Record revenue can still hide a poor Q4.

When somebody tells me ecommerce revenue is up, my first response is normally: good, but why?

That is not me trying to find a problem in a positive result. I want to understand what created the growth and whether the business can repeat it.

Revenue may be up because more customers bought. It may be the result of a deeper discount, one unusually large order or a burst of activity that cannot be repeated tomorrow.

The number tells you what happened. It does not tell you whether it was healthy.

As an ecommerce manager, I was not always close enough to the final profit and loss reporting to say with certainty whether every campaign had been profitable. That fuller picture often sat with insight teams, Finance and the board.

But operating the ecommerce channel gave me enough signals to know when the revenue headline was only telling part of the story.

For me, the better peak question is: how much profitable demand did we convert, could we fulfil it as promised and what did those customers do next?

Start by explaining the revenue movement.

Revenue can increase because more people ordered, customers spent more or pricing and discounts changed.

That is why I always looked at order numbers alongside revenue. Orders told me roughly how many customers were buying, and average order value helped explain how much each was spending.

Although, if I am honest, I normally had average order value in front of me too, so I was not doing the calculation every morning.

What matters is the relationship between the measures.

If revenue rises while order volume stays flat, the change is probably coming from basket value, price or product mix. If orders rise but average order value falls, deeper discounting or a shift towards lower-priced products may be involved.

Neither result is automatically good or bad. They simply tell you to ask a different next question.

More traffic does not always mean better trading.

Traffic gets plenty of attention during peak because it appears quickly and is easy to compare. On its own, though, it tells you very little about the quality of the demand or what customers found when they arrived.

Analysis published by the British Retail Consortium using IMRG and Ecommpay data found that November 2024 traffic was down 4.3% year on year while conversion during Black Friday week improved to 4.6%.

Fewer visits did not prevent stronger commercial performance.

I would not treat those figures as targets. Every brand has a different product range, margin structure, audience and acquisition strategy. The useful point is simply that volume should never be read on its own.

I tend to look at peak performance through five connected areas.

Is the demand any good?

First, understand where demand came from and whether it arrived when expected.

Review sessions by channel, campaign, device, customer type and landing page. Compare them with the plan, the equivalent period last year and recent trading.

Paid media spend needs to sit alongside traffic. Otherwise, it is easy to celebrate an uplift that the business has simply paid more to create.

I would ask:

• Did demand land on products we actually had available?
• Was it more expensive to acquire?
• Did new and returning customers behave differently?
• Did mobile and desktop contribute as expected?
• Which campaigns brought people likely to buy, rather than simply more visits?

Where did the journey improve or break down?

Conversion rate matters, but the blended figure can hide more than it reveals.

Break it down by device, customer type, channel and stock position. Then look through the journey: product views, add-to-basket activity, checkout starts, payment success and completed purchases.

A drop between a product page and basket needs a very different response from a payment failure during checkout.

Revenue per session is useful because it connects conversion and average order value. I would then look at average order value alongside units per order and discount depth.

A higher basket value could be the result of a better product mix. It could also mean customers bought more units because the discount was deeper.

The reporting needs to help the team tell the difference.

Did the orders create value?

This is where a record sales day can stop looking quite so impressive.

I once saw a product managed by another ecommerce team sell out in a single day. A very effective combination of stacked discounts meant customers could buy it for around 20% below its cost price.

It was a consumer’s dream.

A dashboard focused on conversion and sell-through would have made the result look fantastic. Commercially, it was anything but.

That example is why product margin and discount stacking need to be visible while the campaign is live, not discovered during the review afterwards.

Track gross margin and contribution per order alongside revenue and return on ad spend. Contribution should include the costs that change materially during peak, such as discounting, media, fulfilment, payment fees and expected returns.

Useful measures include:

• gross margin value and percentage;
• contribution per order;
• customer acquisition cost;
• new-customer percentage;
• discount cost by product and category;
• cancellation and expected return exposure;
• incremental return from paid activity.

The question is not just whether the business processed more money. It is whether it created more value.

Can the business fulfil what it has sold?

The sale is not complete when the confirmation email is sent.

Operational reporting needs to show whether stock was available and whether the order moved quickly enough to meet the promise made to the customer.

Selling out too quickly can be a problem as well as an achievement. Stock intended to support several weeks of activity can disappear in the opening days, leaving paid campaigns running without enough relevant product to convert the people they attract.

The team then has to pause activity, redirect customers towards a less suitable product or find stock for the next promotional moment.

I would track:

• in-stock rate;
• oversells and cancellations;
• orders dispatched within the promised time;
• time from order to first carrier scan;
• delivery performance;
• warehouse backlog by age.

These measures should be in the same daily conversation as sales and media spend. Otherwise, Marketing can keep scaling demand after Operations has reached a safe limit.

What happens after the order?

Peak creates costs and customer behaviour that will not be visible in the first few days.

Extended Christmas returns make this particularly important. A campaign that looks strong in November can continue changing throughout December and January as items are returned, refunds are processed and stock becomes available again.

Customer contacts, delivery enquiries and compensation costs also tend to appear after the main campaign has finished.

Add these measures to the scorecard:

• contacts per 1,000 orders;
• where-is-my-order contacts;
• first response and resolution time;
• customer satisfaction;
• return rate by product, category, channel and cohort;
• refund cycle time;
• 30, 60 and 90-day repeat purchase rates for peak customers.

Without that later view, a team can celebrate customer acquisition in November and only understand its quality in January.

Agree which numbers the business trusts.

Differences between ecommerce systems are normal. Ignoring them is not.

Trading and marketing teams may be working from Shopify and GA4. Finance may be using Sage, QuickBooks or another reporting platform, with teams manually combining the information in spreadsheets.

It is very easy for each department to arrive at a different version of the same trading period.

Some differences are legitimate. Platforms use different definitions, dates and attribution models. Others point to a tracking or reconciliation problem.

Until those differences are understood, daily meetings become a debate about whose number is right instead of a decision about what to do next. Peak week is the wrong time to start resolving that.

Before Q4, agree four things.

One baseline.

Use last year’s comparable event, the most recent eight to twelve weeks of trading and similar promotions. External benchmarks are useful context, but the brand’s own customers and products normally give you the better comparison.

One definition for each measure.

Agree what counts as net sales, an order, a return, a new customer, an on-time dispatch and a profitable order. Treat cancellations and refunds consistently.

One trusted reporting view.

The daily dashboard should reconcile as closely as possible with the systems used by Finance, Trading and Operations. Any expected differences should be known before peak starts.

One response when a threshold is crossed.

A measure only becomes useful when it changes a decision.

Set thresholds for low stock, falling payment authorisation, warehouse backlog, delivery exceptions, rising acquisition cost and contribution dropping below plan. Give each one an owner and an agreed action.

The five measures I would want every day.

If I had to reduce the daily Q4 view to five measures, I would choose:

1. Net sales against target to show whether trading is on plan.
2. Conversion rate to show how effectively demand is becoming orders.
3. Contribution margin to show whether those orders are creating value.
4. Average order value to help explain customer spend and product mix.
5. Stock and fulfilment risk to show whether the business can keep trading and meet its promises.

I would still use order volume, channel performance and customer contacts to diagnose any movement. These five give me the first signal and tell me where to look next.

A useful daily view should answer:

• Are net sales and contribution on plan?
• Why has performance changed?
• Which products are driving the result?
• What stock or fulfilment risk is building?
• Is the customer journey working as expected?
• What needs to happen today, and who owns it?

You do not need dozens of charts. You need a small number of connected measures that help people make a decision.

The final answer will not be available until January.

The proper peak review should happen once enough returns, refunds and repeat behaviour are visible.

It should explain forecast accuracy, promotional contribution, stockouts, fulfilment, delivery issues, return behaviour and the early value of customers acquired during peak.

Keep a record of important decisions and incidents while trading is live. That context is often the difference between knowing what happened and understanding why it happened.

The review should finish with named actions and deadlines well before the next Q4 planning cycle begins.

The businesses that get better at peak each year do not necessarily collect more data. They connect the commercial, customer journey and operational measures early enough to act on what they see.

At blubolt, we help Shopify brands build that connected view, from reliable ecommerce tracking and reporting through to the journey improvements the data identifies. If your Q4 dashboard starts and ends with revenue, it is still only telling part of the story.

Get the full picture with blubolt.

Ready to uncover the complete story behind your Q4 revenue and build a more profitable strategy for next year? Explore our services to see how we help Shopify brands optimise their end-to-end performance, and browse our work to discover the results we’ve delivered for other ambitious retailers. When you are ready to turn your raw data into actionable growth, get in touch!