The 49% YoY Drop That Wasn’t a Drop
A DTC brand we run shows a 49% YoY net sales decline for April 1–13. Orders down 57%. Items purchased down 61%. Every dashboard glows red.
The business is actually growing.
I want to explain why this happens, what it means for how you read performance data, and why an agency that reports the first number they see is more dangerous than one that reports nothing at all.
What the numbers said
We pulled the April 1–13 period because that’s where we are in the calendar. Thirteen days into April 2026, every headline metric is dramatically negative compared to the same window in 2025.
49% drop in net sales. 57% fewer orders. 61% fewer items purchased. If you handed this to a CFO without context, they’d be on the phone before lunch.
But there’s a single fact that explains all of it.
In 2025, this brand ran their biggest promotional event of the year — a major product sale — from April 1 through April 5. That promotion drove an outsized concentration of orders in the first week of April. In 2026, the same promotion shifted to April 24 through April 27.
You are comparing a window that included the highest-volume promotional week of the year against a window that doesn’t. Of course sales are down 49%. You’re not comparing like for like. You’re not measuring anything real.
The problem with first-number reporting
The most dangerous thing an agency can do is report the first number they see.
Not because they’re lying. Because they’re not asking whether the number means anything. And in a world where dashboards refresh in real time and every metric has a YoY comparison baked in, the first number is almost never the right number. It’s just the loudest one.
Promo calendar shifts are common. They’re not exotic or unusual. Any e-commerce brand running seasonal promotions is going to see this. The brand that ran a Valentine’s Day sale in week two of February 2025, then moved it to week one in 2026, will show an apparent uplift in the first week of February and an apparent decline in the second. Neither is real. Both will show up on the dashboard.
If your reporting process doesn’t account for this, you’re going to make decisions based on noise.
What the number actually tells you — and what it doesn’t
Here’s the distinction that matters.
The YoY net sales comparison for April 1–13 tells you: this brand sold less in this period this year than in the same period last year. That is true and accurate. It is also meaningless for the purpose of diagnosing business health.
What you actually want to know, and the question the number cannot answer, is: is this business growing?
To answer that, you need to wait until April 30 and compare the full month. Once the April 24–27 promotion runs, you’ll have a comparable picture. You’ll be looking at a month that included a major promotional event in both 2025 and 2026. The comparison becomes valid. The trend becomes readable.
You don’t make decisions about the health of an e-commerce business in the first thirteen days of a promotional month. You make decisions when the month is complete and the promo has run. Anything before that is a partial picture being mistaken for a complete one.
If you want to know whether the ads are working
The gross sales decline tells you nothing about paid media performance. Nothing.
Paid media doesn’t create promotional events. It amplifies them. If the promotional event hasn’t run yet, you cannot evaluate the paid media’s contribution to the period by looking at gross sales. The two things are not measuring the same thing.
If you want to know whether the ads are working, you look at three metrics.
ROAS. Return on ad spend. What is the revenue returned for every rand or dollar spent on ads? If ROAS is holding stable or improving, the ads are performing. If ROAS is deteriorating, you have a media efficiency problem. Gross sales decline tells you nothing about which of those is true.
CPR. Cost per result — purchase, lead, or whatever conversion event matters for this brand. Is it going up or down? Is it within the acceptable range for this channel at this time of year? Outside of promotional periods, CPR gives you the clearest signal about whether the media buying and creative are working.
LTV. Lifetime value of acquired customers. Not what a customer spends in the first order — what they spend across the relationship. A promotional period that acquires customers at a higher LTV than your off-promo periods is a good promotional period, even if ROAS looks compressed in the short window.
None of these metrics are hiding in the April 1–13 dashboard. They require pulling data that isn’t in the headline view. They require actually thinking about what you’re measuring before you report it.
That’s the job.
Why agencies report the wrong number
I’m not being cruel about this. I understand why it happens.
Agencies report what’s visible. Dashboards are built to show YoY comparisons because YoY comparisons are useful — most of the time. The problem isn’t the comparison. The problem is applying it mechanically without asking whether the comparison is valid for this specific period.
An agency that’s managing twelve or twenty clients cannot do this analysis manually for every client every week. There isn’t enough time. So they build reporting templates, export data into those templates, and present what the template shows. If the template shows a 49% decline, that’s what the report says.
The alternative — actually interrogating whether the comparison is valid before presenting it — requires knowing the client’s promo calendar well enough to flag anomalies before they appear in the data. It requires a process for checking promotional period alignment before drawing any YoY conclusions. It requires someone on the account who understands the business, not just the dashboard.
That’s the difference between agencies that report data and agencies that interpret it.
What we did
We didn’t panic. We didn’t send an urgent message to the client about a 49% decline. We looked at the promo calendar, identified the shift, and noted that the April 24–27 period would be the actual comparison point.
We told the client: here’s what the dashboard shows, here’s why it looks this way, here’s what to watch for at end of month, here’s what we’re monitoring in the meantime.
That conversation took about three minutes. It saved the client from making a bad decision — pulling budget, changing creative, restructuring a strategy that’s actually working — based on a number that wasn’t telling them what they thought it was.
The full-month comparison will tell us whether the business is growing. The ROAS, CPR, and LTV metrics will tell us whether the media is working. The April 1–13 gross sales number tells us that it’s mid-April and the promotion hasn’t happened yet.
The principle
Here’s the rule we apply before reporting any metric.
Ask: is this period comparable to the period we’re comparing it to? If there was a promotional event in one period but not the other — or the promotional event shifted dates — the comparison is not valid. Flag it. Wait for the comparable window. Report when the data means something.
Ask: what question is this metric answering? Gross sales answers “how much did we sell?” It does not answer “is the business healthy?” or “are the ads working?” Use the metric that answers the question you’re actually trying to answer.
Ask: what decision would someone make based on this number? If the decision would be wrong — if acting on this number would lead to pulling budget from a campaign that’s performing, or restructuring a strategy that’s working — don’t present it as a signal. Present it as context and point to the number that actually matters.
Reporting the first number you see isn’t neutral. It’s a choice to let the dashboard make decisions for you.
The numbers are not self-interpreting. That’s what you’re there for.
