Skip to content
Reason Talent, home

Make it possible to perform

Performance systems built around one question. What does this person need in order to do their best work, and what is currently in the way of it.

Start a conversation

Twenty percent of goals were in the system

McKesson · September 14, 2026 · 2 min read

  • Goal entry in the HRIS sat at twenty percent, so the system held no reliable record of what anyone was working toward.
  • The rebuild was end to end: what to add, what to remove, and the manager journey mapped with HRIS and Employee Experience.
  • Goal entry reached ninety-seven percent and calibration errors fell by three quarters.
  • Ninety-nine percent of teams said the conversation was worth having, which matters more than the ninety-eight percent who completed it.

Performance management across the enterprise needed a redesign. Goal entry in the HRIS sat at twenty percent, which meant the system held no reliable record of what anyone was working toward. Every stakeholder group had a different view of what should change, particularly around calibration, performance labels, and how managers entered notes.

Download the PDFRead it on the site

If the managers just need training, call someone else

Most performance systems were built to settle pay. They measure what pay needs, and managers learn to produce exactly that and skip the rest.

Which is why the usual fixes do not move anything. Train the managers, adjust the rating math, train them again, and goal entry sits where it was. If that sounds familiar, the rest of this page is the work. Three things have to be true before any of it holds.

Say what good is

Goals that state the target rather than describe the job, and rating levels a manager would actually use out loud. This is where it usually breaks, and nobody reports it, because nobody raises their hand to say the goals were never written down.

See it often enough to act

Feedback frequent enough to change something while it can still change. Most systems ask for it once a year because they were built to justify one pay decision a year, whatever the policy says they are for.

Make it count

Pay, progression and accountability that track what was measured. When the first two are weak this one lands as noise: the raise arrives and means nothing, because nobody could say precisely what it was for.

The three depend on each other. A system missing one of them still runs, so nobody notices it stopped working.

The pain is rarely where the break is

The symptom shows up a long way from the cause, so the purchase usually lands in the wrong place.

A vague standard produces no complaint of its own. A feedback problem is felt as a manager problem, so the first purchase is manager training. A consequence that lands flat is felt as a compensation problem, so the second purchase is re-cutting the rating math. Both treat the place the pain surfaced. We made both of those purchases at a Fortune 10 company over several years, and neither of them moved anything.

What had actually gone wrong was upstream and structural. Four in five people had no goals recorded, and the system had been built to justify one compensation decision a year, so it asked for feedback once a year. You cannot train somebody into a behavior the architecture never asks for.

The system is the floor. Manager skill is the ceiling. Most organizations in pain spend on the ceiling.

Training around the system

If you have to train people to work around your performance system, the system is the thing that needs the work.

What we do

Six pieces. Which of them an organization actually needs is what the diagnosis is for.

Rating architecture

The scale, the labels and the definitions underneath them. A label is a word a manager has to say to somebody's face, and if they will not say it they will rate around it. Distribution guidance and calibration that holds the shape.

Where this has been done

Enterprise performance management rebuilt end to end at a Fortune 10 company, over five years, as the person accountable for it rather than as the consultant advising on it.

5x
Goal entry in the HRIS, from 20 percent to 97 percent
75%
Fewer calibration errors
4,000+
People leaders trained on the rebuilt system
99%
Of teams said the conversation was worth having

McKesson, 2021 to 2026. The work took the company’s HR Impact Award.

And a Fortune 500 consumer goods company, where the rating labels were the problem. The word for the middle of the distribution was one no manager wanted to say out loud, so nobody used it and every rating drifted upward. That engagement carries no published figures.

Twenty percent of goals were in the system

McKesson · September 14, 2026 · 2 min read

  • Goal entry in the HRIS sat at twenty percent, so the system held no reliable record of what anyone was working toward.
  • The rebuild was end to end: what to add, what to remove, and the manager journey mapped with HRIS and Employee Experience.
  • Goal entry reached ninety-seven percent and calibration errors fell by three quarters.
  • Ninety-nine percent of teams said the conversation was worth having, which matters more than the ninety-eight percent who completed it.

Performance management across the enterprise needed a redesign. Goal entry in the HRIS sat at twenty percent, which meant the system held no reliable record of what anyone was working toward. Every stakeholder group had a different view of what should change, particularly around calibration, performance labels, and how managers entered notes.

Download the PDFRead it on the site

Start with the diagnosis

Tell us where it hurts and we will tell you where we think it broke, including when the answer is that your system is fine and something else is wrong.

Start a conversation