Deploying Around the Freeze: Retail IT Before Peak Trading

Sometime in mid-November, most Australian retail estates lock. No new hardware, no network changes, no software pushes, nothing that could take a lane down in the fortnight that decides the year. The change freeze usually holds until mid-January, sometimes later for retailers with big Boxing Day and back-to-school trade. Which means the retail deployment year is not twelve months long. It is ten, and the last usable weeks are being spent right now.

The calendar maths nobody does in January

Count backwards from a 15 November freeze. A 300-store program needs its final wave complete by early November, which puts the last waves in October, the pilot in September, and site audits in August. Retailers who start scoping a pre-Christmas rollout in September are not planning a rollout, they are planning a postponement. The freeze does not negotiate, so the schedule has to.

The retailers who get this right treat the freeze date as the fixed point and build the entire program plan backwards from it, with buffer. A wave slipping by two weeks in March is an inconvenience. The same slip in October pushes the tail of the program into February.

What the freeze actually covers

Freezes stop change, not support. Three things keep running through peak:

Break/fix, at higher stakes. A dead lane on 23 December costs more per hour than at any other point in the year, so response arrangements should tighten during the freeze, not relax. Spares forward-staged, escalation paths confirmed, downtime treated as the enemy.

Emergency change. Every freeze has an exception process. The test of a good one is that it is documented before November, with named approvers, rather than invented by conference call on a Saturday.

New store openings. Retail keeps opening stores in November because opening season is trading season. Openings run under their own controls, isolated from the trading estate.

What the freeze is actually for

Deployment teams that treat the freeze as dead time waste a free quarter. The locked months are the natural window for next year's planning: audits of the estate as it actually stands, hardware procurement before lead times bite, staging and kitting in the integration centre, pilot design for a February start. The programs that move fastest in autumn are the ones that did their thinking in December.

The freeze also produces the year's best evidence. Peak trading is the estate's annual stress test, and January's incident data is a ranked list of what to fix, refresh or replace, written by the business itself.

The planning rule

One line worth putting in every retail IT calendar: the year ends in October. Scope in winter, pilot in early spring, deploy through October, and hand the estate to peak trading in the most boring possible state. Boring is the deliverable. Our retail deployment practice is built around that calendar, and the retail chains we support have learned to love an uneventful December.

Planning a rollout that must land before the freeze? Speak to an expert.