How to write a vision and a strategic plan that the people in your company can remember and act on.
Share who could list three of their company’s priorities. Sull, Sull, and Yoder (2018), 124 organisations.
Most managers cannot name their company’s priorities
In a study of one technology company, which the researchers called Generex, 97% of senior leaders said they clearly understood the company’s five strategic priorities and how their own work contributed to them. The priorities had not changed in two years, and the chief executive talked about them often. When the same 132 leaders were asked to write the priorities down, a quarter of them could list three of the five. A third could not list any.
Other companies show the same pattern. Donald Sull, Charles Sull, and James Yoder surveyed 4,012 managers in 124 organisations between 2012 and 2017. Only 28% of the executives and middle managers responsible for carrying out the strategy could list three of their company’s priorities, although more than 80% of the companies had only three to five. An earlier survey of nearly 8,000 managers in more than 250 companies found that only 55% of middle managers could name one of their company’s top five priorities.
The biggest drop is between the top team and the people who report to it. In the typical company in the study, 51% of the top team could list the priorities, and 22% of their direct reports could.
The biggest drop is between the top team and the people who report to it.
Senior leaders usually do not know this. Top teams rated their company’s strategic alignment at about the 67th percentile, and managers elsewhere in the same companies rated it at the 50th. Nearly 90% of middle managers said their leaders talked about the strategy often enough. At one London professional services firm, 84% of staff said in an engagement survey that they were clear on the firm’s top priorities, and fewer than a third of its management team could name two of them.
The companies in the 2015 study had 6,000 employees on average. A company of 80 people has fewer layers for a message to pass through. It still has a top team and people who report to that team, and in a company that size those people run most of the business.
This guide covers the four most common reasons a plan is not understood, and how to fix each one. The plan lists goals instead of making choices. It asks people to remember too many things. Its vision is too abstract to picture. And the people who wrote it assume that because they have said it, it has been understood.
What leaders said, and what managers could list
Most quoted strategy failure rates have no evidence behind them
Talks and articles about strategy often say that 70% or 90% of strategies fail. In 2015 Carlos Cândido and Sérgio Santos looked for the evidence behind these figures. The 90% figure comes mainly from articles by Walter Kiechel in Fortune in 1982 and Planning Review in 1984. Kiechel reported what the consultants, executives, and academics he interviewed told him, and described no method. Cândido and Santos asked the consulting firms involved for the underlying studies and received nothing they could use.
One common 70% figure began as an estimate by Ram Charan and Geoffrey Colvin that 70% of the strategies that fail do so because they are carried out badly. That is a claim about strategies that failed. Robert Kaplan and David Norton later restated it as a claim that 70% of all strategies fail. Across the studies Cândido and Santos could assess, estimated failure rates ranged from 28% to 90%, depending on what counted as failure. They concluded that the real rate is not known.
The figure for change programmes has the same problem. Mark Hughes examined five published sources for the claim that 70% of change programmes fail and found no valid or reliable evidence behind any of them. The claim that 95% of employees do not understand their company’s strategy appears without a citation in a 2005 Harvard Business School working paper by Kaplan and Norton. It appears again in an exhibit whose only attribution is a copyright line for the Balanced Scorecard Collaborative. The measured figures in chapter one are the ones to rely on.
Across the studies that could be assessed, estimated failure rates ranged from 28% to 90%, and the real rate is not known.
Where the strategy failure figures came from
- No source
- Impression
- Misread
- Measured
Careful studies of planning are more positive. A 2019 meta-analysis of 87 correlations from 31 studies found that strategic planning had a moderate positive effect on organisational performance, with an average correlation of 0.23. That is a modest effect, and a reliable one by the standards of management research. Meta-analyses of small firms find the same, and the size of the benefit depends on factors such as the age of the firm,.
In Britain, the Office for National Statistics found that in 2023 firms with 20 to 49 employees scored 0.58 on its index of structured management practices, which includes the use of targets and key performance indicators. Firms with more than 250 employees scored 0.68. The ONS also found that differences in management practices are significantly related to productivity.
Two other findings affect how a plan should be written. In the 2019 meta-analysis, plans made with wide participation showed a weaker link with performance than formal plans. The studies measured participation in different ways. The finding means that involving many people in writing a plan does not by itself make the plan understood.
The second finding concerns agreement. A meta-analysis of research on strategic consensus found that agreement among managers about strategic priorities had a much stronger link with performance, a corrected correlation of 0.34, than agreement about detailed means and ends, at 0.06. The studies are correlational, and the link was weaker where performance was measured objectively. Even so, the result suggests that most of a plan’s value comes from a few priorities that managers share.
The rest of this guide covers why so few people can say what their company’s strategy is, and what to do about it.
A strategy is a set of choices, and each choice gives something up
Roger Martin describes the usual strategic plan in three parts. It starts with a vision or mission statement, continues with a list of initiatives limited mainly by what the company can afford, and ends with five years of financial forecasts. Management commits only to the first year. What the plan rarely says is what the company has decided not to do, or why.
Deciding what not to do is the main work of strategy. Michael Porter wrote that “the essence of strategy is choosing what not to do”. Without trade-offs there is no need to choose, and competitors can copy any good idea. Richard Rumelt calls plans without real choices bad strategy. He lists their usual contents, which include buzzwords, slogans, and financial goals presented as if they were a plan. Donald Hambrick and James Fredrickson say a strategy has five parts, which are where the business will be active, how it will get there, how it will win, the order and speed of its moves, and how it will make money. The parts have to fit together.
Goals are still needed. A company needs to know what it is aiming for, and targets make progress visible. The problem comes when the plan contains only goals. A line such as “reach £15m revenue by 2029” does not say which customers to pursue or which to give up. A result can usually be reached in several ways that conflict with each other, and a team given only the result will often try several of them at once.
Growth forecasts need the same scrutiny. McKinsey’s analysis of thousands of companies found that only about one in 12 moved from the middle tier of corporate performance to the top fifth over 10 years. A plan that forecasts a steep rise should say what the company will do differently to achieve it.
When Sull and Stefano Turconi analysed the strategic priorities published by S&P 500 companies, most did not say what success would look like. Of 1,508 priorities, 44% explained why the goal mattered, 18% included a measure of progress, and 6% included a numerical target. Sull and his co-authors suggest two tests for each priority. Check whether someone could guess the company, or even the industry, from the wording. Then ask a middle manager what they would stop doing because of it.
For example, take a 60-person firm that designs and installs commercial ventilation. A plan made of goals says the firm will grow revenue, delight customers, invest in its people, and embrace innovation. A plan made of choices says the firm will work only on commercial buildings above a certain size and will stop bidding for small retail fit-outs. It will compete on design speed and accept that it will lose tenders decided on price alone. It will employ its own commissioning engineers, although that raises its fixed costs. Each of these choices has a cost, and each tells an estimator what to do with the next enquiry.
Each of these choices has a cost, and each tells an estimator what to do with the next enquiry.
The most reliable way to make a plan include choices is to write a list of what the company will not do next to the list of what it will do. If nobody on the leadership team objects to anything on the not-to-do list, the list is too easy.
Fluff, goal, or choice
Reach £15m revenue by 2029.
Test your own
- 1
- 2
- 3
- 4
- 5
A vision works when people can picture it
Most vision statements are abstract. Andrew Carton, Chad Murphy, and Jonathan Clark analysed the vision statements of 151 hospitals in California and found about 16 abstract words for every word that described something a person could picture. In a subsample, fewer than 8% of the hospitals described anything a person could be seen doing. In a sample of 30 Fortune 500 companies, three did. One hospital’s vision was to distinguish itself through excellence in quality outcomes, which any hospital could say.
The choice of words affects results, because shared images help people coordinate their work. In the hospital data, which are correlational, visions with more imagery went with better results in keeping heart-attack patients from being readmitted. This was true only where leaders also communicated four values or fewer. The researchers then asked 186 employees, in teams of three, to design a toy. Teams given a vision in which children laughed and parents smiled produced better designs than teams given a matching vision about customers enjoying the toys, as long as the vision came with a single value. When nine values were attached, the difference disappeared.
Few leaders write visions this way. Carton and Brian Lucas estimate that fewer than 10% of leaders communicate visions with strong imagery, and that leaders use three to 15 times as much abstract language as concrete language. Part of the reason is that people think about distant times more abstractly than near ones, and a vision describes a time in the future. Part of it is the advice leaders receive. Of 180 sources of guidance on writing visions that Carton and Lucas reviewed, 175 told leaders to choose their words very carefully, which they argue leads to more abstract language.
Carton and Lucas tested a simple exercise. In three experiments, including one with senior executives and one with people working in the UK government, leaders were asked to imagine arriving by time machine just after their organisation had achieved its vision, and to take a photograph of what they saw. The visions they wrote afterwards contained more imagery. In the study with executives, their employees rated the visions as more inspiring.
Research on small firms supports this. J. Robert Baum, Edwin Locke, and Shelley Kirkpatrick studied 183 chief executives and employees of entrepreneurial firms in one industry. They found that the qualities and content of the chief executive’s vision, and how it was communicated, predicted how much the firm later grew. The qualities included brevity, clarity, challenge, and stability. They also included abstractness, which later researchers define as an aim that lasts, as opposed to a goal that is reached and then finished.
The aim and the description are different things, so the two findings fit together. A good vision describes a lasting state of affairs in concrete terms. “Every client can speak to a named partner on the day they call” is concrete and lasting. “Win the Hendry contract” is concrete, and it is finished once the contract is won.
Jim Collins and Jerry Porras recommended in 1996 that a long-range goal should come with a vivid description of what achieving it would be like. They set that goal 10 to 30 years ahead. For a company of 60 or 300 people, that is usually too far ahead to describe accurately, and the research on distance and abstraction explains why such visions become vague. We recommend dating the vision to a specific year three to five years away, which is the horizon Sull and his colleagues recommend for strategic priorities. Write one sentence describing a moment in that year that someone could photograph, and put the numbers next to it as a short list of targets.
Write one sentence describing a moment in that year that someone could photograph.
Check a vision sentence
Write a sentence, or load one of the examples, to see it checked.
Keep the number of things people must remember to about four
Most companies already have a reasonable number of priorities. More than 80% of the organisations in Sull’s study had between three and five, and 78% of the S&P 500 companies that published priorities listed five or fewer. A short list of priorities is not enough on its own, or the figures in chapter one would be better.
The problem is everything else in the plan. At one technology company’s annual executive meeting, the leaders presented the strategy and its objectives. They then introduced 11 corporate priorities that were different from the strategic objectives, a list of core competencies, a set of values, and a list of 21 new strategic terms. When middle managers were asked what made their company’s strategy hard to understand, they were four times more likely to mention the number of priorities and initiatives than unclear communication.
Values add to the total. The hospitals in the Carton study listed 4.9 values on average, and the sample of 30 Fortune 500 companies listed six. As chapter four showed, the benefit of a clear vision disappeared when leaders communicated more than four values.
Research on memory points to a similar number. Working memory, which holds information while you use it, can hold about four chunks, or three to five meaningful items in young adults,. People do not hold their company’s strategy in working memory all day, so the comparison is approximate. The practical point is about attention. A manager deciding how to spend an afternoon can weigh four priorities against each other. With a list of 30, they will act on whichever ones they happen to remember.
With a list of 30, a manager will act on whichever priorities they happen to remember.
Related items can be grouped and remembered as one. This lets you keep the plan short without pretending the company does only four things. Count every item that asks for people’s attention, including values, themes, initiatives, and measures, as well as the items listed as priorities. Then place most of them under a priority. An initiative that serves one priority, or a team goal that supports a company goal, counts as part of that priority. Items listed side by side at the same level each count separately.
How many priorities you can remember
Hold this plan
Four rounds, about two minutes eachYou will see the priorities of a 60-person firm, one every three seconds. After a short task, write down as many as you remember.
Count your plan
- Priorities5
- Values5
- Strategic themes or pillars4
- Company-wide initiatives6
- Measures on the company dashboard8
- Slogans or taglines1
- Other named frameworks0
- 78% of S&P 500 companies that publish priorities name five or fewer.
- Hospital values statements averaged 4.9 values, and a Fortune 500 sample averaged six.
- Imagery helped only with four values or fewer.
- One technology firm’s off-site added 11 corporate priorities and 21 new strategic terms, before its values and competencies.
Repeating a plan does not mean people understand it
A chief executive who has presented the strategy at every monthly meeting for a year may reasonably think it has been communicated. Usually it has not. In experiments by Boaz Keysar and Anne Henly, speakers trying to convey a particular meaning consistently overestimated how often their listeners understood it. Observers who knew the intended meaning did not make the same mistake.
Sull’s surveys show the same pattern in companies. Nearly 90% of middle managers thought their leaders talked about the strategy often enough, and top teams rated alignment well above where other managers rated it. The problem was consistency. Nearly a quarter of middle managers said that top executives changed their messages too often. Of the 69 items in Sull’s survey, the best predictor of alignment was how consistently managers explained their team’s priorities in terms of their own unit and of the company as a whole.
Repetition therefore has two jobs. The first is to use the same words each time, because every rewording makes people wonder whether the strategy has changed. The second is to explain what the priorities mean for each team. A priority about design speed means little to a site supervisor until someone explains what it changes about their week, and that explanation has to come from their manager. In a company of 20 to 500 people, those managers are often the direct reports of the leadership team, which is where the drop in chapter one happens.
A priority about design speed means little to a site supervisor until someone explains what it changes about their week.
Repeating a message through different channels works. In a study of project managers in six companies, Paul Leonardi, Tsedal Neeley, and Elizabeth Gerber found that managers deliberately sent the same message through two or more channels, especially when something unexpected threatened the work. Managers without formal authority over their teams did this early on, to get people’s agreement. Neeley and Leonardi advise managers to ask at least twice.
Timing also matters. A review of 317 experiments on spaced learning found that the best gap between repetitions grew with the length of time the material had to be remembered. Most of those experiments used words and facts, so the comparison with a strategy is approximate. It still suggests that a strategy meant to last three years should be repeated throughout those three years, instead of being launched at one away day and not mentioned again until the next.
The only way to know whether people understand the plan is to ask them. Sull’s method needs no software. Ask the people who have to deliver the strategy to list its priorities without looking them up, and count how many can name three. Leaders tend to measure communication by what they sent, such as emails and town hall meetings. What people can write down a week later is a better measure.
Write the plan on one page and test it on five people
Martin recommends fitting the strategy on one page, in plain words, based on where the company will compete and how it will win. A leadership team can write that page within a week. We suggest working in this order.
- Write the vision as one sentence about a named year three to five years away. Before you write it, do the time-machine exercise and describe something a person could see or hear in that year. Put three or four measurable targets next to it.
- Choose three to five priorities for the period, and for each one write down what it rules out. Then apply Sull’s two tests to each one. Check whether a reader could guess your company from the wording, and ask what a manager would stop doing because of it.
- Write the not-to-do list, and show it to the member of the leadership team most likely to object. If nobody objects to anything on it, go back to step two.
- Count everything else the plan asks people to remember, including values, themes, and initiatives. Remove what you can, and place the rest under a priority.
- If you write the plan at an away day, give the day one clear purpose. A study of more than 650 strategy workshops found that their results depended on a few design choices, including how clear the goals were and how much hard thinking the day required.
- Give the page to five people who were not involved in writing it. A week later, ask them to write down the priorities without looking, and ask how clear they feel about them. Compare the two answers.
- Use the same words in every team meeting and 1:1, and ask each manager to explain what the priorities mean for their team. Repeat the test every quarter.
If few people can list the priorities the first time, your company is in the same position as most of the companies in Sull’s research, and you now know about it.
Test whether people can list your priorities
- Your official priorities
- Send this to five people outside the leadership team
Send it a week after the plan is shared, and let people answer anonymously.
Two quick questions, answered without looking anything up: 1. List up to five of our company’s priorities for the next few years. 2. How clear are you on the company’s priorities? (1 = not at all clear, 5 = completely clear) Optional: pick one priority and say, in a sentence, what it changes about your team’s work.
- Their answersPerson 1ClarityTheir team answer
Headwater is built for companies of 20 to 500 people that want to run on one plan. It keeps the vision, the pillars, and each quarter’s climbs in one place, and brings them back in front of people every Monday, in every team meeting and 1:1, and at quarterly planning, so the same words are used all year. The choices still have to be made by the leadership team. Headwater keeps them in front of people until they can repeat them.
Further reading
- Sull, Sull, and Yoder (2018), “No One Knows Your Strategy (Not Even Your Top Leaders)”. Data on how many managers can name their company’s priorities, with a method you can use in your own company.
- Carton and Lucas (2018), “How Can Leaders Overcome the Blurry Vision Bias?”. Why visions are usually abstract, and the time-machine exercise that makes them concrete.
- Martin (2014), “The Big Lie of Strategic Planning”. The case for a one-page strategy made of choices.
- Cândido and Santos (2015), “Strategy Implementation: What Is the Failure Rate?”. Where the commonly quoted strategy failure rates came from.
Sources
- 1. Sull, D., Sull, C., and Yoder, J. (2018). No one knows your strategy, not even your top leaders. MIT Sloan Management Review. sloanreview.mit.edu
- 2. Sull, D., Homkes, R., and Sull, C. (2015). Why strategy execution unravels, and what to do about it. Harvard Business Review, 93(3). hbr.org
- 3. Cândido, C. J. F., and Santos, S. P. (2015). Strategy implementation: What is the failure rate?. Journal of Management and Organization, 21(2), 237 to 262. doi.org
- 4. Hughes, M. (2011). Do 70 per cent of all organizational change initiatives really fail?. Journal of Change Management, 11(4), 451 to 464. doi.org
- 5. Kaplan, R. S., and Norton, D. P. (2005). Creating the office of strategy management. Harvard Business School Working Paper, 05-071. hbs.edu
- 6. George, B., Walker, R. M., and Monster, J. (2019). Does strategic planning improve organizational performance? A meta-analysis. Public Administration Review, 79(6), 810 to 819. doi.org
- 7. Schwenk, C. R., and Shrader, C. B. (1993). Effects of formal strategic planning on financial performance in small firms: A meta-analysis. Entrepreneurship Theory and Practice, 17(3), 53 to 64. doi.org
- 8. Brinckmann, J., Grichnik, D., and Kapsa, D. (2010). Should entrepreneurs plan or just storm the castle?. Journal of Business Venturing, 25(1), 24 to 40. doi.org
- 9. Office for National Statistics (2024). Management practices in the UK: 2016 to 2023. Statistical bulletin. ons.gov.uk
- 10. Kellermanns, F. W., Walter, J., Floyd, S. W., Lechner, C., and Shaw, J. C. (2011). To agree or not to agree? A meta-analytical review of strategic consensus and organizational performance. Journal of Business Research, 64(2), 126 to 133. doi.org
- 11. Martin, R. L. (2014). The big lie of strategic planning. Harvard Business Review, January to February 2014. hbr.org
- 12. Porter, M. E. (1996). What is strategy?. Harvard Business Review, November to December 1996. hbr.org
- 13. Rumelt, R. P. (2011). Good Strategy/Bad Strategy. Crown Business. penguinrandomhouse.com
- 14. Hambrick, D. C., and Fredrickson, J. W. (2005). Are you sure you have a strategy?. Academy of Management Executive, 19(4), 51 to 62. doi.org
- 15. Bradley, C., Hirt, M., and Smit, S. (2018). Strategy Beyond the Hockey Stick. Wiley. wiley.com
- 16. Sull, D., and Turconi, S. (2017). How to recognize a strategic priority when you see one. MIT Sloan Management Review. sloanreview.mit.edu
- 17. Sull, D., Turconi, S., Sull, C., and Yoder, J. (2018). Turning strategy into results. MIT Sloan Management Review, 59(3). sloanreview.mit.edu
- 18. Carton, A. M., Murphy, C., and Clark, J. R. (2014). A (blurry) vision of the future: How leader rhetoric about ultimate goals influences performance. Academy of Management Journal, 57(6), 1544 to 1570. doi.org
- 19. Carton, A. M., and Lucas, B. J. (2018). How can leaders overcome the blurry vision bias?. Academy of Management Journal, 61(6), 2106 to 2129. doi.org
- 20. Trope, Y., and Liberman, N. (2010). Construal-level theory of psychological distance. Psychological Review, 117(2), 440 to 463. doi.org
- 21. Baum, J. R., Locke, E. A., and Kirkpatrick, S. A. (1998). A longitudinal study of the relation of vision and vision communication to venture growth in entrepreneurial firms. Journal of Applied Psychology, 83(1), 43 to 54. doi.org
- 22. Kantabutra, S. (2005). Improving public school performance through vision-based leadership. Asia Pacific Education Review, 6(2), 124 to 136. files.eric.ed.gov
- 23. Collins, J. C., and Porras, J. I. (1996). Building your company’s vision. Harvard Business Review, September to October 1996. hbr.org
- 24. Brysbaert, M., Warriner, A. B., and Kuperman, V. (2014). Concreteness ratings for 40 thousand generally known English word lemmas. Behavior Research Methods, 46(3), 904 to 911. doi.org
- 25. Cowan, N. (2001). The magical number 4 in short-term memory. Behavioral and Brain Sciences, 24(1), 87 to 114. doi.org
- 26. Cowan, N. (2010). The magical mystery four. Current Directions in Psychological Science, 19(1), 51 to 57. doi.org
- 27. Keysar, B., and Henly, A. S. (2002). Speakers’ overestimation of their effectiveness. Psychological Science, 13(3), 207 to 212. doi.org
- 28. Leonardi, P. M., Neeley, T. B., and Gerber, E. M. (2012). How managers use multiple media. Organization Science, 23(1), 98 to 117. doi.org
- 29. Neeley, T. B., and Leonardi, P. M. (2011). Effective managers say the same thing twice (or more). Harvard Business Review, 89(5), 38 to 39. hbr.org
- 30. Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T., and Rohrer, D. (2006). Distributed practice in verbal recall tasks. Psychological Bulletin, 132(3), 354 to 380. doi.org
- 31. Healey, M. P., Hodgkinson, G. P., Whittington, R., and Johnson, G. (2015). Off to plan or out to lunch?. British Journal of Management, 26(3), 507 to 528. doi.org
More guides
Book a demo and we’ll show you how Headwater would run your week, from the vision down to Monday morning.