How Experienced Merchants Turn August Data Into a Q4 Discount Strategy

In November, somebody on your team will say it out loud. "Last year, 25% worked." It is the most confident sentence in Q4 retail. It also sits on a dataset that could not possibly have proven it. Last December, every visitor who landed on your site saw a discount somewhere. You never ran the version of the store where they did not.

Planning this Q4 off last Q4 feels like the responsible move. It is your own data, your own customers, same season, sitting right there in the admin. The numbers are real. Here is the problem. They describe a market where every store discounted at the same time. That makes them a record of what happened, not evidence of what caused it. So last year's holiday results are the worst input you have for this year's q4 discount strategy. There was no group of shoppers who saw full price and bought anyway. Without that group, you cannot tell which orders the 25% created and which ones it simply paid for.

Experienced merchants noticed this years ago and quietly changed their input. They stopped reading December and started reading August. August is the last month of the year that still contains a control group. Call it what it is: the August Control Group. By the end of this you will have three August measurements and the Q4 decision each one settles. Plus one multiplication that prices your discount before it converts a single extra order.

Last December Recorded What Happened, Not What Worked

A Q4 result only becomes evidence if some comparable group did not get the discount. In November and December, almost nobody in your traffic qualifies. Your list got a code. Your ads carried a promo. The shopper comparing you against three other tabs saw a banner on every one of them. So "25% converted at 3.1%" is not a finding about 25%. It is a description of a fully discounted market.

The Missing Cell in Your Holiday Report

Think of your holiday promotion as a test with one cell filled in. You ran 25% off, you recorded a conversion rate, you wrote it down. What you never ran was the same week at full price, with the same traffic, the same products, and the same competitors. Without that second cell there is nothing to compare against. You have a number floating on its own. That is worse than having nothing, because a number with no comparison still feels like proof. It is how a store discounts deeper every single year while believing it is following the data.

What Q4 Data Is Actually Good For

December is not useless. It is misassigned. It gives you an honest read on your demand ceiling. It shows how much volume fulfillment and support can absorb before quality slips. It tells you how shipping cutoffs move behavior and which product pages hold up under real load. Those are operational answers and they are worth having. Pricing is the one question December cannot answer, because pricing is exactly where every store in the market moved at once.

December told you what people bought. It could not tell you what they would have bought at full price, because by then there was nobody left to ask.

August Is the Last Month With a Full-Price Baseline

August is the only month before Q4 that puts two different shoppers on your site in the same week, at real volume. One has a deadline. One does not. That split is the entire reason August data can answer a pricing question and December data cannot.

Two Populations, One Month, One Store

Back-to-school shoppers behave like holiday shoppers. Fixed budget, hard date, heavy comparison, low tolerance for shipping risk. A parent with a $60 supply list and a Thursday deadline converts the way a dedicated buyer does. Fewer sessions, faster decisions, less price shopping. Meanwhile the rest of your traffic has no deadline at all. They browse, they add to cart, they drift off with an "I'll buy it later" mentality. Most of your catalog outside the seasonal categories is still selling to them at full price. That is your untreated group. Q4 will never hand you one.

The Baseline Has an Expiry Date

Your full-price baseline is perishable. Every promotion you run between now and Thanksgiving shrinks the share of traffic that is still untreated. By the time most stores sit down to plan Q4 in October, the clean read is already gone. This is the real reason August planning beats October planning, and it has nothing to do with being organized. The data is simply better right now than it will ever be again this year.

The August Control Group: the last cohort of the year that will buy from you without being paid to. Measure it before fall promotions erase it.

Three August Measurements, Three Q4 Decisions

A measurement that does not end in a decision is a hobby. So pair every number with the specific Q4 argument it settles. Three numbers do almost all of the work, and two of them are sitting in your store right now.

The Multiplication That Prices Your Q4 Discount

Start with your August full-price conversion rate, measured only on traffic that saw no promotion at all. Say it comes back at 1.8%. Now take your realistic November session forecast. Say 95,000. That is 1,710 orders your own data says arrive with zero promotional support. At a $68 average order value, that is $116,280 of revenue you were already going to receive. Put a 20% sitewide discount on Q4 and you just handed back roughly $23,256 before the promotion converts one additional order. Those inputs are made up for the example. Yours are not. That figure is the entry fee, and you pay it whether the promotion works or not. Most merchants have never run the multiplication. It is why the November argument is about the percentage instead of about who the percentage reaches.

Reading Depth From Behavior, Not From the Market

The next number is the one nobody goes looking for. How small did the offer have to be before a window shopper changed course? If a visitor drifting toward the exit converted at 10%, then 10% is your evidence and 25% is your habit. Watching competitors pushes depth in one direction only. You can see their banners. You cannot see their margins. Your August behavior data pushes the other way, and it is the only input in the room that is actually about your store.

The blocker is rarely willingness. It is that these numbers live in three different places, and none of them split promoted traffic from full-price traffic. Growth Suite's funnel report tracks the whole path from session start to completed order. Its purchase insight report shows how many sessions and how many days a real buyer took before converting. That is the number that should be setting your offer window. Product segmentation then sorts your catalog into groups like Stars, Essentials, and Underperformers. Your Q4 exclusion list gets built from behavior instead of from memory.

September Is Your Test Bench, Q4 Is Not

What August gives you is a hypothesis about depth, window length, and exclusions. A hypothesis is not a plan. Q4 traffic is the most expensive traffic you buy all year, which makes it the worst possible place to discover you were wrong. September has real traffic, real intent, and low stakes. That is the exact combination a test needs.

What a Fair September Test Looks Like

Pick your metric before you start, and pick it honestly. Total profit, not gross revenue. Otherwise you repeat December's mistake in a smaller month. Split traffic between your August-derived depth and the depth you would have used out of habit. Same products, same period, enough sessions that the result is not noise. Then let the loser go without arguing with it. That is the whole value here. Being wrong in September costs a few hundred dollars. Being wrong in November costs the quarter.

Lock the Rule, Not the Calendar

The output of a good test is not a date grid. It is a rule. These categories are excluded. This is the depth floor. This is how long the window stays open. This is who sees an offer at all. A rule survives contact with a competitor's banner. A calendar of percentages does not, which is how stores end up deepening discounts in week two and calling it responsiveness.

This is what Growth Suite's A/B testing module is for. You run two offer variants side by side, set how much traffic each one gets, and watch the results update in real time. The September argument gets settled by your own shoppers instead of by whoever talks longest in the meeting. You choose what it scores on: conversion rate, average order value, or total revenue. Once the rule is set, scheduled campaigns with fixed dates and spend-based tiers let you commit Q4 in advance instead of improvising depth mid-quarter. And every variant reads against the August Control Group you already captured.

September is the last month of the year where being wrong about your discount is cheap. Spend it.

The Sentence You Want Ready in November

Last year's Q4 numbers record what happened in a market where everyone discounted at once. They cannot tell you what your discount caused. August can, because deadline shoppers and no-deadline shoppers are on your site the same week. That is the August Control Group, and it disappears the moment fall promotions start. Three measurements come out of it: what to exclude, how long the window stays open, where the depth floor sits.

So do one thing this week. Pull your August conversion rate on traffic that saw no promotion at all, and multiply it by your November session forecast. Look hard at what your planned discount costs before it earns anything. Then, in November, when somebody says "last year 25% worked," you will have the answer ready. Last year everyone was at 25%. We never found out what worked.

If you are about to set Q4 depth from last year's fully discounted numbers, Growth Suite helps you tell walk-away customers apart from dedicated buyers. It shows a real, expiring offer only to the visitors likely to leave without purchasing. So you keep the full-price orders that were already coming and still recover the ones that were not. It is free to install on the Shopify App Store, with a 14-day free trial.

Frequently Asked Questions