Every September, retail programs with a fall go-live date hit the same fork. The team is close. The vendor says the date is achievable. Finance has already booked the benefits. And someone finally asks the question everyone has been avoiding: do we really want to switch this on weeks before Black Friday?
Most of the time, that decision gets made by the calendar — either “we committed to the date, so we go” or “it’s peak, so we freeze everything.” Both are shortcuts. The right answer depends on what you’d lose in each direction, and how well you can actually contain the downside.
The costs are not symmetric
A delayed go-live costs real money — vendor and team carrying costs, deferred benefits, a program that loses momentum. But it’s a cost you can see, size, and plan around.
A failed go-live in peak season costs you in the weeks when your revenue is highest, your customers are least forgiving, and your own teams are already stretched supporting business-as-usual volumes. You can’t take a week to stabilize in production when every hour of a problem lands on your busiest trading days. The failure is more expensive, and it arrives at the worst possible time to fix it.
That asymmetry doesn’t mean “always wait.” It means the bar for going live before peak should be higher than the bar on any other week of the year — and that bar should be tested, not assumed.
Five tests before you commit to a pre-peak go-live
1. Can you roll back in hours, not days? A rollback plan that has never been rehearsed isn’t a plan, it’s a hope. If reverting would take longer than your tolerance for an outage on a peak trading day, you don’t have a rollback — you have a one-way door.
2. Have you tested the edge cases, not just the happy path? Peak volume doesn’t just mean more orders; it means more of the awkward ones — stacked promotions, split shipments, partial fulfillment, returns and exchanges arriving alongside new orders. A system that handles the standard order flow correctly can still fail on the mix of exceptions that only shows up under holiday conditions. (This is especially true for order management — see where retail transformation complexity actually lives.)
3. Are stores and distribution centres actually ready? A technically correct system that hundreds of newly hired seasonal staff have never seen is not operationally ready. Training bandwidth is scarce in the weeks before peak, and every hour spent teaching a new process is an hour not spent on the seasonal ramp-up.
4. Who staffs hypercare when everyone is busy? The first weeks after go-live need fast, senior response. If your own team is committed to peak operations and your vendor’s delivery team has holiday schedules of their own, “hypercare” may exist on paper only. Name the people, confirm their availability, and confirm who has authority to make a call at 2 a.m.
5. Are the go/no-go criteria written down before the night, not decided on it? Decisions made at midnight, with the team exhausted and the date looming, drift toward “go.” Define the criteria — open defect thresholds, test coverage, rollback readiness, staffing — weeks earlier, and agree in advance who decides.
If you can’t answer yes to most of these with evidence, the honest answer is usually to hold.
It’s rarely a binary choice
“Go now or wait until January” is a false choice more often than not. Options that often sit in between:
- Phase by scope. Go live with a reduced, lower-risk scope now, and hold the complex functionality for after peak.
- Phase by exposure. Start with a single banner, region, or channel that carries less trading risk, rather than a full network cutover.
- Deploy, but don’t switch on. Ship the technical change ahead of peak while keeping the business cutover behind a controlled switch you can time deliberately.
- Move to a post-peak window. January is a natural stabilization period for many retailers — though returns volume means it isn’t quiet, so plan it as carefully as any other window.
Check your own change-freeze calendar early, too. Many retailers already have a freeze window ahead of peak, and a go-live date that collides with it may already have an answer you haven’t acknowledged.
Do the arithmetic out loud
The pressure to hit the date is usually about visible costs — contract milestones, committed benefits, executive expectations. The risk of going live is usually about invisible ones. Force both onto the table: what does three more months of delay actually cost, and what’s the realistic revenue exposure of a bad go-live across your highest-volume weeks? Programs that put those two numbers next to each other tend to make better decisions than programs that argue from momentum.
The bias to watch for
A program under delivery pressure is structurally biased toward going live. Everyone involved has been working toward the date for months, and nobody wants to be the one to say “not yet.” That’s precisely why the go/no-go read benefits from someone who isn’t invested in the date — a person with the standing to say it out loud, and no stake in the answer being yes.
If you’re facing a go-live decision before peak and want an independent read on where the risk actually sits, the Transformation Complexity Assessment is a fast way to get a structured view. For programs already showing strain, see program recovery — or let’s talk through what you’re facing.