The holiday weekend is when every store discounts everything, and the only question shoppers ask is who went deepest. A store that joins the race pays for it twice: once in margin on the weekend and again in January, when nobody buys at full price. The alternative is not to sit it out. It is to make the offer personal, so it cannot be compared.
What a personal holiday offer looks like
A customer opens an email on the Tuesday before the weekend. It shows six products chosen for them, each at a price that is theirs, and an end date on Monday. The rest of the store is at full price. Their friend got a different six. There is nothing to screenshot and post, nothing to wait for, and no banner on the homepage that says the whole store is cheaper.
Everything below is how to set that up with the pieces Randio already has.
The strategy
Make a dedicated holiday strategy rather than editing the one you use all year; a run freezes the strategy it starts with, so the holiday one can be changed without touching the others.
- Guaranteed pool: the products people give as gifts, at a shallow fixed percentage. Everyone sees them, nobody mistakes them for a clearance.
- Chance pool: your main collections at a range, with Higher discounts on Uncommon or Rare. The visible ceiling makes the page feel like the season; the floor keeps the average where you can afford it.
- Limited pool: one hero item at a depth that would be reckless store-wide, capped at the number of customers you would actually sell it to at that price. This is the only deep number in the campaign, and it reaches a set number of people.
Keep the per-product limit at one unit. Holiday shoppers buy multiples; the extras are at full price, which is the point.
Stock
Sell-outs are the holiday failure mode. Set the stock minimum in Settings → Inventory high enough that a product leaves catalogs before the last units go, and set the rule to Replace so the slot refills from the Chance pools. Live catalogs update within about a minute, so a customer who opens the page late sees what is actually there.
Two campaigns, not one
Outreach to your best group first. Order your groups so recent buyers and high-value customers sit at the top, and run the holiday strategy to them with a manual window that starts before the weekend and ends after it. Turn the reminder on; the end date is the message. Because these people hold a live catalog, no other campaign emails them in the meantime.
Storefront for everyone else. Run the same strategy as a Storefront campaign for the same window. Logged In mode gives account holders a personal catalog; Guest mode serves visitors who arrive from ads and search a sticky catalog from a shared pool. The storefront block shows it where you placed it, and nobody has to have been on a list.
Both are manual runs for a reason: the dates are fixed, and the automatic rotation should not decide when a holiday ends. If you run automatic campaigns the rest of the year, their rotation picks up after the holiday window.
After the weekend
Open Analytics, set the scope to each holiday campaign, and read the Discount dependency chart. If orders cluster at the ceiling, the floor did nothing and next year's range can be narrower. If the Limited pool ran out on day one, the cap was too low or the group too big. The Products tab tells you which gifts moved at which depth, which is the start of next year's Guaranteed pool.
Then let the holiday strategy go. In January your recent buyers are a different group with a different offer, and nothing about the holiday run has changed what full price means.
Group order and definitions are in Segment before you randomize; the three jobs a campaign can do are in Three campaign strategies beyond "run a sale".