Introduction
Confusion about key terms often prevents logical discussion about effective planning, let alone robust conclusions about what actually works.
Sure, the terminology mess isn’t unique to planning. But it’s especially harmful here because planning involves making decisions that depend on other decisions.
At the company level, this creates interdependencies across multiple parts of how the business is steered, which demands clear logic about how these parts connect.
That raises the bar for terminology: each definition must not only be individually logical, but all definitions must also form a cohesive whole.
Otherwise, terminology itself becomes an obstacle to rational discussions about how planning should work.
In practice, this results in ineffective and inefficient business steering.
Even well-established terms, like those commonly used in FP&A, often mix basic concepts in illogical ways, leading to flawed conclusions and distracting debates that prevent real improvement.
This article provides clear and cohesive definitions for the core terms involved in planning—specifically plans, targets, forecasts, and budgets—enabling rational dialogue and effective business steering.
Fundamental Components in Planning
Before diving into specific terms, let’s think from first principles: what different fundamental components are involved in planning?
Or, to be more specific, to understand the fundamentals instead of terminology-specific characteristics, what components are involved in any effective ways to influence the future, regardless of what you call it?
The desired future is a major one.
Regardless of what you call it—vision, objective, goal, target, or budget—influencing the future always involves the pursuit of some desired future state. It can be very specific, like ‘$1 billion by the end of the year,’ or very vague, like ‘to make more money.’
Prediction of the future is another major one.
There may be different types of predictions that warrant the use of different terms, but effective ways to influence the future always involve forming predictions about what may happen.
These predictions can be very specific and explicit, such as ‘revenue will be $950 million for 2025,’ or vague and implicit, like ‘this acquisition will help the company make more money.’
Ways to influence the future are also required.
Unless there’s a gap between the desired state and the prediction of the future, there’s no need to influence the future.
For example, if we want revenue to be $1 billion and we predict it will be $1 billion with 100% certainty (more on why stating the probability matters in practice later), no additional influencing is required.
Further, the only way to influence that gap is to make decisions.
And only specific decisions, or a specific group of decisions, will lead to closing the gap (more precisely, lead to the best odds of closing the gap) between the desired future and the predicted future.
In layman’s terms, all effective influencing of the future, regardless of context, details, or the terms used, involves:
- The desired future state – i.e., ‘What we want to happen.’
- Predictions about the future – i.e., ‘What we think will happen.’
- Ways to influence the future – i.e., ‘What we’ll do.’
All are not only dependent on each other but also on what is assumed about the world.
How desirable and difficult ‘what we want to happen’ is depends on real-world context. For example, how desirable and reachable an EBIT of $100 million is depends on how fast the market grows and how raw material prices develop.
How likely our predictions are also depends on the real-world context. For example, how likely the prediction of EBIT of $95 million is depends on how fast the market grows and how raw material prices develop.
How effective ‘what we’ll do’ is in realizing the desired future also depends on the real-world context.
For example, how effective our decisions to launch a new product and increase prices are in realizing the EBIT target of $100 million depends on how fast the market grows and how raw material prices develop.
These dependencies reflect properties of the real world, or, as some might even say, are ‘obvious to common sense’—not based on any revolutionary XYZ-planning concept.
For terminology, this means that in order to have a sensible discussion that is rooted in reality, there is also a need for clear terminology to describe the real-world context.
It’s the Cohesive Whole That Matters
The most important thing about terminology involving effective ways to influence the future, i.e., ‘what we want to happen,’ ‘what we think will happen,’ and ‘what we’ll do’ is that it forms a cohesive whole.
Without that, you end up with circular references or missing critical parts-which means that no matter how you define the individual parts, the sum of the parts will always be ineffective and inefficient.
Translated to a business context, that means how effectively a company is steered-and its performance, however either is defined-will always be inferior to what they could be.
To form a cohesive whole, you need clear terminology at least for the following:
- What we want to happen (normative), for example: target, objective, goal.
- What we think will happen (descriptive), for example: forecast, estimation, projection.
- What we’ll do (to close the gap between what we want to happen and what we think will happen) (prescriptive), for example: plan, strategy, policy.
What the specific terms are is irrelevant; they could be tooth fairy, Easter Bunny, and Santa Claus, as long as they are used and understood the same way by everyone.
The reason terminology needs to form a cohesive whole is simple: how sensible the ‘what we want to happen’ is depends on ‘what we think will happen’ and ‘what we’ll do’ to close the gap.
Further, ‘what we think will happen’ also depends on ‘what we’ll do.’ That includes even when we don’t actively do anything-as not deciding is a decision.
Similarly, the ‘what we’ll do’ also depends on ‘what we think will happen,’ i.e., the context and things outside our control.
In other words, none of the above makes much sense in isolation.
They only make sense as a whole.
For this reason, it also helps to have clear terminology for what to call that whole.
Russell Ackoff called it ‘planning’ as early as 1970 (A Concept of Corporate Planning).
It is also no coincidence that Ackoff is widely considered one of the most prominent systems thinkers and one of the most respected authorities on effective planning principles to date.
The major mistake many companies still make today is that not only do they lack cohesive, logical language for the whole, they also divide those components into silos.
- What they want to happen-target setting-happens as one activity.
- What they think will happen-forecasting-happens as another.
- What they’ll do to close the gap-decision-making-happens as yet another.
Even if they address all the relevant parts, the fact that they address them separately means that the whole doesn’t work effectively or efficiently.
Logical and Cohesive Terminology
What would terminology that covers the fundamental components of planning and forms a cohesive and logical whole look like?
Here is one example:
- Target = What we want to happen, the desired future
- Forecast = What we think will happen, a prediction about the future
- Plan = What we’ll do, an interdependent set of decisions
- Budget = An allocation of resources, a subcategory of ‘what we’ll do,’ specifically an interdependent set of decisions about how resources are allocated
- Planning = Improving the odds that ‘what we’ll do’ will result in ‘what we want to happen,’ based on ‘what we think will happen’
In principle, all you need are clear definitions for target, forecast, and plan—plus a term for the cohesive sum of the whole.
For practical purposes, budget is included too, because it’s an established term and a useful one when used correctly.
Target (What We Want to Happen)
Target = What we want to happen, the desired future
↳ For example, a profit of $100 million (not that static numbers make good targets)
Other terms include objective, goal, aspiration, vision, and so on.
There are countless terms for ‘what we want to happen,’ but what they all share is that they describe a desired future. These terms differ in the characteristics of that future state—such as type, scope, level of detail, and time frame.
The purpose of this article is not to define all such terms, list the most commonly used ones, or present a comprehensive taxonomy (which could be the subject of another article).
Instead, the focus here is on how they can be used logically as part of a system designed to influence the future—in business terms, how to effectively drive performance, however that is defined.
The most critical point is that this terminology must differ from the terms used to describe ‘what we think will happen’ and ‘what we’ll do.’
Although that may sound obvious, in practice it’s often violated.
A notable example is how the word ‘budget’ is frequently used interchangeably with targets or forecasts (discussed later).
Two key characteristics of targets are often overlooked: their desirability and level of difficulty are always context-dependent.
For example, a profit target of $100 million may be highly desirable if the market is flat, but less so if the market grows by 10%. It may be more desirable if it aligns with existing strategic choices, but less desirable if achieving it requires abandoning the strategy.
The same applies to difficulty.
A $100 million profit target might initially be a P70 target, but depending on how the market evolves, it could shift to a P50 or P90 target.
Similarly, if the company adheres to its strategy, it may be a P60. And if the original target assumed a $20 million cost budget, but that is later reduced to $15 million—through layoffs, for example—it could become a P50 target.
In short, the desirability and difficulty of ‘what we want to happen’—such as a profit target—always depend on:
• ‘What we think will happen’
• ‘What we could do’
The practical implication is clear: a system where targets are fixed while forecasts and plans evolve cannot be effective in driving outcomes.
To be effective, a target must always be accompanied by a specification of context and difficulty.
For example: A profit of $100 million is a P70 target, assuming the market remains flat and the organization can adhere to its strategic choices.
Forecast (What We Think Will Happen)
Forecast = What we think will happen, a prediction about the future
↳ For example, a profit of $90 million (not that point estimates make good forecasts).
Other terms include estimate, projection, assumption, and so on.
As with targets, this article does not attempt to define every term used for ‘what we think will happen.’ The focus is on how forecasts function as part of an effective system to influence the future.
However, regardless of terminology, in a business context, ‘what we think will happen’ always has three components:
• Outcomes—including both what we want to happen (for example, profit) and other outcomes that influence it (for example, head count or capacity)
• External factors that influence those outcomes (such as market growth, inflation, raw material prices)
• Internal factors that influence those outcomes (such as product launches, price increases, hiring)
While most companies focus on predicting outcomes, those outcomes are always shaped by assumptions about both external and internal factors.
Internal factors represent ‘what we could do,’ including ‘what we’ll do.’ These are the actions the company controls. External factors represent the real-world context—things outside the company’s control that still influence outcomes.
Forecasts are, in essence, predictions about how external and internal factors will interact to produce outcomes.
One subset of those outcomes is what the company wants to happen—for example, profit. Other outcomes, like head count, may support or hinder those outcomes.
The critical point is that all outcome predictions rely on assumptions about both external context and internal decisions. This applies whether those decisions are specific commitments (‘what we’ll do’) or general options (‘what we could do’).
Some get confused here, believing their forecasts are based on facts and data.
They are not.
Forecasts may be informed by historical data and facts, but those are only inputs. There is no such thing as data about the future.
Once this is understood, it becomes obvious that forecasts should never be treated as promises, commitments, or targets. They are always conditional—dependent on an evolving real-world context and decisions that may still change.
A useful convention is to reserve the word ‘assumption’ for predictions about the external environment.
The difference between assumptions and outcomes is this: assumptions describe ‘what would have to be true’ in the real world for the outcome to hold.
For example: “The market would have to grow 5%, and competitors would have to follow the price increases, for the revenue forecast of $1 billion to be valid.”
Expecting a $90 million profit regardless of market conditions or competitor behavior is illogical. That’s an implicit assumption—and it makes the forecast ineffective for business steering.
In fact, it’s impossible not to make assumptions. They may not be stated explicitly, and not all can ever be listed, but every forecast is based on some assumptions.
The less those assumptions are specified, the more room there is for interpretation—and for excuses when reality doesn’t align with expectations.
“Profits were lower than expected, because the market didn’t grow.”
Well, what had been assumed about market growth?
“Profits were lower than expected, but competitors didn’t follow price increases.”
Well, what had been assumed about competitor behavior?
And it’s not enough to state only external assumptions.
Forecasts also depend on internal factors—what the company plans to do. Whether you call these internal assumptions or the plan is largely semantics.
But it would be illogical to say the profit forecast is $90 million assuming 5% market growth and competitors following price increases, while ignoring whether the company’s own price increases are 5% or 10%, or whether the marketing budget is $10 million or $20 million.
All of this holds true even before accounting for the fact that the impacts of both internal and external factors on outcomes are probabilistic, not deterministic.
Which means that even in the theoretical case where there are no changes in internal or external factors, it’s still illogical to say the profit projection is $90 million.
Despite being the cardinal sin of most business predictions, there is no such thing as ‘no probability,’ nor is there ever a ‘100% probability.’ Yet forecasts are often misread as prophecies—statements of what will happen, no matter what.
A $90 million profit forecast, assuming 5% market growth, 10% price increases, and a $10 million marketing budget, might have a 20% or 80% probability of being realized.
So while it’s not entirely useless to provide forecasts without stated probabilities, doing so invites ambiguity—even if all assumptions play out exactly as expected.
Just like with targets, the practical implication is that forecasts cannot be effective if considered in isolation from targets or plans.
To be useful in steering the business, a forecast must be accompanied by the assumptions behind it and the probability of the outcome.
For example: There is a 70% probability that profits will exceed $90 million, assuming 10% price increases, a $10 million marketing budget, 5% market growth, and competitors following the price increases.
Plan (What We’ll Do)
Plan = What we’ll do, an interdependent set of decisions
↳ For example, decisions to allocate $10 million to marketing and to raise prices by 5%.
Other terms include strategy, budget, roadmap, and others.
Again, it’s outside the scope of this article to dive deep into the intricacies of different terms for grouped interdependent decisions.
The focus here is on how ‘what we’ll do’ functions as an effective part of a system—together with ‘what we want to happen’ and ‘what we think will happen.’
One trivial point to get out of the way: making decisions is the only way to influence the future, in personal life or business.
Sure, there is an abundance of different terms for decisions—choice, policy, rule, priority, goal, allocation—but underneath all of them, someone, somewhere, has made a decision.
At the company level, when the problem is how to steer the business, there are always interdependencies between decisions.
This means that ‘what we’ll do’ is always an interdependent set of decisions. In effective business steering, it’s an integrated set of decisions.
You may have heard the concept used only with specific terms like ‘strategy,’ but that’s a low- to no-value terminology debate, often characterized by strawman arguments and subjective opinions—not by first principles or logic.
Regardless of the terms used, if you’re trying to have a significant impact on the company’s future, you end up with an interdependent set of decisions.
What’s less often given proper attention is that, just like with ‘what we want to happen’ and ‘what we think will happen’—e.g., targets and forecasts—‘what we’ll do,’ e.g., the plan, is always context-specific.
How effective ‘what we’ll do’ is in creating the desired future depends on:
• What we want to happen
• What we think will happen
And, as we remember, ‘what we think will happen’ has three components:
• Outcomes (what we want to happen, but also others)
• External factors influencing the outcomes
• Internal factors influencing the outcomes
That means the effectiveness of ‘what we’ll do’ always depends on:
• What we want to happen
• What we think will happen with outcomes
• What we think will happen with external factors
• What we think will happen with internal factors
The critical point is that the effectiveness of a company’s decisions—regardless of what they are or what they’re called—always depends on predictions of external factors (assumptions about the business context) and predictions of internal factors (i.e., the effectiveness of specific decisions depends on what other decisions will or could be made).
Once this is understood, a few things become trivial:
• Decision-making, decision by decision, is not effective
• Static plans are illogical
The former is often neglected by software sales representatives whose main selling point is the speed of decision-making—in other words, the need for decisions made in real time.
That may work for small decisions with no significant interdependencies. But it can never work for major decisions that always come with significant interdependencies. There is no such thing as an effective decision made in isolation.
The latter is often neglected by old-school executives who take pride in excelling at ‘execution.’ They believe that changing the plan is a sign of weak performance and sloppy management.
That can be true at times.
But more often, it’s the opposite: sticking to a static plan is a guaranteed way to decrease performance, because the plan was effective only under certain assumptions about the world and in connection with specific other decisions.
Once either changes, there is no ‘excellence in execution’ without changing the plan.
An astute reader might note that by ‘executing the plan,’ executives often don’t mean executing the decisions—they mean producing the outcomes, i.e., hitting the targets.
That brings us to another, sometimes even more damaging, misconception about ‘what we’ll do,’ e.g., plans.
To some, this might go beyond the call of duty for an average executive, but it deserves to be mentioned because it has a tremendous impact on performance—and it reveals the illogic of ‘hitting the plan.’
As we saw earlier, the desirability and difficulty of ‘what we want to happen’ always depend on:
• What we think will happen
• What we could do
Another way to say the same is that it depends on:
• What we think will happen with outcomes
• What we think will happen with external factors
• What we think will happen with internal factors
• What we could do
That means the desirability of the plan is also dependent on the same predictions, which are subject to constant change. It’s not just a question of whether actions are effective.
It’s also a question of:
• Does the target still make sense to pursue?
• How difficult the target is?
In simple terms: even if the actions still make sense based on the target, the target may need to be revised before it’s even possible to assess how the actions should change.
This also affects how performance is evaluated. Most companies don’t understand that changes in assumptions about external and internal factors always result in changes in the difficulty of achieving static targets.
Which means that once performance is measured against the target, the organization will often conclude that performance is poor—that there is an execution issue—even when performance and execution are excellent.
It is the changed difficulty that has altered the probability of reaching the target, even with excellent execution, that is missed.
In other words, even if the target stays at $100 million, a change in assumed market growth from 10% to 0% may change the target difficulty from P90 to P50.
From the perspective of maximizing business performance, the best option might be to change the target and keep the plan as is. Or, it could be to change both.
The other side of the same coin is that everything that can be done about the future is also probabilistic.
Hiring and training 100 people at a factory before the season is subject to uncertainty. There may be a 99 percent probability it can be done—or only 50 percent.
There may be a 90 percent or 60 percent probability that a product launch will be on time and within budget.
There may be a 95 percent or 85 percent probability that a capacity investment is delivered on time and within budget.
‘Excellence in execution’ can influence those odds, but never make them 100 percent.
As with targets and forecasts (or any other terms for ‘what we want to happen’ and ‘what we think will happen’), the practical implication is this: a system where plans (or any other term for ‘what we’ll do’) are considered in isolation from the other two fundamental components will never be effective in creating the desired future.
It also means that for a plan to be effective in business steering—or in other words, worth executing in the first place—it must be accompanied by a specification of the context and the probability.
For example: There is a 50 percent probability that the profit target of $100 million will be reached (or exceeded), with decisions to raise prices by 10 percent, allocate $10 million to marketing, and assuming 5 percent market growth and competitors following the price increases.
In other words, it’s a P50 plan.
Budgets (Subcategory of What We’ll Do)
In a business context, budget is a widely used term. But does it have a specific function, and where does it fit under the fundamental components?
Budget = Allocation of resources
↳ For example, a marketing budget of $10 million = an allocation to spend $10 million on marketing
Other alternative wordings include reservation of resources or commitment of resources.
A sensible use of the word budget is to treat it as part of ‘what we’ll do.’
As a set of decisions, budgets are forms of plans—interdependent decisions about how resources are allocated under certain assumptions about the real world.
Your target may be $100 million in EBIT, and what you’ll do to achieve it is allocate $20 million to R&D and $10 million to marketing.
A budget as a target to spend $20 million on R&D or $10 million on marketing would be illogical. You may need to allocate those amounts to realize the desired future, but that spend is not what you want to happen.
If you could achieve the same desired future by allocating only $15 million and $8 million, you’d probably prefer that instead.
Even when companies talk about cost budgets as targets, they don’t really mean that ‘what they want to happen’ is to spend $10 million on marketing.
What they mean is that as long as marketing spend—and all the other parts of the budget—stay within the budget, the whole remains sensible.
And that’s the right use of budgets.
It’s an aggregated allocation tool that empowers the organization to make decisions within agreed limits aligned with the rest of the organization.
The advantage is that instead of trying to align thousands of smaller decisions, you can align maybe tens of bigger ones. Or instead of millions, all you need to align is hundreds.
As with any other types of plans, it’s illogical to have a budget that would remain the same regardless of what the company wants to happen, what it thinks will happen, or what it does—or to assume that its impact would ever be deterministic.
For budgets to be effective in business steering—they as well should be accompanied by a specification of the context and the probability.
For example: There is a 70 percent probability that the profit target of $100 million will be reached (or exceeded), with decisions to allocate $20 million to R&D and $10 million to marketing, and assuming 5 percent market growth and the company adhering to its strategic choices.
Planning
Planning = improving the odds that ‘what we’ll do’ will result in ‘what we want to happen,’ based on ‘what we think will happen.’
In other words, planning is improving the odds that the interdependent decisions the company makes will create the desired future, based on predictions about the future.
Put yet differently, planning is improving the odds of reaching desired targets by making and revising plans based on forecasts.
Planning involves making and revising decisions, but as discussed earlier, for it to be effective, it must also involve creating and revising targets, as well as creating and revising forecasts.
The practical implication is that it’s ineffective to have separate ‘target-setting’ or ‘forecasting’ processes, as many companies do.
While that might ‘kinda’ work, it’s like trying to run with stones in your shoes while your hands are tied behind your back. Sure, you can technically call it ‘running,’ but anyone with proper shoes and free hands will outpace you easily.
Separating target setting, forecasting, and planning can never be effective because all three depend on real-world assumptions—such as how the market develops.
As discussed earlier, a profit target of $100 million may be desirable when the market is flat, but no longer if the market grows by 10%.
Therefore, to find the most effective ‘what we’ll do,’ we first need to understand whether the ‘what we want to happen’ is still in fact ‘what we want to happen.’ It is not possible to find effective decisions without considering the two together.
Many strategy experts have long understood this.
For example, Richard Rumelt writes in The Crux: “The answer is that good strategic goals are an outcome of working the gnarly problem of strategy, not an antecedent.”
Similarly, Roger Martin describes the process of making strategic choices as one integrated loop, one of which is the ‘strategic aspiration’ (another term for ‘what we would like to happen’).
While strategy experts often like to point out several ways in which strategy (strategizing, really) is vastly different from ‘planning,’ they are in fact often describing universal characteristics of effective ways to influence the future—not only those specific to making strategic choices.
Similarly, since both ‘what we want to happen’ and ‘what we can do’ depend on predictions about the future—i.e., ‘what we think will happen’—planning always involves forecasting.
It also follows directly from the characteristics of ‘what we want to happen,’ ‘what we think will happen,’ and ‘what we’ll do’ (what we can do), that effective planning must be probabilistic.
As we’ve learned from earlier, ‘what we want to happen’ (e.g., targets), ‘what we think will happen’ (e.g., forecasts), and ‘what we’ll do’ (e.g., plans or any interdependent decisions) are always subject to probabilities.
So, since all the components are probabilistic, effective planning must also be probabilistic.
This also means that its purpose is not to create static plans or to produce guaranteed outcomes. Its purpose is to improve the odds of reaching specific targets—but it can never guarantee them.
For example, planning is deciding to allocate $10 million to marketing and to raise prices by 10% in order to improve the probability of reaching EBIT of $100 million by the end of the year from 40% to 50%, based on the assumptions that the market will grow 5% and competitors will follow price increases.
Sure, that’s not exactly how most executives think about planning, or decision-making for that matter, but that’s what all of them nevertheless do anyway.
The fact that their spreadsheets contain only 100% precise numbers, creating the illusion of a ‘$2 million gap to target’ or whatever the case may be, is just that—an illusion.
In reality, all their numbers and decisions are probabilistic, and all they can do is change the odds—never remove the underlying uncertainty and influence of luck.
“But we are constantly hitting our targets.”
Well, maybe you are.
However, if that’s the case, there is about a 100% probability that either:
• The targets are easy.
• There is a whole lot of cheating and lying going on.
Probably a bit of both.
One major contributor to both is that the planning processes encourage both. It’s beyond the scope of this article to go into detail, but one example everyone understands is annual budgeting.
The reason why it ‘kind of works’ is that the targets are never based on great performance—they are what you can negotiate, and later they’re influenced by dumb luck.
And even if you’re not able to get ‘a good deal,’ and even if you’re unlucky, there is always the option to ‘cheat’ to hit the targets—at the expense of the long term, which isn’t even in scope.
So, in practice, even when the P70 targets would change to P40 targets under the assumption that strategic choices would be adhered to and no long-term performance would be sacrificed, no such constraints apply to annual budgeting.
Sell more economy products instead of premium ones, cut long-term investments and marketing costs, and voilà—the same target that under the old assumptions was a P40 suddenly becomes a P95.
Nobody just notices, because ‘probabilities are theoretical.’
The Terminology Mess
To comprehensively cover the terminology would require a book on its own—or a series of books, if we’re being honest.
That’s outside the scope of this article, but here are some examples that illustrate the issues:
- Budgets as Targets, Forecasts and Plans
- Revenue and Profit ‘Budgets’
- Plans as Targets
- Forecasts as Targets
- Forecasts as the ‘Most Likely Outcome’
- Rolling Forecasting
Budgets as Targets, Forecasts and Plans
The Beyond Budgeting movement got the problem right. Traditional (annual) budgets have three conflicting definitions: ‘what we want to happen’ (target), ‘what we think will happen’ (forecast), and ‘what we’ll do’ (plan-more specifically, decisions about financial resource allocations).
In other words, budgets are what we want to achieve, by allocating resources as budgeted, since the budget is what we think will happen.
That’s circular reasoning.
If you were to put that in Excel, it would give a circular reference error. The outcome of the formula is nonsensical, as it uses itself to calculate itself.
You cannot effectively create ‘what you want to happen’ unless you understand ‘what you think will happen’-and if you constrain ‘what you’ll do’ by setting resource allocations in stone.
However, while Beyond Budgeting was spot-on about the problem, its solution-dividing budgeting into three separate processes-is at best ineffective, and at worst, actively harmful.
Their reasoning is that separation will solve the gaming issue: people trying to massage the numbers in their favor.
Here’s the problem with that. As discussed earlier in this article-and as the Beyond Budgeting movement also recognizes-the three concepts are all dependent on the same assumptions, like what the market does.
That means if you separate them without integrating the assumptions, you get three nonsensical processes (which, of course, is exactly what happens when companies run annual budgeting and later try to disentangle the pieces).
For example:
- A target of $100 million—desirable and realistic but challenging—based on assumptions of high market growth and the ability to raise prices.
- A forecast of $90 million, based on assumptions of flat market growth and no price increases.
- A plan to decrease prices, based on a $90 million forecast and assumptions of declining market growth.
At a minimum, you need to connect the assumptions-which include both external (like the market) and internal (like decisions) factors.
However, as long as that’s the case, people will still know that the forecast-no matter where or how it’s prepared-will be used in decision-making. So, the original incentive to game it hasn’t changed.
Beyond Budgeting suggests that this can be ‘solved’ by separating the people who do the forecasting from those who make business decisions. But that results in disconnecting forecast accountability from business accountability.
In other words, it disconnects forecasts from real decisions and what happens in the real world, which often makes forecasts irrelevant-or even counterproductive.
Aligning forecasts and business accountabilities doesn’t mean that executives who decide prices, for example, should enter forecast numbers into the system.
It means that a person accountable for delivering a number-like expected sales-must own and approve that number.
That requires understanding the key business drivers behind it, and therefore being actively involved in its creation.
Not by entering numbers or running statistical algorithms, but by understanding external drivers and how their own future actions are included.
In the end, there’s no way to keep forecasts connected to reality if you disconnect the people accountable for the decisions that influence those forecasts from accountability for the forecast.
That, by the way, is different from just getting the same level of accuracy. A finance team may reach the same ‘accuracy’ with a highly tuned Excel calculation-and that would be fine if the purpose were just to maximize accuracy.
But if-and when-the purpose is to improve decision-making, it’s a moot point.
That’s how you get into conversations like:
- “Whose number is that?”
- “That’s not my number.”
- “What’s in that number?”
And in the end, you still cannot hold anyone accountable for a ‘highly accurate’ forecast.
So, separating target setting, forecasting, and decision-making (planning) cannot solve the gaming issue-without opening a whole other can of worms, with significant negative consequences.
And that’s not even accounting for the fact that the same dynamics apply to all planning-not just resource planning, which is the core of annual budgeting.
Companies run (or should run) multiple planning processes for different types of decisions, and applying the ‘separation solution’ would mean tripling the number of processes.
So, instead of one process, you’d need to run three. That would not only be inefficient, but impractical-since keeping them connected would require sequencing them.
While that might be possible for some decisions, like resource planning, you’d again run into problems trying to align vertically across strategic, tactical, and operational levels.
Even with just those three levels, you’d be running nine processes-and that doesn’t even account for business units, geographies, or functions.
What a mess.
Instead of separating targets, forecasts, and plans into three processes, it’s more effective to keep them in one process-and deal with the gaming issue differently.
The simple solution is to hold people accountable not only for targets, but also for forecasts.
Many get confused here.
Common misconceptions include:
- “That will change the targets.”
- “That will lead to more KPIs.”
- “That will inflate the forecasts.”
None of that is true, nor does much of it even make sense. Accountability for a forecast doesn’t mean accountability for delivering the forecasted number.
Yes, that sounds confusing. So, what does it mean?
It means accountability that the forecast is not bullshit.
That’s different from requiring the forecast to come true precisely. It means the forecast does not contain systematic error. That it’s not consistently exaggerating or sandbagging what actually happens.
BIAS is a simple tool to accomplish this. It allows forecasts to be missed all the time. The forecast doesn’t have to come true even once-or even be close to the actual number-for the BIAS to be good.
So, accountability for forecasts doesn’t replace targets or create parallel targets.
Here’s the distinction:
- Accountability for the target = accountability for what we want to happen.
- Accountability for the forecast = accountability for the business logic behind ‘what we’ll do’
The forecast is the basis for decisions (i.e., the plan), and as long as the forecast is unbiased (free of systematic error), the decisions-‘what we’ll do’-are based on the best possible understanding of the world.
That’s also what enables the highest odds of achieving the target.
And accountability for targets ensures accountability for the decisions.
Revenue and Profit ‘Budgets’
The concept of budgeting revenue is an oxymoron—which also makes profit budgets oxymorons.
You can make predictions about profits or revenue. You can take actions to influence profits or revenue. But you cannot make allocations, reservations, or commitments for profit or revenue.
You don’t control revenues or profits in the same way you control costs.
If by revenue budget you mean target, use the word target.
Or at least use a specific word for ‘what you want to happen.’
If you mean a prediction or expectation of revenue, use the word forecast.
Or at least a specific word for ‘what you think will happen’—which is different from ‘what you want to happen,’ because those are two separate things.
Those are the only two types of words you need for revenues.
There is no real-life action for allocating revenue like there is for allocating costs.
So, while it may be useful to use a specific term for the action of allocating costs, there’s no use for a specific term for ‘allocating revenue.’
Plans as Targets
The most common use case for this confusion is again annual budgeting: it produces the plan, which becomes the target. But if that’s the case, you’re left with a critical problem—what do you then call ‘what you’ll do?’
Forecast?
And if that’s the case, what do you call ‘what you think will happen?’
The basic issue is the same as with all other misconceptions: once you merge two fundamental components, you’re in trouble. You either omit one of the core parts entirely, or you unnecessarily constrain yourself.
When viewed through the lens of the fundamental components, the illogicality becomes obvious:
‘What you’ll do’ = ‘What you want to happen’
You raise prices 5% because you want prices to be 5% higher?
You launch new products because you want to launch new products?
Regardless of their impact on revenue, profits, costs, or performance?
Who thinks that way?
No one.
Okay—of course no one actually thinks that way when they say plan = target. But here’s the problem: that faulty logic sneaks into how companies operate.
The practical issue is this: the plans become ends in themselves—even when they conflict with business performance.
Budgets are classic examples, but the same flawed idea underpins much of what’s called ‘strategic planning.’
Executives love to make 5-year strategic plans with revenue and profit projections.
Those projections then become synonymous with the plan, which becomes the strategy that will be implemented.
This is pure insanity.
Those projections are predictions of the future—i.e., what we think will happen, i.e., forecasts. So it’s not even that the plan is the target—it’s that the forecast attached to the plan becomes the target.
But as we now know, forecasts are always subject to real-world conditions (like what the market does) and the company’s own actions (‘what we’ll do,’ the plan). What you really have is just another circular reference—where companies end up chasing their own tails.
Forecasts as Targets
This is not so much about confusing forecasts with targets as it is about the absence of a holistic and effective system for business steering—and forecasts (and often targets as well) being disconnected from reality.
The problem is that in companies that run annual budgeting—i.e., they set static point-estimate targets—people are held accountable for the targets, whereas no one (at least no one who matters) is held accountable for the forecast.
This leads to the belief that forecasts should match the targets—since otherwise, the targets won’t be reached. And after all, what’s the downside, since nobody is really accountable for the forecast anyway?
While there is some logic to that, it neglects the basic difference between ‘what we want to happen’ and ‘what we think will happen’: if there is no difference, no decisions are needed.
It also neglects another basic fact: both are dependent on the real-world context.
Even in companies where the targets are precise, deterministic numerical points set in stone—unless the world around stood still, and unless the organization made no new decisions—the forecast would constantly change.
So, the real problem is that forecasts are not grounded in what happens in the real world: the company’s plans (decisions that are probabilistic) and external assumptions (like what the market does).
If they were, it would immediately become clear that the forecast ≠ target.
Forecasts as the ‘Most Likely Outcome’
This is commonly used by finance experts, and while it may serve as a definition for a particular type of forecast, it is not a particularly useful definition for a forecast in general.
A ‘most likely’ forecast typically means a point estimate believed to be the most probable outcome at a given time. However, that’s a bit of an oxymoron—because the probability of any static, precise point estimate actually occurring is always 0%.
So not only is it the ‘most likely,’ it is also not likely at all to occur. Of course, the practical interpretation is different: it’s ‘the best guess’ of the general ballpark where the future will land.
Another way to put it: it is the point around which there’s an equal probability of exceeding or falling short—a 50/50 forecast.
In reality, though, that’s often not the case either.
It’s more likely the ‘most likely hoped-for outcome.’ That means there’s a somewhat realistic possibility of hitting that number—but unless everything goes well, it’s unlikely to be reached.
And because there’s usually no clarity around the probabilities of exceeding or falling short, ‘most likely forecasts’ are often not particularly effective for business steering—even though that’s what most companies use, whether or not they use the term.
A more useful approach would be to include probabilities of exceeding or falling short.
For example, a P70 or P90 forecast: there is a 70% or 90% probability that sales will exceed $1 billion. That’s just as simple to understand as ‘most likely’ (which, if true, means—in other words—a P50 forecast) but infinitely more useful for business steering and performance management.
Or a variation using the original language:
“There is over 50% probability that sales will be between $990 million and $1.01 billion.”
That’s significantly more precise and actionable than saying:
“There is a 50% probability that sales will be between $900 million and $1.1 billion,”
—or worse—
“Between $500 million and $1.5 billion.”
In all those cases, the so-called ‘most likely’ outcome might still be $1 billion—but the decisions the company should make could differ dramatically.
Those who get caught up in whether the distribution is symmetrical are missing the point. But for the sake of completeness, the same ‘most likely’ outcome of $1 billion could apply to these cases as well:
50% probability that sales will be between $800 million and $1 billion,
or between $1 billion and $1.4 billion.
Rolling Forecasting
One of the worst terms in business planning—ever. For several reasons.
First of all, try Roger Martin’s strategy test: “Is the opposite stupid on its face?”
Not only does it work for testing strategic choices, but it also works equally well for business terminology in general.
So—rolling forecasting instead of static forecasting? Rolling instead of setting the forecast in stone? That, if anything, is indeed stupid on its face, which makes “Rolling Forecasting” a nonsensical term.
But even if we put the obvious absurdity of the term aside, it’s illogical when viewed through the framework of ‘what we want to happen,’ ‘what we think will happen,’ and ‘what we’ll do.’
It’s not just about ‘what we think will happen’—which is what forecast means if you look it up in a dictionary or ask a normal person. Rolling forecasting often includes not only that, but also parts of ‘what we want to happen’ and ‘what we’ll do.’
Finance experts frequently present rolling forecasting as an alternative to annual budgeting. Sounds good, right?
Wrong.
Even if it were used as an alternative to planning—that is, as a substitute for the cohesive whole of all three—it would require redefining forecasting in a way it’s never used elsewhere (outside of this narrow finance context).
And if you do that, you can no longer have a sensible definition for planning.
If this were just about semantics, it might not matter. But since planning is already an established—if often misunderstood—term, replacing it with rolling forecasting only creates further confusion.
And even if that weren’t a problem, the actual characteristics of rolling forecasting reveal it’s not a true substitute for planning. One major reason: it’s typically a finance-led numbers exercise that’s disconnected from how real decisions are made.
So if you look at what rolling forecasting actually is—who does what, and how—it turns out the name is appropriate: it’s less about the business deciding ‘what we’ll do,’ and more about finance fine-tuning their Excel models at a faster pace.
All of that leaves us with a fundamental confusion and contradiction:
- The name refers to ‘what we think will happen’ (forecasting).
- It’s presented as covering ‘what we want to happen,’ ‘what we think will happen,’ and ‘what we’ll do’ (planning).
- In reality, it focuses only on ‘what we think will happen’ (forecasting).
That could suggest the name is actually accurate—but in that case, there’s a mismatch with how it’s presented: as some kind of dynamic business steering mechanism, an alternative to annual budgeting.
No matter how you twist it, it’s impossible to make a coherent or effective system out of the concept itself—let alone use it to steer a business.
Conclusion
Many companies don’t have logical and cohesive terminology for fundamental components that influence their futures—that is, components required to steer the business.
The definitions of individual terms are often illogical—or even counterproductive—and the sum of them doesn’t add up to a sensible whole: it results in circular references, contradictions, oxymorons, or missing parts.
That means the systems behind their business steering mechanisms (regardless of what they are called) are ineffective by design and cannot ever result in effective decision-making or maximize business performance over time (regardless of what is meant by that).
Even when the fundamental components are logically defined and form a cohesive whole, many companies make the mistake of working on them in silos instead of together.
Transforming a company’s business steering mechanisms into effective systems is a massive undertaking that has many moving parts, involves high risk, and requires significant time and effort.
A good practical starting point for many organizations would be to clarify the terminology around the basic components by which the future is influenced.
That alone won’t change the way companies are steered, but it will enable rational discussion about how the current system works and how a more effective system could work.
Here is one example of such a cohesive whole, based on the most commonly used terms:
• What we want to happen (the desired future) = Target
↳ For example, a profit of $100 million (not that static numbers make good targets)
Or, better yet,
↳ A profit of $100 million is a P50 target assuming strategic choices are adhered to.
• What we think will happen (a prediction about the future) = Forecast
↳ For example, a profit of $90 million
Or, better yet,
↳ A profit of $90 million is a P70 forecast assuming 5% market growth and a $10 million marketing budget.
• What we’ll do (an interdependent set of decisions) = Plan
↳ For example, decisions to allocate $10 million to marketing and to raise prices by 5%.
Or, better yet,
↳ The plan to invest $10 million in marketing and raise prices by 10%, assuming 5% market growth and competitors following price increases to reach the profit target of $100 million, is a P40 plan.
Budget = Allocation of resources (specific type of plan)
↳ For example, a marketing budget of $10 million = an allocation to spend $10 million on marketing
• Cohesive whole (‘what we’ll do’ to make ‘what we want to happen,’ based on ‘what we think will happen’) = Planning
For example, planning is deciding to allocate $10 million to marketing and to raise prices by 10% in order to improve the probability of reaching EBIT of $100 million by the end of the year from 40% to 50%, based on the assumptions that the market will grow 5% and competitors will follow price increases.
What the exact terms are is (almost) irrelevant, as long as they cover all fundamental components involved in influencing the future, form a cohesive whole, and are understood the same way by everyone in the company.
The opposite is equally true: regardless of the terms used, unless they form a cohesive whole and are understood the same way by everyone, effective planning—i.e., effective decision-making—i.e., effective business steering—is impossible.
Practical insights
- Most companies use illogical terms in their business steering.
- Fundamental components for influencing the future must form a cohesive whole
- A target = ‘what we want to happen’
- A forecast = ‘what we think will happen’
- A plan = ‘what we’ll do’
- A budget = resource plan, i.e., what we’ll do to allocate resources
- A budget ≠ a target or a forecast
- A forecast ≠ a plan
- Planning = (‘what we’ll do’ to make ‘what we want to happen,’ based on ‘what we think will happen’)
