There is a step that gets skipped in nearly every planning process I encounter, in organizations of every size and in individual efforts of every kind. It is not a difficult step. It requires no expertise and usually very little time. It is simply this: before you begin, write down where you are starting from.
The reason it gets skipped is not oversight. It gets skipped because the honest number is almost always worse than the number people have been carrying in their heads, and because writing it down converts a vague unease into a specific fact that can be compared against later. That is uncomfortable, and the discomfort is doing real work in preventing the measurement.
It is also the single cheapest source of value in any engagement, and skipping it produces a specific downstream failure that is nearly impossible to diagnose once it has occurred.
The failure it produces
Without a baseline, a plan cannot be evaluated. That sounds abstract until you consider the practical consequence: when the plan underperforms, you have no way to distinguish a defective plan from poor adherence.
Those two situations require opposite responses. A defective plan needs to be redesigned; executing it harder produces the same result plus discouragement. Poor adherence to a sound plan needs accountability structure; redesigning the plan rewards the avoidance. Choose wrong and you will make the situation worse in a way that looks like effort.
In the absence of data, the default attribution is almost always to adherence. The person or team is assumed to have underperformed, because that explanation is available immediately and requires nothing to be reconsidered. So plans that were defective from the outset get executed harder, repeatedly, while the people executing them accumulate evidence that they are the problem.
The organizational version: the number is wrong
In organizations, the problem usually is not that no number exists. It is that a number exists, is quoted confidently, and does not measure what everyone believes it measures.
A regional manufacturer reported on-time delivery at approximately 82 percent. Every executive could quote it. It appeared in board materials. Attempting to verify it against source systems revealed that on-time delivery was calculated against the most recently revised promise date rather than the original commitment to the customer. Every schedule slip had been silently absorbed into the definition. Against original commitments, the figure was 61 percent.
Nobody falsified anything. A reasonable definitional choice was made years earlier by someone solving a reporting problem, propagated into every downstream report, and never re-examined. The company had been managing a problem one third the size of the one it had.
The improvement plan under discussion would have targeted a nine-point gain against a twenty-one-point gap. It would have been executed diligently, produced real improvement, missed the target substantially, and the failure would have been attributed to operational discipline. The definitional defect would have survived the entire engagement, because nothing in a conventional planning process ever goes back and questions the metric.
This pattern is common enough that I now treat baseline verification as an independent objective rather than as preparation. In organizational engagements, discovering that a core operating metric contains a definitional defect is frequently the most valuable single finding of the entire engagement — and it is available in week two, for the cost of asking how the number is calculated and then checking.
The individual version: the number is avoided
For individuals the failure mode is different. The number usually does not exist at all, and its absence is being actively maintained.
A person planning to rebuild their physical condition typically does not want a baseline fitness test. Someone intending to get their finances in order does not want a complete accounting of current debt. A professional planning to improve their writing does not want to reread what they produced last year. In each case the avoidance is rational in the short term: the honest measurement is worse than the working estimate, and knowing it precisely feels like it would make things worse.
What actually happens is the reverse, for a reason that is not obvious until you have experienced it. An unmeasured deficiency is unbounded. It occupies as much psychological space as your worst-case imagination allows, and it is available for rumination indefinitely. A measured deficiency is finite. It is a specific number, it has a specific distance to a specific target, and that distance can be divided into blocks.
My own baseline, when I finally took it after roughly a decade of avoidance, was worse than I expected and considerably better than my fear. A ninety-second plank. Eighteen push-ups in a minute. A two-kilometer row at nine minutes forty. For a former athlete those numbers were unpleasant to write down. They were also the first time in ten years that I knew where I actually stood, and the first time any subsequent measurement could mean anything.
Six weeks later, a second measurement was possible. That is the entire point. Without the first number, the second number is just a number. With it, there is a direction and a rate, and both are facts rather than impressions.
What a real baseline requires
Three properties, and all three are routinely violated in ways that make the measurement useless later.
First, it must be recorded rather than known. A number you remember is a number you will unconsciously revise. Six weeks from now, if progress has been modest, your memory of the starting point will drift downward to make the change look larger. This is not dishonesty; it is ordinary memory reconstruction, and it is precisely why the record has to exist outside your head, with a date on it.
Second, the measurement method must be documented alongside the number. A figure without its method is not reproducible, and a baseline you cannot reproduce is not a baseline. If on-time delivery is 61 percent, the record must state that it is measured against original customer commitment, excluding customer-initiated changes, using the ship date from the source system. Otherwise your next measurement will use a slightly different method and the comparison will be meaningless.
Third, it must be taken from source rather than from a report. Management reports are aggregations, and aggregations contain definitional choices made by people who are not in the room. If the baseline matters, go to where the data originates.
The gate this creates
There is a consequence worth stating plainly, because it is where clients most often push back.
If the objective is measurable, execution does not begin until the baseline is recorded. Not because of methodological purity, but because a plan launched without a baseline is a plan that cannot be corrected, and an uncorrectable plan will consume the engagement's full cost while producing no reliable information about whether it worked.
Clients experience this as a delay, and it usually is one — a week, sometimes two, occasionally longer when an organization discovers its systems cannot produce the number it manages by. That discovery is not a delay. That discovery is the finding, and it is worth more than the plan that was waiting on it.
There is a related gate worth mentioning, because it comes up in personal engagements. Where verification requires an outside authority — medical clearance, legal review, financial audit — the plan waits at the gate rather than routing around it. A coach is not a physician. A framework that pretends otherwise is a liability to the client, and the temptation to proceed is strongest exactly when the client is most eager.
The objection I hear most
The most common objection to baseline discipline is that some things worth pursuing genuinely cannot be measured. Culture, trust, judgment, the quality of a relationship with a difficult board. This objection is correct, and it deserves a better answer than insisting everything is quantifiable.
Here is the answer. When an objective resists direct measurement, you do not abandon the baseline. You record a description precise enough to be compared against later, and you record it in writing before you begin.
For an unmeasurable objective, write down what you currently observe: what specifically happens in meetings, what people say and do not say, what decisions get made and how. Three or four paragraphs of concrete observation. Date it. Do not revisit it during the block.
Six weeks later, write the same description again without rereading the first one. Then compare them. You will find differences you would never have noticed otherwise, because the drift is gradual and memory smooths it. This is not measurement, and it does not pretend to be. It is a fixed reference point, and that is the actual function a baseline serves.
An executive I worked with used this for an objective about becoming less of a bottleneck for her leadership team. Her opening description mentioned, almost in passing, that she was copied on essentially every external email. Her six-week description did not mention email at all — the practice had stopped and she had not registered it as a change. Without the first document, that improvement would have been invisible to her, and she would have judged the block a partial failure on the basis of the things she happened to remember.
Why the number gets avoided
It is worth being direct about the mechanism, because understanding it makes the avoidance easier to overcome.
An unmeasured problem lets you hold two beliefs at once: that the situation is bad, and that it might not be as bad as you fear. That second belief has real psychological value, and measuring destroys it. So avoidance is not irrational — it is protecting something.
What it is protecting, though, is comfort rather than capability. The cost is that you cannot plan. You cannot size the effort, you cannot sequence the work, and you cannot tell whether anything you do afterward helped. You have traded the ability to solve the problem for the ability to remain uncertain about its size.
There is also a second-order cost that is easy to miss. People who avoid baselines tend to also avoid the reviews that would reveal them, which means they lose the correction cycle entirely. The avoidance does not stay contained to one measurement. It propagates into the whole method, and what remains is effort without feedback.
Where to start
Take the objective you care most about right now. Write down its current value, the date, and the method you used to obtain it. If you cannot obtain a value, that is your first finding and your first assignment.
If you lead an organization, take the three metrics your leadership team manages by and ask independently how each is calculated — what is included, what is excluded, and which date is used. Compare the answers. In my experience it is more likely than not that at least one contains a definitional problem that has been quietly shaping decisions for years.
The number will probably be worse than you thought. That is not a reason to avoid it. It is the reason it is worth having, because a problem with a known size can be planned against, and a problem of unknown size can only be worried about.