Why month-over-month comparisons mislead
Most businesses do not sell the same amount every week. Retail peaks before the winter holidays. Garden centers peak in spring. Tax preparers peak before filing deadlines. Compare December with November and almost any retailer looks brilliant; compare January with December and the same campaigns look broken.
The size of the peak can be dramatic. Adobe reported that U.S. consumers spent $14.25 billion online on Cyber Monday 2025, and $44.2 billion over the five days from Thanksgiving to Cyber Monday. When demand jumps like that, conversion rates usually rise, and so does competition for ad space. A good cost per sale in late November tells you little about what to expect in February.
The fix is simple: compare each period with the same period a year earlier. Use rolling 7-day or 28-day totals rather than single days, so a busy Monday does not swing the whole picture.
Holidays move, and they differ by country
Year-over-year comparisons have their own trap: holidays do not fall on the same date each year. U.S. Thanksgiving is the fourth Thursday of November, so the shopping weekend after it lands on different dates. Easter can fall in March or April. Lunar New Year, Ramadan and Diwali follow other calendars. Compare November 2026 with November 2025 by date and you may be comparing a holiday week with an ordinary one.
If you advertise in several countries, each has its own calendar. A sales dip in Germany on a public holiday is not a campaign problem. Keep a simple list of key dates for each market and mark them on your reports.
- Align comparisons by event ("the week before Thanksgiving") as well as by date.
- Mark national holidays and big shopping events for each country you sell in.
- Watch the days before a holiday, when shipping cut-offs can end online sales early.
- Remember that a time zone difference can push late-night orders into the next day.
- Thanksgiving$6.4
- Black Friday$11.8
- Cyber Monday$14.25
Each bar is total U.S. online spending that day in billions of dollars, as measured by Adobe; demand on peak days is far above a normal day, which is why holiday weeks should be compared with the same weeks last year.
Source: Adobe
Telling ad platforms about short events
Automated bidding learns from recent conversion rates, so a sudden flash sale can confuse it. Google Ads offers seasonality adjustments, which Google describes as a way to inform Smart Bidding of expected changes in conversion rates for future events like promotions or sales. Google says they are ideal for short events of 1 to 7 days. For longer seasons, the bidding usually adapts on its own, so check before overriding it.
Currency: convert once, the same way, every day
If you pay for ads in one currency and sell in another, exchange rates can make results look better or worse without a single customer behaving differently. Say you spend in U.S. dollars and sell in euros. If the euro strengthens 5% against the dollar, the same European sales are worth 5% more in dollars, and your return on ad spend rises with no change in the campaign.
Two rules keep this honest. First, pick one source of exchange rates and one rate per day. The European Central Bank publishes euro reference rates that are usually updated around 16:00 CET every working day, for information purposes only. Second, convert spend and sales with the rate for the day each happened, not with today's rate applied to the whole year.
When comparing years, it can help to look at results in local currency as well. If European sales grew 10% in euros, that is the real business growth; whatever the dollar figure adds or subtracts is the exchange rate. Accurate conversion matters most when you reconcile ad data with your Stripe or store sales, because a small daily rate error adds up across a month.
Putting it together
Read every international result through three filters: is this a seasonal pattern, is a holiday distorting it, and is currency moving it? Only what is left over is the effect of your marketing. For a cleaner read on what ads really caused, a holdout test run outside peak season is hard to beat, and the same thinking applies when you split your budget across channels.
Matrix ROI, a Smith App Studio product, converts every amount into one currency using European Central Bank reference rates and shows platform claims beside verified sales day by day, which makes these comparisons easier to see. A 14-day trial needs no card.