The Six Box

The nine box was built for the room where leaders argue about people. The six box was built for the system where a whole company’s worth of those arguments have to land in the same fields. When a company assesses talent at scale, placements become dropdowns, reports get pulled across functions, and “ready” in one division has to mean “ready” in another. At that size, fewer boxes with written definitions beat more boxes held together by tribal knowledge.

Six boxes come in a few shapes. The one I work with makes three trades, and I think all three are good ones.

It uses definitions instead of numbers. Nobody argues about whether someone is a five, because each box has a name and a sentence attached. That clears up a whole category of confusion. It costs you the dots-not-bins argument about someone sitting on the line.

It makes the middle row one box. Whatever their potential, a solid performer gets the same play, which is to keep them growing. That builds the bell curve right into the grid. On a nine box I have to teach leaders that most people belong in the five. Here the wide middle box makes the point before anyone says a word.

It splits the bottom row by potential instead of severity.Low performance with no real upside gets formal action. Low performance with real upside gets support and a window. That’s the fork that changes what you do next.

What you give up is resolution. The clock difference between a one and a four, the wrong-seat seven, and the line between a strong five and an early six all flatten out. So run the grid that matches your job. If you coach leaders on their people every week, run the nine box, because the coaching lives in the detail. If you’re standing up talent review across a whole company, run a six box, because one shared language matters more than detail at that size. And if you lead inside a company that runs a six box, do what I do. Keep the nine box as your working tool and treat the six box as the record.

Think of the six box as three shelves. The shelf is performance over time. The bottom shelf is below the bar, the middle is solid, and the top is exceeding, judged across a real stretch of time instead of one good quarter. The whole middle shelf is one box, whatever the potential.

Potential only splits the top and bottom shelves, and it asks a narrower question than the nine box does. Do they want more scope, and could they carry it? The answers are not seeking more, growing into more, or ready now. A bigger canvas counts, even without a new title.

Box names vary by company. These are mine. Open each one to see where my nine box placements land.

I read the grid through two lenses. Performance looks back at what someone actually delivered. Potential looks forward, about two years out, at what they could take on next. You need both to place anyone. If your company also rates people once a year, don’t mistake that rating for the box. The rating is the backward look at one year of impact. The box takes the longer track record, adds the forward look, and ends in a plan.

Performance over time. Judge the track record, not the spike. Look at what they delivered, how they delivered it, and whether the impact lasted. It’s the same consistency rule I use on the nine box. Average out the spikes and see what’s actually trending. One practice worth keeping on any grid is that when someone moves roles inside the organization, you don’t re-score them from zero. Start them at solid and let the new role prove itself.

Potential. The six box asks whether someone wants more scope and how soon they could carry it. That’s narrower than the nine box’s read on next-level competencies, and it’s why translating between them takes judgment. I look for the same things I always do, attitude and aptitude, plus appetite. Do they want more, and could they carry it without you holding it up for them? Leave out what doesn’t belong. Tenure, likability, and how much someone reminds you of yourself are not potential.

The companion questions. A good talent review asks three more things about each person, and they’re worth asking on any grid. What breaks if this person leaves? How likely are you to lose them? And is their seat critical, meaning it matters to the strategy and would be hard to refill? Hold that last one to a high bar and don’t default to yes. When someone does hand you their notice, you should already know whether it’s a crisis and what you’re doing about it.

Every box needs a plan, and the plan can be simple. Sometimes the plan is to keep them growing, keep watching, and let them simmer. That’s still a plan, as long as you chose it on purpose.

What I try not to do is peanut-butter my attention across the whole team. If I give everyone the same amount, the places that need me most get too little, and that isn’t doing right by the organization. My attention belongs on the top and bottom rows. The bottom row is a performance problem, and it gets a bespoke plan with a date, either to get someone performing or to help them out of the organization. The top row gets a bespoke plan too, because that’s where you develop, challenge, and keep your best people.

The Core gets a different kind of care. Make sure they’re paid well, taken care of, and supported, and that they understand the vision and where the team is going. That’s real work, just not a bespoke plan for every person.

Go box by box, by what each box means, not by where the dot sits. This is the mapping I use:

One warning, because I made this move myself the first time. My nine box strata are diagonal layers, and the six box’s shelves are flat. If you translate layer to layer, a three and a seven land in the same box, and they’re opposites. The three is exceeding and the seven is struggling. Layer to layer gets the one, four, five, and nine right, misplaces the two, three, and seven, and can’t tell you where a six or an eight belongs. Translate one person at a time.

The six box also can’t tell your six from your eight, since both land in Reaching. Keep the rank-or-reach difference in your own notes, because it changes the plan.

If you keep a nine box under a company six box, three habits keep the seam clean.

Answer the two questions, then read the box. When you enter people into your company’s system, you’re really answering two dropdowns, a performance level and a potential level. Answer them from your nine box evidence and let the box fall out. Don’t start from the box that sounds best and work backwards.

Keep your clocks yourself. The six box won’t tell a one from a four, and it won’t remember that a Support placement is really a wrong-seat seven with a landing plan in motion. Your working grid carries that. The company grid doesn’t need to.

Calibrate in their language. When you brief up or sit in a company calibration, speak the six box, because a shared language is the whole point of the simpler grid. Translate once, then commit to the room.

Whatever grid the room runs, the same biases do the damage. Name them out loud before you calibrate.

Skill and will. A two-by-two of capability and motivation, usually used to pick a coaching style for one person. It’s a fine conversation tool. It’s too coarse to run a team on, and the will axis invites exactly the bias a good process tries to squeeze out. Use it to decide how to coach, not to make talent decisions.

Ratings alone. A single performance score with no potential axis. The classic failure is promoting your best individual performer into a job they never wanted. The second axis exists so that excellent right where they are can be a celebrated answer.

Stack ranking. As a forced distribution with consequences attached, it earned its reputation. As a private exercise it still has a use, and I keep it as one. Rank fresh twice a year and study what moved. The movement asks better questions than the list.

OKRs and scorecards. Goal attainment tells you about the work. It doesn’t tell you about the person’s trajectory, and it can be unfair across unequal territories and assignments. Feed the numbers in as evidence, and don’t mistake them for the assessment.