A back in stock email goes to someone who asked for one named product and left an address to get it. Nothing else you send carries that much intent. The ceiling on the whole flow is not your copy or your send time, it is how often you run out of something people wanted, and that is a number you can pull this afternoon.
What actually has to happen for it to fire
The trigger is the Subscribed to Back in Stock metric, written when someone submits the notify-me form on an out-of-stock product page. Those subscribers then sit in a back in stock delay, which is not a time delay. It holds them until your catalog tells Klaviyo the item is available again, however long that takes.
Two settings on the store side decide whether that moment ever arrives. Quantity tracking has to be on for the product, and continue selling when out of stock has to be off. Leave the second one on and the product is never out of stock as far as the catalog is concerned, so it never restocks, and a waiting list quietly grows behind a message that will not send. Nobody notices for months, because a flow with subscribers in it looks alive in the dashboard.
The tracking layer, and why this flow mostly escapes it
The notify-me form is an onsite script, the same layer that browse abandonment depends on and loses most of its audience to. This flow gets off much lighter. The subscribe is a deliberate submit that hands over an email address, so Klaviyo does not need to have recognised the visitor beforehand, and the identification problem that caps browse abandonment does not apply here at all.
What is left is narrower. The form has to render on the product template, and it has to post. Blocked scripts, a theme update that dropped the embed, or a form hard-coded to one variant will all produce the same silence as a broken Started Checkout, which is the failure we pulled apart in abandoned cart under-triggering. The variant case is the one that bites in practice: the form renders, the shopper subscribes, and the event carries the wrong variant, so the restock that matters never matches.
Ten minutes settles it. Pick one product you know sold out recently, take its product page views for that window in Shopify, and compare them against Subscribed to Back in Stock events for the same product in Klaviyo. A handful from a few thousand views is normal. Zero from a product you watched sell out is a form problem, not a demand problem.
Sizing it before you build it
Three numbers, last 90 days, all of them yours.
In your store admin, count the products that hit zero and came back. In Klaviyo, open Metrics → Subscribed to Back in Stock over the same window for the subscriber count. Then pick a conversion rate you are willing to defend for people who asked to be told, and multiply it by average order value.
Say 180 subscriptions in a quarter, a 5% conversion, a £48 order. That is £432 a quarter, and the arithmetic is worth doing with your own figures rather than mine, because the first number is the one that moves. A store that sold out of four products last quarter has a flow worth a few hundred pounds a year. A store where the best sellers are out of stock most weekends has one worth many times that. Published benchmarks for this flow are high and they tell you nothing about which of those two you are.
If you have not gone out of stock at all, the flow is worth nothing this quarter. Build it anyway, once, because the build cost is an hour and the demand it captures arrives without warning.
The two settings that decide what it earns
When 40 people are waiting and 12 units land, notifying all 40 at once sells 12 and annoys 28. Klaviyo handles this with a restock minimum, which holds the notification until enough units come back, and with batching, where you set how many subscribers to alert per restocked unit and how long to wait between batches.
Both fail in the same direction when they are set carelessly. A restock minimum set high means small replenishments never notify anyone, and the waiting list ages until the interest has gone. Set the minimum near the smallest restock you would actually want to tell someone about, and batch at two or three subscribers per unit so the first email goes to a group that can all buy.
The email that arrives after the product sold out again
Klaviyo does not check inventory again after the message has sent. If a small restock clears in an hour, the emails already out the door keep pointing at a page that is out of stock, and the fastest readers you have get punished for opening quickly.
A conditional split on current inventory before the send is the standard fix, and it matters more the faster your restocks clear. The same batching that protects the units protects this, by keeping most of the list unsent until there is something to sell them.
Where it belongs in the build order
Late, and that is not a criticism. Welcome, abandoned cart and post-purchase carry the volume, and the order to build flows in puts the supporting flows after the core ones for good reason. Back in stock earns its place because it runs on its own once built, on an audience that rebuilds itself every time you sell out.
That last part is what separates it from the flows that quietly stop working. A flow aimed at a fixed pool mails the same people until they are used up, and the size of that pool ends up mattering more than the copy does, which is the argument in your engaged segment is probably too small. A back in stock list cannot be exhausted, because every person on it put themselves there this week.
How many emails should a back in stock flow have?
One, then a second only if your restocks last. Send the first immediately on restock, because the entire value of the email is being early. A follow-up a day later to people who did not click makes sense when stock holds for days, and is pointless when it clears in an hour.
Three emails is a sequence built for a different situation. The person already told you what they want, so reminding them twice more is a stock availability question rather than a persuasion one.
Should a back in stock email include a discount?
No. This is the one email where the product is doing all the selling and the reader has already committed interest. A discount here pays margin on a sale you had, and it teaches a patient shopper that waiting produces a code.
Keep it short, lead with the product image, link straight to the product page, and say how much came back if the number is small.
Why did only some subscribers get notified?
Check the restock minimum first, then the batching. If the batch interval is set in days and the stock cleared on day one, everyone after the first batch is still waiting on a product that has gone, which looks identical to a flow that stopped working.
Then check whether the notification matched the variant the person subscribed to. A subscriber waiting on one size will not be released by a restock of another, and on a store with deep variants that accounts for most of the people who say they never heard back.
