T-Shaped, I-Shaped, X-shaped, M-shaped... Why the obsession with alphabetical labels?

David Turner
Jul 7
5 min read

Updated: Aug 10

T-Shaped, I-Shaped, X-shaped, M-shaped… Why the obsession with alphabetical Labels?

The letter is not the person: the problem with X-shaped and M-shaped skills

Somewhere this quarter, in a talent review, a manager will describe an engineer as M-shaped. Heads will nod earnestly and a box will be ticked. The engineer, who is not in the room, will spend the next year receiving work shaped like the letter rather than shaped like their ability. Nobody will check whether the label was accurate, because it does not feel like a claim. It feels like an observation, but it’s not really. It’s just the re-emergence of the letter-shaped profile, awoken this time by people who need something - anything - to say about AI.

The vocabulary persists, and AI has given it fresh fuel, because it solves a real problem cheaply, quickly and badly. Leaders need to know who can do what and where the team is thin. Tangible evidence of that is expensive to gather and tedious to maintain, whereas a gut-feel label is free. So organisations substitute the cheap proxy for the costly measurement, and the proxy becomes the target: people start performing their letter instead of doing the work. A tidy taxonomy also feels like control, which is more comfortable than understanding. And nobody's incentive ever pointed at correcting an error. The manager will not spend hours disproving a snap judgement. The HR team has a completed spreadsheet. The employee probably can't see the label at all.

Now add the current advice that machines will do the deep work, so everyone must become ‘M-shaped’ or ‘X-shaped’, and a lazy shorthand acquires the urgency of a survival strategy. The advice sounds rigorous, yet it’s no more than a metaphor in a lab coat. And frankly with a new letter-profile emerging almost weekly, the whole thing is starting to sound like an episode of Sesame Street. The only saving grace being that presumably it stops after 26 iterations (assuming we don't move on to the 50,000 characters of Mandarin).

The cost of course compounds quietly. Scullen, Mount and Goff analysed thousands of manager ratings in 2000 and found that more than half the variance in a rating reflected the rater, not the person rated. Actual performance accounted for about a fifth. The letter in the spreadsheet mostly records the manager's exposure and memory, and staffing decisions then run on that noise. The mislabelled specialist gets narrow work, managers stop inviting them to cross-team conversations, and within a year their record actually does look narrow. Snyder and colleagues demonstrated the mechanism in 1977: false beliefs about people draw confirming behaviour out of them.

The firm pays twice. Once for a measurement that does not measure, and again for the capability it buries.

All of us have at some point experienced the feeling of not being considered for a project we were deeply knowledgeable in, because we were assumed to be a specialist in something else. No-one wins in that scenario. The organisation loses out on the knowledge, the individual subconsciously (or in some instances consciously) becomes a fraction more cynical and apathetic. Which in turn can manifest in lower levels of engagement and maybe start or accelerate a journey toward leaving the organisation. All while the project is missing the value they could have provided, were it not for a letter of the alphabet being arbitrarily attached to them.

Shape is a property of the work, not the person

Shape is a property of the work, not the person. The important question is never whether someone is a generalist or a specialist. It is whether the ground under the work is stable or shifting. Teodoridis, Bikard and Vakili found in 2019 that specialists do better when a field holds still and generalists gain the edge when it moves. The same person is the right answer in one situation and the wrong answer in another. No letter captures that, because the letter describes the person and the value lives in the fit.

Nor are the roles the letters gesture at new. Adam Smith opened The Wealth of Nations with a pin factory because specialisation produced output no generalist could match. Ronald Burt spent years showing that people whose relationships bridge disconnected groups produce better ideas and get promoted faster, in ordinary firms, two decades before anyone said "M-shaped". Depth and coordination built cathedrals and railways. They are features of complex collective work, not of any technology. AI does not create these roles. What it plausibly changes is the ratio the market wants, by automating some of the routine execution that sat between them. A change in proportion therefore, rather than a new kind of valuable person.

Look at the automotive industry. For decades it’s been made up of exactly that mix. Vehicle programs are made up of specialists like software devs, procurement category buyers, dynamics and powertrain engineers. And they’re all held together and pointed toward a launch date by generalists. People who have some level of appreciation of the different specialist disciplines involved, but can perceive the overall system of information flows, funding, timing, resource profiles, regulatory compliance milestones and supply-chain risks. For most of that time no-one in Turin, Stuttgart or Aichi Prefecture were concerning themselves with which letter of the alphabet they felt most like. It was clear that the roles and personas required to succeed were varied. The work and the outcome dictated that, not a label.

What to do instead

Never apply the metaphor to named people. "This team is all deep specialists who do not talk to each other, so the next hire should bridge them" is a sensible sentence, and the ‘M’ can express it quickly. Pinned to a person in a review cycle, the same letter imports the rater noise and distorts that person's next year. These letter-shaped profiles are no more than lazy shorthand with the potential to negatively influence an employees output over the next cycle, and should be treated as such.

Instead, replace the letter with a tangible record. What this person has delivered, in which areas, with what result. Keep it current, let the person own part of it, and treat it as a record of the past rather than a verdict on the future. It costs more than a letter, which is the point. A fact you can check does not trigger the labelling effects, because "you delivered work in three areas" is not an identity anyone absorbs.

Employee rating procedures often consist of more than enough pseudoscience and pop-psychology as it is. Three-sixty reviews are famously the most effective way to get teams to play out their grudges against each other anonymously. Myers-Briggs at least produces variety, in that you get a different answer every time someone takes it. And recency bias accounts for the majority of a manager’s 1-5 staff ratings anyway. Before of course they all get hammered into a Gaussian distribution profile. Let’s not add alphabet-based shapes as well.

Of course the real genius of the alphabet-shape trend is how it utilises an age-old mechanism to appear credible. "Depth in your core area with real curiosity across neighbouring fields" fits nearly everyone, which is why nearly everyone nods. It is of course a Barnum statement, the long-favoured device of astrologers and fortune-tellers.

Forer demonstrated that mechanism in 1949, and it built the personality-testing industry. Flattery that fits everyone measures no one. The old consultancies repackaged this truth with a letter so they could sell it back to you. The specialist and the bridge-builder were valuable before a management consultancy somewhere needed a hiring slide, and they will be valuable after the alphabet runs out. And AI doesn't change that.

David Turner is the founder of Kói, an independent technology consultancy advising investors, founders, and boards on technology assessment and organisational capability.

You can reach him at: enquiries@dkoi.design