In a nutshell

  • Start With the Flow: Read it from delivery to clicks to orders. You can have plenty of opens and still end up with very few purchases.
  • Look Closely at Email 1: Does it deliver on the signup promise? Is the offer right for the person who signed up? Does the link take them where they need to go? And is the delay making sense? This is usually where most of the welcome revenue comes from.
  • Check Who’s Entering: If past buyers or Shopify checkout opt-ins are getting into a new-subscriber flow, that can change both engagement and revenue.
  • Keep a Few Possibilities Open: Don’t settle on one explanation too early. Keep checking until the account gives you enough evidence.
  • In This Case: The biggest revenue changes came from the flow filters, the offer match, and where the first link sent people, rather than simply offering a bigger discount.

How to read “not converting”

When someone says a welcome flow is “not converting,” start with three things: did the email arrive, did people click, and did anyone buy?

A flow can look fine on the first two and still bring in very few orders. When that happens, it’s worth slowing down for a moment, because there are a few different reasons behind it.

Two other numbers can help make the picture clearer.

One thing I always check here is the denominator, meaning the number on the bottom of the calculation. In Klaviyo, placed-order rate is the number of orders attributed to the message within the attribution window, divided by the number of unique recipients.

That’s different from asking, “What percentage of new subscribers bought within 14 days?” Both numbers can be useful, but they’re answering different questions, so keep them separate when you compare results.

Then look at where the revenue is coming from. If Email 1 brings in almost all of it, take a closer look at what Emails 2 and 3 are doing. If every email gets opened but almost nobody buys, check the offer, who’s receiving it, where the link takes them, and what happens once they land on the site.

Pro tip: I’d pull the revenue by email before rewriting Email 2 and Email 3. If Email 1 is doing most of the selling, the later emails may need a different job rather than simply pushing the same offer harder.

Where a welcome flow can thin out

There are a few places to check. One might explain the drop-off, or you could find a few things happening at the same time.

The Form and Email Say Different Things

Say the popup promises 15% off your first order, but Email 1 gives you 10%. Or maybe the code is there, but it’s buried under a brand story.

If someone signed up for the discount, that’s what they’re expecting to see when they open the email. When the email leads with something else, you can lose that person before they even get to the offer.

The Offer Arrives Late

Another thing to check is timing. Maybe Email 1 starts with the brand story and the discount code doesn’t come until Email 2. Or Email 1 takes a few hours to arrive.

If someone signed up to get a code, a few hours is a long time to wait.

A delay of a few minutes is usually fine. Hours of waiting, or having no code in Email 1, is worth checking. Klaviyo’s welcome-series guide recommends sending Email 1 immediately after an Added to List trigger.

So if Email 1 feels late, I’d check the trigger first. With an Added to List trigger, the flow should send almost immediately. A segment-triggered flow can take 15 to 60 minutes while the segment updates. And with double opt-in, the clock starts when the person confirms their subscription, not when they submit the form.

Past buyers are in the welcome path

The trigger is Added to List, but there’s no filter to keep people who have already ordered out. So someone who’s bought before can get a first-order welcome email. They might ignore it. Or they might use the new-subscriber code on an order they were already planning to place. Either way, the numbers you see will be different.

There’s another thing to check here. Shopify checkout opt-ins often go onto the same newsletter list, so someone can subscribe and buy in the same session, then get dropped into a flow written for a first-time visitor.

Two filters help here. The first keeps past buyers out by looking for people who have placed zero orders over all time. The second checks whether someone has placed an order since entering the flow and stops the remaining emails if they have. Klaviyo covers both in its welcome-series guide and flow-filter guide.

The click goes to a general page

The button says “Shop Now,” but it sends people to the homepage. The logo does the same thing. So someone who just signed up has to figure out where to go from there and what to buy.

The button is hard to reach on a phone

A big hero image can push the code and button below the first screen. People still open the email because the subject line caught their attention, but the next step is easy to miss on a phone.

The page after the click is slow or confusing

The email can do everything right and still lose the sale after the click. Maybe the collection page loads slowly on mobile. Maybe the code is hard to find. Or the products on the page don’t match what the email showed.

This is worth checking on the site itself (I always want to see what happens after the click), because the flow report won’t tell you what happened once someone left the email.

Gmail engagement looks weaker over time

You may also see more welcome emails landing in Promotions as people stop interacting with them. That can happen when the wrong people enter the flow, or when the email doesn’t match what they expected when they signed up.

Promotions is still an inbox placement. So check this after looking at who entered the flow and what they were expecting to receive.

Case file Fashion brand on Klaviyo

Case study: Applying Email DDx to a Welcome Flow with Strong Opens, No Sales

1

Presenting Symptom

A fashion brand’s welcome flow was getting a 65% open rate and 4% click rate, but only 0.4% of emails led to a purchase, well below the 1.97% benchmark.

They increased the discount from 10% to 20%, but orders barely moved. That told us the discount probably wasn’t the main issue.

2

Possible Root Causes & Evidence Gathering

We listed out the possible causes, then looked at the evidence for each one and rated how strong the evidence was. That helped us avoid jumping to conclusions.

Possible Root cause Evidence Signal Tier
Offer mismatch (popup 15% vs email 10%) Side-by-side content check Strong Confirm
Past buyers entering welcome flow Export: ~25% of “new” subscribers had prior orders Strong Confirm
CTA points to homepage Link Activity: most clicks go to homepage Strong Confirm
Discount code timing/visibility No direct data Indeterminate
Mobile layout (CTA below fold) 60% mobile opens, but no scroll map Weak Confirm
Gmail Promotions tab Lower Gmail opens/clicks, but confounded Weak Confirm
3

Prioritized Root Causes

We focused on the issues where the evidence pointed to a problem. For anything we couldn’t confirm yet, we left it alone for now.

  1. Past buyers in the flow: They were included in the numbers even though they had already bought something.
  2. Offer mismatch: The discount in the email didn’t match the offer people saw in the popup, which could make the offer less convincing.
  3. Homepage CTA: Sending people to the homepage added an extra step after they clicked.
4

Treatment Plan

We made one change at a time and watched the results before moving on. This helped us see which change was making a difference.

  • Fix 1: Excluded anyone with a previous order from entering the flow.
  • Fix 2: Changed the email discount to 15% to match the popup.
  • Fix 3: Sent people to a collection page instead of the homepage.
5

Results with Clean Attribution

Because we made the changes one at a time, we could see what happened after each fix.

  • After Fix 1: Conversion rose from 0.4% to 1.2% after removing past buyers from the flow.
  • After Fix 2: Click rate rose from 4% to 6.5%, and conversion reached 1.9%.
  • After Fix 3: Click rate reached 9.2%, and conversion climbed to 2.8%.

Revenue per recipient increased several times, and spam complaints dropped too.

6

My thought on this

A high open rate can make an email campaign look healthy at first, but it doesn’t tell you what happens after someone opens the email. People can open and click your emails and still not buy.

So when sales are down, I’d look at what happens after the open. Are people clicking? Where are they going? And what happens once they get there?

That’s how we approach Email DDx. We look at the possible causes, check what the evidence tells us, and focus on the causes we can confirm. Then we can see what actually helped, instead of changing five things at once and having no idea which one made the difference.

What to look at in your Klaviyo welcome flow

Use this as a map while you gather evidence. Your account may point to a different mix of issues, so work through the list and follow what the data shows.

  1. Flow Filters and Recipients: Check who can enter the flow. Can past buyers still get these emails? Add an entry filter to exclude anyone with an order history, then use an in-flow filter to stop the series after a purchase. In Recipient Activity, check who was skipped and why. To compare signups with first purchases, build a segment or export the data and compare the list-add date with the first Placed Order.
  2. How People Joined the List: Check where subscribers came from. If checkout opt-ins and popup signups go into the same list, recent buyers can end up in a first-order welcome flow.
  3. Email 1 Delay and Trigger Type: For a list-triggered flow, Email 1 should send immediately, or within a few minutes. If it’s delayed, check the trigger type, double opt-in settings, and whether Email 1 contains the promised discount code.
  4. Signup Form vs. Email 1: Compare the signup form with the first email. Does the discount match? What about the deadline and first-order wording? Any mismatch is worth checking.
  5. Click Destination: Use Link Activity to see where the main button and logo send people. Check whether they go to the homepage, a collection, or a product page.
  6. Mobile Layout: Open Email 1 on a phone. Check how far you have to scroll to find the discount and the main button.
  7. Message-Level Revenue: Compare Email 1 with the emails that follow. If almost all the welcome revenue comes from Email 1, the later emails need to be looked at separately.
  8. Skipped Profiles: Check the skipped reasons in Recipient Activity. Someone who never received the email should be separated from someone who received it but didn’t buy.
  9. The Page After the Click: Check page load time, add-to-cart activity, and whether the discount field is easy to find.

Creative tests can still help, but it’s easier to interpret them once you know what the account is pointing to.

Start here tomorrow morning

Open Recipient Activity and check who entered the flow, who was skipped, and who had already placed an order. Then put the popup and Email 1 side by side and compare the offer. After that, open Link Activity and see where people went after they clicked.

That gives you a quick first read without having to rebuild the whole flow. If you want to take the same approach across the full account, that’s the diagnosis we run.

Sources

FAQ

Why do people open my welcome email and skip the click?

If people are opening the email but fewer than 2% are clicking, there are a few things I’d look at first. The offer or CTA might be sitting too far down the email, or the button could be buried under a big image. It could also be that some of these people are already customers, so a welcome offer doesn’t really give them much reason to click.

I’d start by checking the layout of Email 1, especially what people see on the first screen on mobile, and whether past buyers are still being included in the flow.

What is a decent conversion rate for a Klaviyo welcome flow?

It varies by offer, average order value, and traffic quality. Klaviyo’s 2025 Benchmark Report (US) puts many welcome flows near a 2% placed-order rate, with the top 10% closer to 10%. A more useful read is the trend plus revenue per recipient, split by email.

Start by comparing Email 1 with the rest of the series, then look at whether past buyers are getting mixed in with your new subscribers. And one thing I’d keep separate is the conversion rate of each individual email versus the percentage of subscribers who end up buying within two weeks. Those numbers answer two different questions.

How long should a Klaviyo welcome flow be?

Many stores run 2 to 3 emails. Email 1 delivers the promise. Email 2, a day or two later, shows products. Email 3 adds reviews or proof. Klaviyo’s starter welcome series uses three emails over a week, with Email 1 sent immediately.

Some stores do run longer series, but I’d figure out the length after Email 1 is doing its job properly. Once that first email is clear and working, it’s much easier to decide whether you actually need more emails.

Can existing customers in the welcome flow affect conversion?

They can. If someone has bought from you before, they might just ignore the welcome email. Or they might use the new-subscriber discount on something they were already planning to buy anyway.

A filter that excludes purchase history keeps the welcome path pointed at first-time subscribers. A second filter should drop anyone who buys after they enter. Returning customers can then get a separate series that makes more sense for them.

Should I raise the welcome discount if orders are low?

You can test that, but in this case, the brand went from 10% to 20% and saw little change while other confirmed issues were still in the flow.

I’d check the form promise, who’s entering, where the link goes, and whether Email 1 actually delivers the code alongside the discount test.

Does Gmail Promotions placement explain a low welcome conversion rate?

It could be part of the picture, especially if Gmail opens are weaker than other providers. Engagement can affect where emails land.

But if the email is showing up in Promotions, that’s still the inbox - it’s not the same as going to spam. And if we’ve already confirmed that past buyers are getting the wrong offer, I’d look at that first.