Skip to contentAnemo
EN
Contact

Peak Sales Week: A Systems Readiness Checklist

· 4 min read

Peak sales week does not break the part of your business that customers see. It breaks stock synchronisation, payment provider limits, notification queues and the admin panel your own staff use, and it breaks them in that order. Six weeks out is enough time to find those limits deliberately rather than discovering them on the day.

This guide is written for an owner or manager who has to ask the right questions of a development team, without being able to read the code themselves.

Key takeaways

Ask what load has actually been tested against

The question to ask is not "will it hold". It is "what number was it tested against, and where did it start to fail". Any team that has done the work can answer both. A team that answers only with reassurance has not done it.

Pick the number together from last year's peak, then test well above it. Two useful figures: concurrent visitors, and orders per minute. The second is the one that matters, because it is what drives writes to stock, payments and order records.

Know which parts break first

The storefront is usually the most resilient part of the system, because it is read-heavy and can be cached. The fragile parts are the ones that write.

Freeze what can be frozen

Agree a date after which nothing is deployed except a fix for something that is broken. Two weeks before the campaign is a reasonable point for most businesses.

The reason is simple: during the week itself, every change is made quickly and under pressure, which is exactly when mistakes happen. A discount rule adjusted at nine in the evening has ended more campaigns than any traffic spike.

Decide the fallback before you need it

For each of the failure modes above, decide now what the response is and who makes the call.

Write these down on one page. The page is the deliverable, not a document nobody reads.

Agree the support arrangement in writing

Your development partner's normal working hours are not the campaign's hours. Before the week, agree response times, who is reachable, through which channel, and what counts as an emergency. If this is not written down, it does not exist.

The same applies internally. One named person with the access and the authority to act beats five people waiting for approval.

Measure the week while it happens

Decide in advance the three or four numbers you will watch: orders per minute, checkout completion rate, error rate, and time to process an order internally. Having these visible on one screen turns arguments into observations, both during the week and afterwards.

If preparing for the campaign uncovers work that needs product, engineering or integration support before November, Discuss Your Plan.

Frequently asked questions

How far ahead should preparation start?

Six weeks is a practical minimum for load testing, provider limit increases and a change freeze. Four weeks is possible if the only work is testing and agreeing a fallback plan. Two weeks is enough to write the escalation page and not much else.

What if we have never load tested anything?

Start with the order path alone. Testing checkout, payment and order creation against a realistic number gives you most of the information for a fraction of the effort of testing everything.

Is this only about the website?

No, and that is the most common mistake. The internal tools, the integrations and the people processing orders are part of the system, and they are usually where the queue forms.

Related services

Related reading

Building the product for what comes next

We would rather deliver one product that holds up than three that have to be rebuilt. That standard is the same on every project, whatever its size.

Ali Boran GazelCEO

Contact us