Evidence, not optimism: introducing The Peak Survival Guide

Golden Quarter starts before Black Friday does, and so should your plan.

Diwali runs 6–10th November. Singles' Day lands on the 11th. Black Friday and Cyber Monday close the month on the 27th and 30th. By the time most retailers start treating peak as urgent, three trading events have already tested the platform.

That's the gap The Peak Survival Guide is built to address: a practical, prioritised breakdown of what to fix first across the five risk areas that take retail platforms down at peak, built from what we've actually seen work with real clients.

Why we didn't write another checklist

Every vendor in retail tech has a peak readiness guide. Salesforce has one. So does Shopify. They're not wrong, they're just a bit basic. Many are more of a marketing post than something a team can actually work from.

"We wanted to produce something that was properly based on our experience, based on what the guys are really doing with big clients, and distilled into something useful. It's a guide that technical teams can go deep on, but that also frames it for tech leaders."

Steve Dennis

It's deliberately not a playbook either. The Peak Survival Guide gives you the broad areas to focus on and the freedom to adapt them to your own context, rather than a script to follow step by step.

The five risks that take retail platforms down at peak

The guide is structured around five risk areas, covering fourteen activities in total, each graded Must, Should or Could so teams can prioritise rather than trying to do everything at once.

How should you prioritise peak readiness when time is short?

Start with the Musts. These are the eight activities to complete before your code freeze, and they come first because everything else depends on them. You can’t sensibly test in production before you’ve written and proven basic scripts.

• Rank your critical journeys
• Model scenarios on real traffic
• Know your test environment
• Build your toolkit
• Plan your test schedule
• Use the right test data
• Know which metrics to trust
• Rehearse, rehearse, rehearse

If time allows, move on to the Shoulds: trace the root cause, close the gap in production, build observability and quantify third-party risk. The Coulds, making readiness a habit and adding performance testing to the pipeline, are for after peak.

The full guide covers what each activity involves and which risk areas it mitigates.

Grounded in what testing can and can't tell you

A lot of peak confidence comes down to grey areas: data sets, integration points, hardware limits, message throttling. None of it maps directly from test to live.

"Everything we do in test is a simulation, and it doesn't directly correlate back to live. A 100% pass rate in test isn't 100% production ready. What we're trying to do is understand where the confidence levels are, understand where the risks might be, and make sure those are clearly flagged. It's not somebody putting their head on the block and just saying yes, it's good to go."

Darryl Kennedy

That's the real value of the guide: not a false promise of certainty, but a clear-eyed way of knowing what you don't know, and where to focus the time you've got left.

What evidence shows a platform is ready for peak?

Not a 100% pass rate in test. Evidence of peak readiness covers:

• Load tested at realistic peak scale, in an environment you trust to reflect live
• Performance proven end to end in production, not just system by system
• Monitoring everyone can see, with the response rehearsed
• Third parties and SaaS providers confirmed to scale and respond
• The right people briefed on thresholds and on hand to make go/no-go calls.

Built on real peak trading, not theory

The five risk areas in the guide aren't hypothetical. They're the same ones we've worked through with clients trading at genuine peak scale.

ASOS handles millions of Black Friday orders on a platform that's tested continuously, not once a year. Currys saw a 29% year-on-year uplift in peak conversion after treating testing as an ongoing discipline rather than a pre-peak scramble. And Not On The High Street went into peak carrying 3 million monthly site visits, confident the platform wouldn't buckle under it.

Confidence isn't just technical

SaaS platforms scale. That was never really the question. The harder question is whether the rest of the machine scales with it: fulfilment, warehousing, the people who need to be in the room when something goes wrong at 3am.

"It's not just about having technical confidence in the scale. It's having confidence that the organisational machine will work. It's the people side of it, in terms of who needs to be involved in monitoring and making decisions, and whether they're on hand to do it. That's what lets you walk into your go/no-go decision with evidence, not optimism."

Steve Dennis

It's why the guide doesn't stop at the technical checklist. Each risk area pushes into the operational and commercial questions too, because a platform that scales technically but can't fulfil a next-day delivery promise has the same outcome for the customer as one that falls over completely. For the warning signs that peak is already in danger, read Peak paranoia is real. Here’s how to use it.

What's inside

• The full Must/Should/Could framework across all five risk areas
• Fourteen prioritised activities, graded so you know what to tackle first
• A structure built for both technical teams and business leaders to work from together
• Guidance that adapts to your own context, budget and stage of readiness. This isn't a one-size playbook.
  

It's not too late

If you're reading this and thinking it's already too late to change your peak plan, it isn't. Mitigating 80% of a risk beats mitigating none of it. And knowing where your flags are beats not knowing.

 

GET YOUR PEAK SURVIVAL GUIDE