---
title: "Both Axes of Your Nine-Box Come From the Same Manager: Walking Through a HiPo Nomination Round"
description: "A step-by-step walkthrough of a high-potential nomination round, and why a nine-box grid filled in by one manager measures performance twice rather than measuring potential."
canonical: https://www.assessall.com/blog/both-axes-of-your-nine-box-come-from-the-same-manager-walking-through-a-hipo
published: 2026-10-23
updated: 2026-10-23
author: "AssessAll Editorial"
source: AssessAll
---

# Both Axes of Your Nine-Box Come From the Same Manager: Walking Through a HiPo Nomination Round

A step-by-step walkthrough of a high-potential nomination round, and why a nine-box grid filled in by one manager measures performance twice rather than measuring potential.

A high-potential nomination is a prediction, not a reward: it claims a person will perform well in a role they have not yet held. Performance is evidence about the job someone holds now. Potential is a forecast about a different job. A nine-box grid displays both claims but supplies only the first.

What follows is an illustrative walkthrough — a composite of decisions that recur in talent reviews, not an account of any real organisation or client. The numbers in the scenario are invented for the example; every research figure is sourced.

## The round

A 900-person services firm runs an annual talent review. Line managers nominate for the high-potential pool; HR plots each nominee on a nine-box grid of performance against potential; a committee confirms the list. This year eight names arrive. Six sit in the top-right box. Two sit one box left.

The committee's first question is usually "are these the right eight?" That is the wrong first question. The right one is: **where did the potential rating come from?**

In this round, as in most, the answer is that the same line manager filled in both axes, in the same sitting, from the same impression of the same person. The grid looks like two independent judgements. It is one judgement, plotted twice.

This is not an argument against the grid. The grid is a reporting format — it has no measurement properties of its own. Whatever reliability and validity the review has are entirely the properties of the two inputs. If one input is a copy of the other, the grid is a performance rating drawn on a larger canvas.

## Step 1: Name the role the prediction is about

"High potential" is meaningless until you say potential *for what*. In Silzer and Church's survey of 20 large corporations, organisations split on exactly this point: 35% defined potential by role, 25% by level, 25% by breadth of contribution and 10% by track record ([Silzer & Church, 2010](https://www.crforum.co.uk/wp-content/uploads/2025/04/Identifying-and-Assessing-High-Potential-Talent-Current-Organisational-Practices.pdf)). Those four definitions produce four different lists from the same workforce.

Write the target down before the names are discussed. "Ready for a first people-leadership role within 24 months" is a testable claim. "Has potential" is not.

## Step 2: Check whether your performance signal predicts the target role

This is where most reviews quietly fail, and the evidence is unusually clean.

**Sales, 131 firms.** Benson, Li and Shue studied 38,843 sales workers, including 1,553 promoted into management ([*Quarterly Journal of Economics*, 2019](https://academic.oup.com/qje/article/134/4/2085/5550760)). Their findings:

- A worker ranked top in sales on their team was about **three times as likely to be promoted** as an average salesperson.
- Doubling a manager's pre-promotion sales was associated with each of their subordinates' sales falling by about **6.1%**.
- A promotion policy optimised for managerial quality rather than sales rank would have raised average managerial quality by roughly **30%**.

**Football, 55 years.** Schleu and colleagues tracked 407 Bundesliga head coaches, 309 of them former professional players ([*Journal of Business and Psychology*, 2024](https://link.springer.com/article/10.1007/s10869-023-09893-9)). Playing performance and coaching performance showed an average correlation of **r = .06** immediately after promotion, and path models controlling for squad quality returned effects statistically indistinguishable from zero at every time point. The authors' conclusion is blunt: they found no link between employee performance and leader performance.

Two very different industries, two very different measures of output, the same answer. Excellence in the current job is weak evidence about the next one — and in the sales data, the correlation with *managerial* outcomes runs the wrong way.

## Step 3: Add one piece of evidence the nominating manager cannot supply

The fix is not to abolish manager nomination. Managers see things no instrument sees: reliability under pressure, how someone handles a client who is wrong, whether they tell you about the problem early. The fix is to make the potential axis depend on at least one source that is not the nominating manager.

Practical options, in rough order of cost:

**Low cost**
- A structured work sample built from the target role, not the current one — a budget trade-off, a staffing conflict, a performance conversation.
- A scenario-based judgement exercise scored against a pre-agreed rubric, so every nominee faces the same situations.

**Higher cost**
- A development centre with trained, calibrated assessors.
- A cognitive and personality battery interpreted by someone qualified to interpret it.

The point of any of them is the same: a second reading, taken under standard conditions, that the committee can compare across nominees. Platforms like AssessAll exist for this step — AI-graded scenario assessments with a common rubric and integrity bands, so eight nominees can be put through the same target-role exercise without building a development centre to do it.

## Step 4: Decide list size and cycle before you see the names

In the same survey, 20% of organisations set a target distribution for the pool — usually 5–20% of salaried employees, most commonly 9–10% — and 75% held formal talent review meetings, 60% of them annually. Deciding the size afterwards turns the list into a negotiation about people rather than a standard.

## Step 5: Write down what removes someone from the list

Sixty per cent of surveyed organisations tracked high-potential turnover, which ran between 3% and 11% a year. Pools leak, and an unmanaged pool silently becomes a tenure list. Agree in advance: what evidence would take a name off, who reviews it, and how often. A pool nobody can exit is a pool nobody can trust.

## The checklist

1. State the target role and the time horizon, in writing, before names are raised.
2. Confirm who supplied each axis. If one person supplied both, treat the grid as one rating.
3. Ask what your performance measure is known to predict — and whether anyone has checked it against the target role.
4. Require at least one standardised, non-manager source of evidence per nominee.
5. Fix pool size and review cadence before the list exists.
6. Define exit criteria at the same time as entry criteria.
7. Keep the records. Under the US [Uniform Guidelines on Employee Selection Procedures](https://www.ecfr.gov/current/title-29/subtitle-B/chapter-XIV/part-1607), promotion is a selection procedure like any other, and a nomination process that gates promotions is inside that definition — whatever you call it internally.

## What this does not fix

Replacing manager intuition with an instrument only helps if the instrument is better. Be careful with claimed predictors of potential. A meta-analysis of 20 field studies reported corrected correlations of around **ρ = .74** between learning agility and leader performance ([De Meuse, 2019](https://articlegateway.com/index.php/JOP/article/view/1088)) — a figure far above the validities reported for structured interviews or work samples in the mainstream selection literature. A number that large in a young construct with many competing measures is a reason to ask what the criterion was and who rated it, not a reason to buy. Silzer and Church made the definitional problem explicit a decade earlier in [*The Pearls and Perils of Identifying Potential*](https://doi.org/10.1111/j.1754-9434.2009.01163.x), and the field has not fully settled it.

The honest position: potential is harder to measure than performance, every available method is imperfect, and the appropriate response is triangulation plus documented reasoning — the evidentiary discipline the [*Standards for Educational and Psychological Testing*](https://www.testingstandards.net/open-access-files.html) ask of any high-stakes inference.

**When the simple approach is right:** in a small team where the nominating manager has directly observed the target work — a team lead stepping up in the same function, with the manager having seen them run the team for three months — a nomination and a conversation may be the proportionate answer. Formal assessment earns its cost when the target role differs from the current one, when nominees come from managers who rate differently, or when the decision will be contested.

The takeaway: a nine-box grid tells you how a conversation went, not how a person will do. If the potential axis has no evidence source of its own, you have not measured potential — you have measured performance twice and promoted the rating.

---

Source: https://www.assessall.com/blog/both-axes-of-your-nine-box-come-from-the-same-manager-walking-through-a-hipo