Writing

Why Most Companies Set the Wrong Goals

Goals get copied rather than derived, then measured as activity rather than result. Both mistakes are cheap to make and expensive to live with, and together they explain most of what looks like a talent problem.

I once watched an HR team set a goal to reduce average time to hire.

They set it because a venture firm had set that goal for another company, and it looked like the kind of thing a serious HR function measures. Nobody in the room asked the obvious question first: does hiring faster actually produce a more talented team, better output, and more profit?

Generally it does not. Speed in hiring mostly produces misalignment, and misalignment is far more expensive than an empty seat. A better approach is slower and almost never chosen: forecast your needs 6-18 months in advance, proactively build a pipeline of talented people, run small paid projects with outside contractors so both sides get a real sample of working together, then hire the best fit when the seat opens.

That team spent a year optimizing a number that, if they hit it, would make the company worse.

This is not an HR problem. It happens in every function, and it happens because goal setting gets treated as an administrative step to clear rather than the decision that determines everything downstream. A planning session lands on the calendar, a template gets filled in, three or four numbers get written down, and the deck goes into a folder. It takes an afternoon and it feels productive because something was produced.

What is lost is the nuance. Companies and the people in them are often in such a rush to be busy that they blow through goal setting without considering whether the goals are doing more harm than good.

An engineer cannot work this way, because the consequences of an unexamined assumption are visible and expensive. A structure built on a number nobody derived does not produce a disappointing quarter. It falls down.

I trained as an engineer before I ran marketing teams, and the contrast is what makes this hard to unsee. In engineering, deriving the number is the job. Most companies treat it as a step to rush through on the way to the work, which is the habit I have found transfers most usefully.

In practice: for every goal, ask why until the answer stops being circular. If the honest reason is that another company measures it, or that it is easy to count, you have found a number worth removing rather than hitting.

Input goals and output goals

The second mistake compounds the first, and it is the more common of the two.

When you understand the fundamentals of a function, you can set goals that fit the work and the person doing it. When you do not, you fall back on measuring activity, because activity is the only thing visible from outside.

Input goals measure effort. Posts published, calls made, features shipped, campaigns launched.

Output goals measure result. Qualified pipeline, activation rate, revenue retained.

Both get called goals. They behave completely differently.

The three problems with measuring activity

One: you cannot find the failure.

When a company misses and every goal underneath it was an activity number, there is nowhere to look. Every activity number can be hit while the outcome misses. Content shipped on schedule, calls made to target, features delivered. Everyone did their job and the company still lost, and now the diagnosis is a matter of opinion rather than evidence.

Two: it teaches people to think in inputs, which is worse than it sounds.

If shipping the feature is the goal, then shipping the feature completes the job. Was it any good? Did anyone adopt it? Not the question that was asked, so not the question anyone answers.

I shipped it. My job is done. Whether it was world class is somebody else's problem.

That is a precedent, and it spreads. Once a company establishes that hitting your number and doing excellent work are separate things, people rationally optimize for the number, because that is what gets reviewed. Quality becomes a residual, something that happens if there happens to be time. Companies die of this and the postmortem usually blames execution.

Three: it requires micromanagement by construction, and micromanagement is expensive in a way that never shows up as a line item.

If you are measuring activity, you have to watch the activity. So you get standups to confirm work is happening, check-ins to confirm it is still happening, and status reports about work that could have been done in the time it took to report it.

Then the arithmetic gets ugly. A manager watching activity can genuinely handle three to five people. Beyond that they are not managing, they are sampling. So the company hires more managers, and the cost of the management layer rises faster than the cost of the work.

Managing to output, I have directed a dozen full-time staff comfortably, alongside contractors and agencies at the same time. Not because I am unusually capable at supervision, but because supervision is not the job. I am watching outcomes, which is a far smaller number of things to watch.

Hire the right people, understand where each of them is strongest, set goals that fit, supply the resources those goals require, and give them an environment autonomous enough for their efficiency to be maximized and their creativity to show up. The rest largely handles itself.

How I actually set them

I work alongside the team to set output goals for the quarter, for the department and for each individual. Aligned to their strengths, and agreed as reasonable by them at the start rather than handed down.

That last part matters more than it sounds. A goal someone accepted is a commitment. A goal someone received is an assignment they will meet the letter of.

Done properly, this removes the need for micromanagement entirely. If you hit the output, where and when and how you did it stops mattering. And if you missed it, that is the same conversation regardless of how much activity happened along the way.

Far too many companies spend enormous energy forcing people into rigid structures built around input goals. Not because rigidity works, but because very few leaders actually know how to set output goals and hold people to them properly, and rigid structures at least look like management.

In practice: for each person, write the one result their quarter should produce. If you cannot state it without describing activity, you do not yet understand the role well enough to manage it.

What output goals unlock

The benefit runs well past cleaner reporting.

Autonomy, which produces better work. When the output is agreed, the person controls their own cadence. They work when they are sharp rather than when the calendar says to look available, and creative work in particular is far better under those conditions than under observation.

Recruiting through referral. People on a team that operates this way describe it to others in their network, most of whom have never experienced anything like it. Those people apply. I have seen this happen repeatedly, and it is the cheapest and highest-quality source of hiring there is.

Retention, and everything that follows from it. Longer average tenure, lower attrition, lower recruiting cost, and institutional knowledge that stays in the building instead of walking out every eighteen months and having to be rebuilt from scratch.

None of that appears in the quarter you make the change. All of it compounds.

Why this keeps happening

None of the above is difficult to understand. It persists because deriving a goal takes an afternoon of uncomfortable thinking and copying one takes ten minutes, and both produce a document that looks the same in a board deck.

The consequences also arrive late and land somewhere else. A goal set badly in January produces a miss in September, attributed to whoever was carrying the number by then. So the person who set it experiences no feedback, and the practice continues.

Sadly, most companies are too busy looking busy to stop and examine any of it. When you are traveling too fast, everything around you looks blurry. When you slow down once in a while, it becomes crystal clear.

Engineering bakes that pause into the process, because a bridge built on an unexamined number is not a disappointing quarter. Most other disciplines never had to learn it.

Frequently asked questions

Why do so many companies set the wrong goals?

Because goal setting is treated as an administrative step rather than the decision that determines everything downstream. Numbers get borrowed from another company or another industry without anyone asking whether hitting them would actually help here. An HR team reducing time to hire is the classic case: faster hiring generally produces misalignment, which costs far more than an unfilled seat.

What is the difference between input goals and output goals?

Input goals measure effort, such as posts published, calls made or features shipped. Output goals measure result, such as qualified pipeline, activation rate or revenue retained. Both get called goals and they behave completely differently, because every input number can be hit while the outcome misses.

Why are input goals a problem?

Three reasons. Failure becomes impossible to diagnose, because every activity number can be hit while the company loses. They teach people that completing the activity completes the job, which makes quality a residual rather than the point. And they require micromanagement by construction, because measuring activity means watching activity.

How does measuring output reduce management cost?

A manager watching activity can genuinely handle three to five people; beyond that they are sampling rather than managing. So companies add managers, and the management layer grows faster than the work. Managing to output, one person can run a dozen comfortably, because the number of things to watch is much smaller.

How do output goals reduce the need for micromanagement?

If someone hits the output, where and when and how they did it stops mattering. If they miss it, that is the same conversation regardless of how much activity occurred along the way. The supervision was never the mechanism, so removing it costs nothing.

How should output goals be set?

Work alongside the person rather than handing goals down. Align the goal to their actual strengths, and get agreement at the start that it is reasonable. A goal someone accepted is a commitment. A goal someone received is an assignment they will meet the letter of.

What are the second-order benefits of output-based goals?

Genuine autonomy, which produces better creative work because people control their own cadence. Recruiting through referral, because people describe how the team operates to their network. And longer tenure, which lowers attrition, lowers recruiting cost, and keeps institutional knowledge in the building.

Why do companies keep setting goals this way if it does not work?

Because deriving a goal takes an afternoon of uncomfortable thinking while copying one takes ten minutes, and both produce something that looks identical in a board deck. The consequences also arrive late and land on someone else, so the person who set the goal never experiences the feedback.