similar organization.

  • Our adaptive strategy process: Here is where we reassure readers that the organization's leaders have a sound approach to adapting and improving plans over time.

  • Strategies that apply in all scenarios currently analysed: Some strategies make sense almost regardless of what happens in future, so they can be described with little qualification.

  • Uncertainties and strategies that are contingent on future circumstances: Some strategies will only make sense if the future unfolds in a particular way. The fact that the option to act in that way in those circumstances has been thought about and perhaps is being prepared for or kept open is important and may be interesting to readers, even though no definite decision to go down that route has been taken.

Although these elements would make sensible sub-headings in a strategy document it is not necessary to use them in that way. They may appear within other sections, or be repeated for different areas of the plan. Also, although the order presented above makes good sense and will often be the best choice it is not necessary for the elements to appear in this order to be helpful. If a regulator's document calls for particular headings then those have to be used.

Area #1: The environment and inevitable uncertainties

There's general agreement that writing knowledgeably about the organization's environment is a good thing, and it is also the opportunity to start talking about the things that cannot be controlled or predicted completely and therefore have to be managed more actively. For example, if a company's profits rise and fall with oil prices, and those prices are controlled by a world market not the company, then this is clearly an external uncertainty the company has to live with.

The wording should not sound like excuses. It should be simply a statement of fact. The organization should have taken steps to understand, track, and manage the impact of this kind of uncertainty. Take the opportunity to mention at least some of those steps (see Area #2).

A writer who shows understanding of the environment, an awareness of the importance of changing and unpredictable circumstances, and an understanding of how these affect results will come across as knowledgeable, alert, and responsible. In contrast, a writer who leaps straight into apologising for, or trumpetting, recent results looks like a press officer, not a leader. Similarly, if the document dives straight into plans it looks as if the leader has little or no ‘situational awareness.’

This section should not turn into a dreary litany of bad things that might happen and good things that cannot be guaranteed. The SEC's requirements for a discussion of risks led companies listed in the USA to write just this sort of material, and it seems particularly unhelpful and pointless. I asked an American lawyer about it and he explained that the SEC had made it clear they did not want text about potential good things that might happen and there were definite tactical, legal advantages if you listed many specific circumstances that might cause problems. As a result, although these disclosures seem alarmingly negative to those unfamiliar with the usual approach, all such disclosures are the same, leading knowledgeable readers to ignore their frightening tone.

In contrast the Accounting Standards Board's OFR Reporting Standard 1 (May 2005), which sets out the new requirements for Operating and Financial Reviews by UK companies, encourages them to write about both the good and bad things that might happen (paragraph 57).

Area #2: Our adaptive strategy process

Having established that the organization faces environmental uncertainties and therefore needs to adapt continually to changing circumstances the next step is to talk about how this is done and will be done in future. What is done and how often?

For many organizations the management ritual usually seen as fulfilling this role involves setting targets annually that stay fixed until replaced a year later, and incentivizing people to try to coerce performance into matching those targets. If the plan describes this process then it is effectively saying that the organization thinks about what it is doing only once a year and for the rest of the time it pursues out-of-date targets because its bureaucracy cannot churn faster.

Fortunately, even in organizations with a grinding budgetary control process there are usually other, more natural, management processes going on that reflect a more responsive approach to management. For example, if a director reads a shocking and relevant story in a newspaper and finds as a result that a dramatic technical breakthrough by a competitor needs to be responded to then only a fool would wait for the next annual planning round to suggest some action.

Is there regular monitoring of information about the environment, including statistics and news feeds? Do managers at various levels regularly discuss the potential implications of events and consider options for action? Are plans adjusted frequently? Are there rolling forecasts? Are the financial implications of alternative future scenarios modeled? All these points and many more, if true, portray an alert and responsible organization capable of taking on an uncertain future environment.

Here's an example that brings out some of the reasons for describing adaptive processes rather than setting out one path to one destination. Imagine a pharmaceuticals company that does research in four areas, each with a team of top specialists. The company's current belief is that all these areas of research are worth continuing with, though some have better prospects than others. Currently the flu research is hot because of demand for drugs in this area. However, a breakthrough in any of the other research areas that brings a marketable drug within relatively easy reach could change priorities almost overnight.

A strategy document is to be written. Should it say which research areas are to receive priority funding in future? If it did then the potential embarrassment of having to change those priorities in the next planning round is the least of the problems that could result. Announcing fixed priorities would close off other options because the research specialists in the areas not favoured would seek employment elsewhere. Circulating a one-destination, one-path strategy document could help kill the company.

On the other hand just saying ‘We will continue to invest in high quality research in promising areas.’ is not giving readers enough. This is when we need to see what can be said about the process by which investment in research is managed to cope with the uncertainties involved. Maybe we can write something like ‘We have invested in a portfolio of research work covering four main areas: flu, circulatory diseases, blood disorders, and arthritis. All research is planned to provide natural opportunities to evaluate progress. This progress and all other factors relating to the commercial prospects of the research are evaluated quarterly by a multi-disciplinary team that will in future include independent medical specialists, and priorities are reviewed. We will continue to invest in high quality research in promising areas and to keep valuable options open.’ (With this wording the only researchers likely to leave are those who have been pretending their research is going well but now will have to convince the independent medical specialists.)

There will be many sections of the strategy where discussing policies and processes for adaptive management is all that is needed. However, in others this will not give readers enough. We will need to press on and give more detail.

The explanation of our adaptive planning processes lays the ground for the next two areas.

Area #3: Strategies that apply in all scenarios currently analysed

To write about this area and the next, one needs to have done some serious thinking about alternative futures the organization might face, and looked at which actions are appropriate in each. This is classic scenario planning. If nothing like this has been done it will be difficult to write convincingly about plans for the future because a worthwhile plan doesn't exist. Either the intention is to be entirely reactive or the plans cover just a tiny subset of potential futures (usually just the one currently perceived as most likely or desirable).

The activities that fall into this area typically include projects to improve processes, systems, and to cultivate people in core areas. All this is reassuring to readers and needs to be covered. However, it can lack the flavour of being specific to the organization and its circumstances so usually we need to say more and go on to Area #4.

Area #4: Uncertainties and strategies that are contingent on future circumstances

This area is likely to be the most difficult to write, but offers the best opportunities for demonstrating a good understanding of the future of the organization and how it could respond. Here is where the writer demonstrates ‘vision’, not in the sense of a fantasy about what would be nice to achieve, but in the sense of an awareness that extends through the organization, into the wide world, and far into the future, coupled with the inventiveness needed to imagine how the organization might best cope with alternative events.

Since the plan document has already explained (in Area #1) the main drivers of results and what is uncertain about them, the next step is to set out the range of possible alternatives and to simplify these into scenarios that can conveniently be described, perhaps named, and used to explain how plans would be adjusted in response to each scenario if it began to unfold.

We can now discuss how we would adapt to each scenario if it were to begin to emerge, reinforcing this with facts about steps we are already taking to develop options, monitor events, and so on. Providing information about what we are doing now in preparation for potential futures make the description informative and useful, where it might otherwise read like just a list of hypothetical future situations with no relevance to today.

Now that we are writing about alternative futures, potentially very different from each other, it is likely that we will hit the problem of writing about long term objectives. The one-route-to-one-destination philosophy tempts us to state target numbers for financial and non-financial numbers. For example ‘We will increase customer satisfaction by 4% per year for the next three years.’ or ‘We will increase production by 10% next year.’ This is a route to embarrassment, as companies like British Telecommunications and BP have found. Within months conditions change and those targets can easily become inappropriate.

Unless our objectives are very general it is quite likely that different futures will lead us to adopt different objectives. How can we meet demands such as this one from paragraph 35 of Reporting Standard 1: ‘The OFR shall discuss the objectives of the business to generate or preserve value over the longer-term.’? This may seem particularly challenging for organizations that prefer not to use fixed targets (of which there are many successful examples, some huge).

Here are three techniques that may be useful:

  1. Make statements about objectives that are contingent on future circumstances.: For example, ‘If the market moves towards the Full Deregulation scenario our objectives will most likely focus on X, Y, and Z.’

  2. Concentrate on certain types of objective, and avoid stating targets.: This idea is well established in management science but rarely seen in ordinary business conversations, so needs more explanation.

    Suppose a business has ‘customer churn’ as one of its measures of performance. Something that is a measure is not necessarily an objective but as soon as they start to value different levels of customer churn differently it becomes an objective. As soon as they decide that lower churn is better than higher churn a rough objective arises. They might go further and map different levels of churn to a utility score, so that they can say for example, that churn of 1% is three times more valuable than churn of 2%, and half as valuable as customer acquisition costs at 12% of revenue. This may seem odd but it is quite common in quantitive decision making.

    A target arises when one particular level of churn is selected and given special status. In many management systems it is as if achieving that particular level is valuable but anything else is not.

    Typically, the system of values placed on different levels of achievement moves more slowly than targets. So, even if we cannot say anything long lasting about targets, we may be able to say something about how we value different outcomes. For example, if the plan document said ‘Our intention is to continue focusing on being leading innovators and we see this as our enduring competitive advantage.’ most readers would understand that high value is put on high innovation relative to competitors, but not on having very low prices or offering unusually high levels of customized service.

  3. Avoid giving precise numbers: Quantitively vague descriptions using words like ‘high’, ‘low’, and ‘improve’ sacrifice precision for a longer period of relevance. Targets and utility functions expressed numerically are the shortest lived objectives because they use precise numbers. If they are not to be obsolete they have to be revised frequently. Mentioning these precise numbers in a strategic plan for wide circulation is not a good idea.

Obviously some scenarios are likely to be more favourable to the organization than others, but this should not influence the amount of discussion each is allocated.

Disclaimers and safe positions

I once saw Sir Christopher Bland, Chairman of BT, presenting some quarterly results. Almost immediately he explained that, as he was going to be talking about future prospects, he had been advised for legal reasons to make some standard disclaiming statements. His tone and words were both apologetic and annoyed. He was apologetic for telling a room full of professional investors and financial journalists that the future could not be predicted with certainty and that plans he discussed might not come off, and he was annoyed at the necessity of making such patronising statements.

Clearly disclaimers can be useful, but they can also be annoying and lose impact through being artificial. This section discusses safe positions to take throughout, rather than standard paragraphs to insert.

The two potentially embarrassing mistakes are (1) to announce a particular course of action saying it will be taken, and then not do so because circumstances have changed, and (2) to make a prediction about what will happen in the future and be wrong. Fortunately, it is possible to say a lot that is informative and useful without announcing fixed courses of action and without making predictions. Furthermore, it is rare for regulators to insist on fixed plans or firm predictions.

How could a regulator reasonably expect an organization to announce future actions as fixed no matter what happens, or to predict the future? Usually they don't. For example, although OFR Reporting Standard 1 is clearly influenced by a one-route, one-destination, rather fixed view of strategy it stops short of assuming companies take that approach. It calls for discussion and analysis of the future, but does not insist on predictions. It insists on seeing the directors' strategies for achieving the objectives of the business, but it does not preclude contingent strategies or insist that those objectives or strategies stay fixed. Even paragraphs 40 and 41, which were clearly written with fixed targets in mind, do not prevent companies from changing their targets in between setting them initially and later measuring whether they are achieving them, provided those targets were not presented as fixed. This is a crucial point because a continuously managed organization will naturally be adjusting the precise value of its targets frequently as circumstances and prospects change. Although most organizations don't yet do this formally some do and with great success.

In fact, OFR Reporting Standard 1 says in paragraph 11: ‘Given the nature of some forward-looking information, in particular elements that cannot be objectively verified but have been made in good faith, directors may want to include a statement in the OFR to treat such elements with caution, explaining the uncertainties underpinning such information.’

The four area approach I suggested earlier weaves this disclaimer into the whole presentation of strategies. From very early on we start to explain the inherent uncertainties, the need for the organisation to be cautious and flexible, and so on.

Building on this position it goes on to develop the idea that the strategy document is today's view and, in a month's time, the current view will have been adjusted – probably slightly but possibly more fundamentally. Area #2 discusses the way the strategy is frequently reviewed, adapted, and improved. In effect this is another disclaimer.

However, this second disclaiming position would be frustrating for readers if they felt that the organization was explaining a strategy that was more than likely to be completely obsolete in a month's time. Therefore it is vital that it be clear that the current document is the result of hard work, not something casually thrown together just to have ‘something’ to talk about, and that the strategy that is then explained has the flexibility and robustness to remain stable for some time to come.

A one-route, one-destination strategy is unlikely to project the robustness and flexibility required no matter how it is expressed. However, a strategy that goes through scenarios as suggested in Areas #3 and #4, if well thought out, should project the necessary qualities. The reader will think ‘OK, I see why you need to keep certain options open, you've covered the things I was most worried about and it's good to know you are doing something about them. There are some exciting possibilities and you've given me a feel for how things could change in future.’

As far as possible, avoid stating courses of action that will be taken no matter what, and avoid making predictions about the future. Also, I would suggest avoiding statements about probabilities.

Although it is rarely possible to talk about the future of an organization with certainty it is, logically, possible to talk about the probabilities of various things happening. However, although in theory this is a reasonable thing to do and should not lead to unfair criticism once outcomes are known, misconceptions about probability are so widespread that bad reactions should be expected often.

People often confuse probabilities with certainty. Imagine I say that something is likely to happen, but in fact it doesn't happen and as a result several people lose money. From one test there is barely any evidence that I was wrong in my statement but still some angry person will say I made a mistake. (In contrast, if I made many probability statements then my judgement could be tested. For example, if I quote a 0.7 probability of an event on 10 occasions and the event only happens twice instead of 7 times out of 10 then there is now good evidence that my probabilities are badly calibrated.)

Another problem is that many people think probabilities are objectively verifiable facts, unconnected with how much is known to the person stating the probability. Someone thinking this way might say, for example, that the probability of getting Heads on tossing a coin is 0.5 and that's a fact so if someone says it's 0.3 they must be wrong. Similarly, if a company says a certain market will ‘probably’ expand over the next year many people will think this is a factual statement that the company should have checked carefully to make sure it is correct. These beliefs are demonstrably wrong, but the demonstration is not easy and I would be astonished if the beliefs died out in the next ten years.

A safer, and often more informative, alternative to stating probabilities is to write about evidence that bears on the likelihood of alternative outcomes. For example, ‘In allocating resources to developing this market we have considered various sources of evidence indicating the likelihood of further expansion. For example, we know that this market has expanded by more than 5% annually for the last 7 years and is still far from saturated.’ By writing in this way we can restrict the evidence to specific points we know we can substantiate, rather than saying expansion is likely and being exposed to any evidence that might conceivably bear on that statement.

To go back to the coin tossing example, this would be like saying ‘In betting on Tails we have taken into account relevant evidence such as the obvious bend in the coin and the fact that on the previous 100 occasions where it was tossed in the manner expected next time the result has been Heads in only 5.’


The future is complex and impossible to predict fully. Writing as if we have such foresight, or such control, that one sequence of steps will inevitably lead to the visionary goal we describe is a recipe for embarrassment and sometimes worse (because it can kill options that should have been kept open). It is also unnecessary, because realistic, adaptive, contingent plans can be described in ways that give a stronger sense of leadership and understanding.

To do this we need to look closely at the way plans are written, and weave discussion of risk and uncertainty throughout the document, rather than confining it to a special sub-section. We also need to look at the way strategy is made by, for example, adopting elements of scenario planning, and using financial models that go beyond just adding up managers' estimates for one possible future trajectory.

Related reading

The Accounting Standards Board's new requirements for OFRs are here.

The listing rules for the UK's stock exchange that relate to profit forecasts are set out in the FSA Handbook at paragraphs 13.5.32 to 13.5.35, and at paragraphs 9.2.18 to 9.2.19. Note that the definition of a profit forecast is very wide, so even comments you might think of as no more than a hint of the maximum or minimum results to expect are considered forecasts and caught by the rules. The gist of the rules is that if you make a forecast or estimate you will usually have to explain the difference between your estimate and reality at some point.

Companies listed in the USA should be aware of the ‘safe harbor’ provisions of section 27A of the Securities Act 1933 and 21E of the Securities Exchange Act of 1934. Subject to various restrictions this makes it worthwhile pointing out when you are making a forward looking statement and warning readers that they are subject to specific uncertainties, at least some of which you list. Here is a typical disclaimer of that type:

‘SEC Safe Harbor Statement

Projections, estimates and business plans in this Website are forward-looking statements that involve risks and uncertainties. Actual future market growth, capital expenditures, costs, earnings, events, financial performance and plans could differ materially due to, for example, changes in market conditions, the outcome of commercial negotiations, changes in operating conditions and costs, technology developments and other factors discussed in these web pages and in Item 1A of the Company's Form 10-K on file with the Securities & Exchange Commission.’

UK companies with listings in the USA will usually have to make an annual return using the SEC's form 20-F, which includes a requirement for a summary of risk factors as item 3D. The wording of the requirement is as follows:

‘Risk factors. The document shall prominently disclose risk factors that are specific to the company or its industry and make an offering speculative or one of high risk, in a section headed “Risk Factors.” Companies are encouraged, but not required, to list the risk factors in the order of their priority to the company. Among other things, such factors may include, for example: the nature of the business in which it is engaged or proposes to engage; factors relating to the countries in which it operates; the absence of profitable operations in recent periods; the financial position of the company; the possible absence of a liquid trading market for the company's securities; reliance on the expertise of management; potential dilution; unusual competitive conditions; pending expiration of material patents, trademarks or contracts; or dependence on a limited number of customers or suppliers. The Risk Factors section is intended to be a summary of more detailed discussion contained elsewhere in the document.’

It's clear that they have only downside risks in mind, but that is not firmly stated in the wording. Surely upside risks are what really make an investment speculative.

The meaning of probabilities is a long fought controversy. My contribution is ‘A way of thinking about probability.’

The conflict between popular notions of what business leaders should do – such as setting out ‘visions’, having away-days, and so on – and the reality of what actually works has been explored in a lengthy research project described in ‘Living Leadership: A Practical Guide for Ordinary HeroesLeaders in transition: the dramas of ordinary heroes’ by George Binney, Gerhard Wilke, and Colin Williams.

There are many good sources on scenario planning and Googling ‘scenario planning’ is a excellent way to start. I read about it first in Paul Schoemaker's book, ‘Profiting from uncertainty’.

For examples of companies that profit from not using fixed targets visit the website of the Beyond Budgeting Round Table. Their book, ‘Beyond budgeting’, is packed with case studies and I understand new ones will appear in a second book in 2006 called ‘Reinventing the CFO: How Financial Managers Can Transform Their Roles and Add Greater Value’.

Made in England


Words © 2005 Matthew Leitch. First published 26 October 2005.