Why holiday weeks produce false alarms
The Monday after Thanksgiving, someone opens the store-by-store report and finds a dozen locations down hard on the week before. A few are flagged red. Somebody wants to know what went wrong at those stores.
Often nothing did. Holiday weeks change the base the report is built on. A store that closed for the holiday, ran short hours on the Wednesday before, or took one huge catering order the week prior will look like it fell off a cliff. The demand didn't move the way the chart says. The denominator did.
Thanksgiving 2026 is Thursday, November 26, so the week of November 23 is the one that will produce these reports. Here are three checks to run before anyone ranks a store, plus a log that makes next year easier. This is our practice framing, not a benchmark. We're not going to tell you how big a typical holiday dip is, because it depends on your stores, your menu and your guests.
Check 1: compare to the right week
Week-over-week is the default view in most reports, and in a holiday week it's the wrong one. The week before Thanksgiving is a normal week. Comparing a holiday week to a normal week tells you the holiday happened, which you already knew.
Compare against last year's holiday week instead, and align it by the holiday, not the calendar date. Thanksgiving fell on Thursday, November 27 in 2025. The fair comparison for the week of November 23, 2026 is the week of November 24, 2025, so each Thursday lines up with a Thursday and each closure day lines up with a closure day.
The same rule matters later in the season. Christmas Day falls on a Friday in 2026 and fell on a Thursday in 2025. Line up December 25 with December 25 and you're comparing different weekdays, with different closures and a different shape to the week. Line up the holiday and the days around it, then compare.
- Use the same stores in both years. A store that opened since last year has no fair comparison yet.
- Use the same days. If last year's report ran Monday to Sunday, keep it that way.
- Write down which week you compared against, on the report itself, so nobody reads it as week-over-week.
Check 2: orders per open hour
Total orders punish a store for being closed. That's fine when you want to know total orders, and misleading when you want to know how the store performed while it was open.
Build an hours file: one row per store per day, with the hours it was actually open. Not the posted hours, the real ones, including the early close on the day before the holiday and the store that shut for a broken freezer. Then divide orders by open hours and rank on that.
Orders per open hour helps when the question is "did this store do well with the hours it had?" It misleads when hours were cut because demand was low, since the ratio can then look healthy on a quiet day. It also flatters a store that opened only for its busiest hours. Read it next to total orders, not instead of them.
The hours file is also the hardest part of this. Most brands don't have one in a single place. If yours lives in a scheduling tool, a franchise portal and a few emails, that's worth fixing before the next holiday, and one location key across every system is what lets the hours file join to everything else.
Check 3: pull out catering and large orders
Holiday weeks bring office parties and family gatherings, and one big catering order can lift a single store's week on its own. Rank stores with that order in and the store looks like a star. Next week, without it, the same store looks like it collapsed.
Set a size threshold for what counts as a large order. Pick it yourself, based on what's normal for your menu, and keep it the same all season. Then rank stores twice: once with every order, and once with catering and large orders removed. If a store's position changes a lot between the two, the ranking was about one order, not about the store.
If catering comes through a separate form or another platform, it may not be in the same export as everyday orders at all. The Catering Order Nobody Counts covers finding where those orders end up.
Keep an annotation log
The three checks fix this year's report. The annotation log fixes next year's. Keep it next to the report, in the same place people already look, not in someone's inbox.
Record one line per event, per store:
- Closures, full or partial, with the date and the hours lost.
- Changes to posted hours, and when they changed back.
- Outages: ordering site down, payment problems, a delivery partner pausing a store.
- Weather that closed roads or kept people in.
- Local events that pulled traffic toward or away from a store.
- Large orders that you excluded under check 3.
Next November, the person running the comparison can read why a store looked strange last year without having to remember it. That person may not be you.
What this does not fix
These checks give you a fairer read. They don't give you a forecast, and they don't account for everything.
- Stores that opened, closed or were remodeled since last year don't have a clean prior-year week. Leave them out of the ranking or show them separately.
- Menu and price changes since last year change what an order is worth. Order counts may be steadier than sales for the same reason.
- Weather varies by region, so a chain-wide comparison can hide a storm that hit one market.
- Some holiday weeks also move inside your own tools. If a store was listed on a new channel or dropped one since last year, its total shifts for reasons that aren't demand.
If a store still looks bad after all three checks, that's a real signal worth a phone call. The point of the checks is to make sure the phone calls go to the right stores. Store-level reporting also depends on where each order came from, and tagging each Google Business Profile listing by location is one way to see which listings actually send orders to which store.
A four-question self-check
- Does your holiday report compare to last year's holiday week, aligned by the holiday?
- Do you have a file of the hours each store was actually open?
- Can you rank stores with and without catering and large orders?
- Is there an annotation log next to the report, and does anyone fill it in?
If any answer is no, that's where to start before the next holiday report goes out. If you want help getting store-level reporting to hold up, book a data and tracking audit. If you also want to see what your restaurant site loads before a guest makes a privacy choice, request a free privacy scan. We review each request within 48 hours.