Development bought before the role has a standard is a coaching program or manager curriculum purchased against last year's job description, competency row, or "meets expectations" language that still describes last year's job. The live work already changed. The purchase looks like a people-development decision. The standard never moved, so coaching has nothing to coach against.
This is the leftover-standard problem. How much a leadership development program costs already covered the buy. Leadership development versus training already covered the event that never becomes a week. This post stays on last year's JD, the competency row, or the "meets expectations" line that a coaching program still uses as the brief.
Why Did Companies Buy Development Before the Role Had a Standard?
Because last year's JD already looked finished.
Role standards exist for a reason. A Workday profile. A competency row. A "meets expectations" paragraph Comp and the HRBP can defend in calibration. Those files tell recruiting what to hire and the review cycle what to score. That model is a compensation artifact on purpose. The owner sits with Total Rewards or the HRBP. Tuesday does not have to rewrite them. "We already have a standard" is a true sentence when the job is last year's description.
The 2026 purchase was copied from that script. Same JD. Same competency row. Same "meets" language. Different job. The live work is no longer only the deliverable last year's file named. The week already moved. A person hired, reviewed, or coached against last year's description does not become able to do this year's work. What leadership development actually is already drew that line: a program is supposed to change behavior in the job. A leftover JD blurs the line. It looks like People work and runs like a compensation file with a new noun on the kickoff deck.
In 2026, Deloitte published AI and the Future of Job Architecture: From Static Frameworks to Living Systems by Greg Stoskopf, Melissa Murphree, Will Jirik, Steven McWilliams, Sheila C. Sever, Adam Ekl, Sarah Hildebrandt, and Sidney Ozcan (Deloitte Consulting LLP, copyright 2026). The authors wrote that traditional job architecture was designed for stability, not speed, and that static role definitions and slow update cycles are now misaligned with how work actually changes. They reported that organizations are planning work redesign at scale, yet fewer than 30 percent have fully linked skills to their job architecture, and that the gap between documented structures and work reality continues to widen. Greg Vert, Deloitte's HR Strategy and Technology Leader, said job descriptions stop being set-and-forget and become a moving target as more work becomes AI-enabled. Last year's JD is that gap, printed as a standard. Most companies still buy coaching against it.
The skip felt reasonable. The profile is in Workday. The competency model was approved. The "meets" language still prints. Those files were not wasted. The failure is using last year's description as the brief for a coaching program. How much leadership development actually returns already argued that the spend only pays when behavior in the job moves. A named JD answers the req. A named "meets" on the current job answers the live work. Most teams shipped the PO against the old file and called the quarter a development plan.
What Happens When Last Year's JD Counts as the Standard?
They fill a SOW. They miss the live job.
Finance signs. L&D picks a vendor. The kickoff deck says "manager coaching for FY26." Nobody writes what good looks like on the work people actually do this year. Nobody rewrites "meets expectations" to name the current deliverable. Nobody tells the coach which Tuesday conversation is the job. The coach inherits last year's JD. The person is still measured on the old sentence. The purchase looks complete because the file already existed.
The dangerous standard is not the missing one. It is last year's.
An empty JD at least admits nobody named the job. A complete, approved, last-cycle description is how the company files the role as handled. Coaching has a brief. Comp has a grade. Recruiting has a req. The person who does this year's work still gets rated on last year's sentence, and the coach still prepares against a file that no longer describes Tuesday.
A workshop, a cohort kickoff, a competency model reprint: those artifacts still look like activity. Someone can sit in the room. Someone can reprint the row. Last year's role standard is the one artifact Comp and People already own. It is the easiest one to file as "we already know what good looks like" because the JD is approved and someone can paste it into the SOW. That is why it is the most dangerous.
| On last year's standard | What a role standard for coaching needs |
|---|---|
| JD / Workday profile from last cycle | The live jobs people actually do this year |
| Competency row / "meets expectations" | Observable "meets" on the current work, not the old deliverable |
| Approved grade / last-cycle language | A behavior a coach can see this week |
| "We already have a standard" | Proof the standard moved with the work |
The left column is what most 2026 development POs still buy against. The right column is the minimum for a standard you can coach. If you cannot fill the right column, you do not have a role standard. You have last year's file with a new noun on the kickoff deck.
That is why the business case for coaching cannot be a seat count pasted into the people pack. A purchase answers "did we buy development." A standard answers "what good looks like on the job as it is now." A system can store a JD. It cannot tell you whether the person who "meets" still does last year's work, or whether the coach was asked to grow a behavior the file never named.
Stale standards also teach the company that the live job is optional. If last year's JD is on the SOW, nobody has to write what "meets" looks like this year. The team copies that the way they copy a boss who still runs the week the old way. The cost of bad managers already put a number on what that copy costs. Buying development against a leftover standard is how the manager layer gets skipped while still looking funded.
What Should Happen Before Development Gets Bought?
Three things. About the live job. Written before the PO, not after the kickoff.
A usable standard needs the same specificity a management development program is supposed to start with: a named gap in the current job, not a leftover Workday paragraph. The JD can transfer last year's language. It cannot assign this year's Tuesday.
Rewrite what "meets" looks like on the work that already changed. Not "refresh the competency model." The job they already own: the conversation they run, the deliverable they ship, the decision they make, and what a miss looks like this year. If they cannot point at that sentence before Finance signs, they bought a program. They did not buy a standard. What management coaching is is the mechanic that needs that sentence. This JD is the artifact that pretends last year's "meets" already covers it.
Buy coaching against that sentence, not against last year's file. The useful brief is a behavior a coach can see. A conversation that went sideways. A deliverable that still starts the old way. A week that made the leftover standard feel safer, and whether they stayed with the new work anyway. If the brief is "coach to the JD" and the JD still describes last year's job, you do not have a program. You have a PO. What actually works in manager coaching already said the same thing: a workshop transfers language. A leftover file transfers a grade. Neither tests the week.
Put the new standard on the week, not only in the job-architecture folder. Comp can keep a grade. The manager has to ask, after the PO, whether the live job looks like the rewritten "meets." That is not a second job analysis. It is the weekly conversation last year's JD was standing in for. Leadership skill gaps persist when the standard never names the current work. Training transfers language in a room. A JD transfers a file. Neither tests the week unless someone who assigns the work names what good looks like now.
Stoskopf and colleagues wrote that the gap between documented structures and work reality keeps widening, and Franz Gilbert at Deloitte said the velocity of job change forces more frequent updates to job descriptions. The operational burden is the constraint. Last year's file is the left column. A "meets" someone can see this week is the right column. The PO cannot produce it.
Use coaching for managers only if the brief names the current job, not last year's JD. Why new-manager promotions fail is what happens when the role grows and the standard stays put. Across our client base, competency scores improve 23 percent on average through coaching. That line only means something because the competency had a behavior a coach, and a manager, could see in a week. A development buy against last year's standard cannot move 23 percent. It can only produce a kickoff.
Why Is a Stale Role Standard a People Problem, Not a Job-Architecture Problem?
Because the JD was filed as a compensation file. The job is a people system.
Comp can keep a grade. Recruiting can keep a req. Total Rewards can keep last year's profile in Workday. They cannot sit with a manager on Tuesday and say what good looks like on the work people actually do now. They cannot tell People which live job is still scored against last year's sentence. Those are People jobs that now sit inside the manager role. When they go unfinished, the failure shows up as a healthy "coaching program launched" writeup nobody can translate into a changed week.
Calling it a job-architecture backlog is how last year's JD survives another quarter. Another Workday refresh, a cleaner competency row, a nicer "meets" paragraph will not add an owner if the proof is still the old file. The manager who still cannot coach the current work does not have a catalog problem. They have a standard that never required them to name the live job, and a People partner who was asked to treat last year's JD as the brief.
The Deloitte paper is blunt about where the mismatch lives: documented structures on one side, work reality on the other, and a refresh cycle that was built for stability. That is not a workshop skill. It is not a reprint skill. It is the unnamed part of the 2026 manager job: sit with a person while the work changes, and keep them on the current standard when the first run is clumsy. Last year's JD has no column for it.
That is why this belongs next to the development PO, not next to the compensation calendar. Coaching in that window is how a manager gets close enough to the work that the standard has somewhere to land, and how a person gets a chance to grow the behavior before HR reprints "we already have a JD." Sessions stay on Zoom. Slack and Teams carry the prep and the follow-through. The file is the artifact. The coaching is what makes it honest. A leftover Workday profile is not a brief. An approved last-year JD is not a standard.
FAQ
What is development bought before the role has a standard?
It is a coaching program or manager curriculum purchased against last year's job description, competency row, or "meets expectations" language that still describes last year's job. The live work already changed. The purchase looks like a people-development decision. The standard never moved, so coaching has nothing to coach against.
Why does buying coaching against last year's JD fail?
Because the file is a leftover, not a people-development program. It is last year's description, silent on the current work, and easy to paste into a SOW. The JD becomes the brief. The live job never has to be named.
Is an approved JD the same as a current role standard?
No. An approved JD answers whether Comp and recruiting have a file. A current standard is a "meets" sentence that names the live job as it is now, and a behavior a coach can see this week. A complete last-year profile with no rewritten "meets" is a catalog entry, not a brief.
How is buying against a leftover JD different from buying a workshop?
A workshop is an event that transfers language in a room. Leadership development versus training already covered why the room is not the program. This artifact is last year's JD, competency row, or "meets" language that the coaching buy still uses as the brief. If that file only produces a SOW, you have a leftover standard on the kickoff deck and no proof the live job was named.
What should happen before the development PO is signed?
Rewrite what "meets" looks like on the work that already changed. Buy coaching against that sentence, not against last year's JD. Put the new standard on the week, not only in the job-architecture folder. Keep the old file if Comp still needs a grade. Do not let it replace the brief the live job requires.
How does coaching sit next to a stale role standard?
Coaching is the calendar that keeps the week on the current job, and gives the manager time to grow the behavior after the PO is signed. It is not another job analysis. It is the conversation about the live job the old JD never named, where it got clumsy, and what Tuesday will show.
The Standard Is the Brief
If the quarter ran and development was bought against last year's JD, you do not have a mysterious culture problem. You have a people-development gap that the leftover standard made official. Stop counting the old file as a brief. Write what "meets" looks like on the live job. Then coach against that sentence.
Boon is one operating system for people development that lives in Slack, Teams, and MCP, and gets measured. SCALE, GROW, EXEC, and TOGETHER share a current standard. Last year's JD does not. Count the behavior. Then coach the week the file is supposed to describe.