The Task Force, Not The Steering Committee · The Reluctant Guide to Shopify Migrations

You, specifically—the one about to send the calendar invite for the thirteenth meeting of the migration steering committee. Before you hit send, we’d like a word.

We have sat in your steering committee. Not yours exactly, but enough of them that the differences have stopped registering. It meets every two weeks. The agenda is the agenda from last time, lightly edited. There is a slide that says On Track in green, and a slide that says Risks, and the risks are the same risks, aging in place like wine that isn’t getting better. Someone presents. Heads nod. A decision is, occasionally, requested—and then deferred, because the person who could actually make it isn’t in the room, or is in the room but wants to “take it back to the team,” and the team is three floors down and will hear about it on Thursday. The meeting ends on time. Everyone agrees it was a good meeting. Nothing that happened in it unblocked anyone.

We are not going to spend long on this, because there isn’t much to say once you’ve seen it. The steering committee, as most enterprises run it, is a body that receives a migration. It is briefed on the migration. It is reassured about the migration. It has opinions about the migration, which it expresses every other Tuesday, and which arrive at the team like light from a dead star—technically real, hopelessly late.

The thing a migration actually needs is a different animal, and it’s worth naming the difference precisely, because the two get confused constantly and the confusion is expensive. What a migration needs is a task force.

A task force is four to six people. The e-commerce director. A senior engineer. A design lead. An operations lead. Possibly one more, if the brand’s shape demands it. They are not chosen for their titles. They are chosen for one property above all others, and we’ll get to it, because it’s the whole game.

They show up to standups. Not to a biweekly ceremony—to the actual daily rhythm of the work, or close enough to it that they know what’s stuck on the day it gets stuck, not on the Tuesday after. When something is blocked, they unblock it, in the room, without escalating it up a chain and waiting for the answer to filter back down. They are not spectators to the migration. They are inside it. The migration is, in part, theirs—and a person treats a thing they own differently from a thing they’re briefed on.

A task force has skin in the game daily. A steering committee has skin in the game quarterly. That is the entire distinction, and almost everything else follows from it.

Here is why the cadence matters so much, and it’s not a preference. The decisions a Shopify migration generates do not arrive on a two-week schedule. They arrive whenever the work hits the next thing nobody decided yet, which is most days. The ERP integrator needs to know, by Thursday, whether wholesale tax is computed on their side or yours. The design lead needs to know, this afternoon, whether the gift-set bundle is one product or three, because the PDP template forks depending on the answer. Should the legacy customer tags map to Shopify tags or to a metafield? That one’s blocking the import, and the import window is Friday.

None of those can wait for the thirteenth meeting. If they wait, the team does one of two things, both bad. It stalls—and a stalled day on a five-month project is not a rounding error, it’s a brick out of the wall. Or, more often, it guesses, picks the plausible answer, and keeps moving, and the guess is wrong in a way nobody notices until assembly, when the wrong guesses have married each other and had children. A biweekly meeting cannot absorb a daily decision rate. It was never built to. Asking it to is like asking a mailbox to function as a phone.

A vintage curbside mailbox with its door open and flag up, a rotary telephone handset perched improbably on top of its lid, the coiled cord drooping down the post into the grass. — Built for letters. Not for emergencies.

So who’s on the task force. The roles are easy to list and the temptation is to list them and stop: e-comm director, senior engineer, design lead, ops lead. But the roles are not the qualification. The qualification—the one credential that actually matters, the thing you should screen for before you screen for anything else—is that the person can make a decision without leaving the room to ask permission.

That’s it. That’s the bar. Not seniority, though it often correlates. Not technical depth, though you want that too. The qualifying property is decision authority that travels with the person into the room and doesn’t have to phone home. A “senior engineer” who has to escalate every architectural call to someone not on the task force is not on the task force; they’re a courier for someone who is, and you should have put that someone on it instead. A design lead who can’t approve the bundle template without a committee is a bottleneck with a nice title. Put people in the room who can say yes, do it that way and have it mean something the moment they say it. If you cannot find four such people in your organization, you have learned something important about your organization, and it is better to learn it now than in month four.

At Greycott the task force settled out about the way you’d expect once you knew the cast. Owen Caldwell, Head of E-Commerce, who wanted this done right and was out of his depth but learning, and—crucially—who could actually decide things about the storefront without convening anyone. Daniel Okafor, the lead engineer who knew where every workaround in the old Magento stack was buried, and who could therefore say that one dies, that one comes along without a meeting. A design lead. Priya Shah, Director of Operations, who ran wholesale invoicing and retail forecasting and reconciliation and knew, better than anyone in the building, what the new system would have to actually do—and who could speak for operations without taking it back to operations, because she was operations. Four people who showed up to the work, not to a presentation about the work. Naomi Brennan, the PM, didn’t have to chase them for decisions, because the decisions lived in the room with them.

Four completely mismatched chairs—a folding chair, an office chair on wheels, a wooden stool, and a battered armchair—pulled into a tight circle around a tiny table with four coffee cups on it, all seats empty but visibly mid-use. — The operating body. No slide deck required.

How To Know If Your Steering Committee Is A Steering Committee Or A Performance.

Six questions. Answer them honestly, which is the only way they work.

One. In the last six weeks, name one thing the steering committee decided that the team could not have decided without it. (If you’re still thinking, that’s the answer.)

Two. When a decision is requested in the meeting, is it usually made in the meeting—or is it usually “taken back to the team”? (A committee that takes everything back to the team is a relay station, not a decider.)

Three. Could you cancel the next meeting and have the project lose nothing? (Try it. Observe what happens. Notice, in particular, whether anyone notices.)

Four. Are the Risks slide and the On Track slide telling the same story? (If everything’s been on track for eleven weeks and the risks haven’t moved, one of those slides is lying, and it’s usually the green one.)

Five. Who in that room can say yes, ship it that way and have it be true the instant they say it? (If the answer is “nobody, we’d have to check,” your operating body is somewhere else, and you should go find it.)

Six. If you described the meeting to the team three floors down—the ones doing the actual migrating—would they recognize it as the meeting where their work gets unblocked, or as the meeting where their work gets described? (You already know which.)

If the answers point the wrong way, the fix is not a better agenda. The fix is to stop asking the steering committee to do a job it was never shaped for, hand that job to four people who can do it daily, and let the committee do the rarer, slower, genuinely senior thing it’s actually good at.

This is the part where someone—possibly you, possibly the person three chairs down who scheduled the thirteenth meeting—objects: but we still need the steering committee. And you do. We are not telling you to fire the senior stakeholders into the sun.

You need a steering committee for the handful of decisions that genuinely require a senior altitude, the ones the task force should not and cannot make alone. A budget overrun. A scope change that moves the launch date. A trade-off that quietly threatens the business case that got the migration approved in the first place—the kind of drift that the people heads-down in the work are least equipped to notice, because they’re heads-down. For those, you want people removed enough from the daily to still hold the original vision clearly, and senior enough to unblock the project when it’s the project that’s stuck, not a ticket. That cadence is real and it’s valuable and it operates at a different altitude than the task force, which is precisely why it should never be confused with the operating body.

The error is not having a steering committee. The error is pretending it’s the operating body—running the whole migration as a series of status reports delivered to a room that meets twice a month, and mistaking the absence of objections in that room for progress. The steering committee governs. The task force operates. When you collapse the two, you get the thing we opened with: a body that meets every other Tuesday to be reassured, and a team three floors down quietly guessing.

At Greycott, Andrew Whitfield—the CFO, who asked continuous questions and received, by the nature of the thing, point-in-time answers—sat on the steering committee, where his questions belonged. He was not on the task force. He has been told this, gently, and the telling went fine, because the distinction, once you say it out loud, is not insulting. It’s just true. A CFO who asks good questions every two weeks is exactly what a steering committee is for. A CFO who has to be in the room every time the gift-set template forks is a CFO whose week you have ruined for no reason, and a task force whose velocity you have capped at the speed of his calendar.

That’s the whole chapter, and it’s short on purpose, because the idea is short. Build the small body that decides and ships, made of people who can decide without asking permission. Keep the larger body for the few decisions that need real altitude, and stop pretending it runs anything. The distance between a migration that moves and one that meets is mostly the distance between those two rooms—and which one you put the deciders in.

So: that calendar invite. Before you send it, ask who in the room can actually say yes. If the answer is nobody, you haven’t scheduled a decision. You’ve scheduled a description of one. Send a different invite.