Klaviyo Email Going to Spam: Isolate the Content Before BFCM

When one campaign email lands in spam while the others reach the inbox, the question is specific: is it the content, or something else? Klaviyo describes a test method to answer it, and reserves it for a particular case. This article goes through it step by step, with the limits the documentation puts on it.

TL;DR. Klaviyo says to use the content test only after ruling out all other factors, or when the problem affects a single message. The principle is to rebuild the email with a test send after each addition: stripped-down email, text paragraph by paragraph, links one by one, then images. You need a test account with the mailbox provider you are targeting. In our view, this test should be prepared before Black Friday week, not during it.

Want an outside view before you build? We offer a Q4 readiness audit: 30 minutes, free, no strings attached.

When to run this test, and when not to

The content test page is clear about timing: go through this process only when all other factors are ruled out, or when the spam placement affects one specific message and no others (Klaviyo, how to test your email content to identify spam triggers).

The troubleshooting page gives the general context. Mailbox providers sort with algorithms that are not made public. Klaviyo writes that they rely on several signals, including sender reputation, content and authentication, and that repeatedly sending to dormant subscribers can lower your reputation (Klaviyo, troubleshooting why emails go to spam). It lists, among the main metrics mailbox providers take into account, engagement, sending volume and frequency, bounces, spam complaints and unsubscribes.

In our view, a single message that lands in spam while the rest of your sends get through is the typical case where the content test is useful. If all your emails are affected, we would look at reputation first, which the troubleshooting page lists among the signals, before copywriting. Our article on the Deliverability tab shows where to read these indicators in Klaviyo.

The test, step by step

The method is to build the email iteratively, with a test send after each element you add, to see what causes the move to spam. It requires a test account with the mailbox provider you are testing.

1. The header. The page advises reading the email header: it shows the network path of the message and lets you see whether an authentication problem or an issue with the sender or recipient is involved.
2. A stripped-down email. If the header shows nothing, remove the text, links, images and ads, keeping the unsubscribe link and the required footer information, then send a test (Preview, test address, Send test). This is a starting point for comparing versions, and it does not isolate the HTML as the only possible cause. If it does not arrive or still lands in spam, the page asks you to revisit the other factors before continuing. Preview emails count toward the preview sending limit for the billing cycle, and their placement does not necessarily match that of a live send.
3. Text. If the stripped-down email reaches the inbox, add the content paragraph by paragraph, with a test send between each addition, to spot the words or phrases that trigger the filters.
4. Links. Once all the text has gone through without an issue, add the links one by one, with a test send after each. The goal is to see whether any URL is considered suspicious by the provider.
5. Images and ads. Add them one by one, with a test send after each addition, to make sure image repositories or third-party ads do not affect placement.
6. Remove the culprit and continue. When an element causes a problem, remove it from the following tests and carry on. Once all the content has been checked, send a final test of the complete email, with the problem content removed or revised.

The page also mentions that a third-party tool like Litmus can be used for this test: it lets you preview campaigns in different email clients and check for potential spam issues. To check authentication, you need to test an email sent from Klaviyo: an HTML export alone does not include the email headers. For rendering previews inside Klaviyo, see our article on testing and previewing before BFCM.

In our view, the test is long for a complex template, since it takes one send per element. That is a reason to do it on the base template, before pouring in the BFCM offers, rather than the day before a send.

What the troubleshooting page says about content

The same page lists content guidelines. They are Klaviyo's advice, with no promise of results: it states itself that spam filters are designed so that evading them cannot be perfected.

• Subject lines. Avoid all caps, strings of symbols and exclamation marks, and phrases such as "JUST THIS ONCE FOR A LIMITED TIME ONLY", "100% FREE!" or "ACT NOW!".
• Image and text balance. An email made only of GIFs and images, or very image-heavy, may trigger the filters. The page cites an Email on Acid finding that an email should contain at least 500 text characters, and recommends alt text for images. It also advises against including just one large image.
• Links. Keep URLs to the necessary links, especially to sites other than your own.
• Code. For Klaviyo's drag-and-drop editor, the page says Klaviyo takes care of this. With custom HTML, extra tags can trigger the filters. Text pasted from Word brings styling that weighs the code down: the page advises pasting without formatting with Ctrl + Shift + V (Windows) or Cmd + Shift + V (macOS).
• Personalization. The page says the more personalized an email is, the more likely email clients are to deem it important.
• Two unsubscribe links. One at the top and one at the bottom, which, according to the page, reduces spam complaints and helps if the email is clipped because it is too large.

The thresholds Klaviyo quotes

The troubleshooting page gives figures, introduced with "in general" for unsubscribes and volume: aim for a spam complaint rate, as reported in Klaviyo, under 0.01%, an unsubscribe rate under 0.3% per email campaign, and treat more than one email per inbox per day as risky. For bounces, the page says to aim for under 1%. It states that a high complaint rate calls for immediate action, and that Gmail does not report individual complaints to Klaviyo: a low rate in Klaviyo therefore does not confirm a low rate among Gmail recipients, for which it refers to Google Postmaster Tools.

These values come from this troubleshooting page. In our view, they work as reading guides: Klaviyo does not present them as thresholds set by a mailbox provider.

Authentication: rule it out before content

Before rebuilding an email, the troubleshooting page reminds you that an authentication failure (SPF, DKIM, DMARC) can lead to rejection or spam placement, and asks you to check that your DMARC policy is valid. It explains that to be DMARC compliant, you need a dedicated sending domain whose root domain matches that of the sender address. The choice between a branded and a shared domain is covered in our article on the sending domain.

Checklist before BFCM

These points are Deliver's advice.

1. The problem affects a specific message, or all other factors are ruled out.
2. A test account exists with each targeted mailbox provider.
3. The test starts with a read of the header, then a stripped-down email.
4. Each addition (text, link, image) gets its own test send.
5. The faulty elements are removed from the following tests.
6. The corrected template is saved before the BFCM offers go in.

FAQ

When should you use Klaviyo's content test?

The page says to use it only when all other factors are ruled out, or when the problem affects a specific message only.

What do you need to run this test?

A test account with the mailbox provider you are testing, and test sends between each content addition.

Can you use a third-party tool?

Klaviyo mentions Litmus, which lets you preview campaigns in different email clients and check for potential spam issues. For authentication, the page says to test an email sent from Klaviyo, since an HTML export alone does not include the email headers.

Is content the most common cause of spam placement?

The troubleshooting page does not rank causes. It lists several signals: sender reputation, content and authentication, and says that repeatedly sending to dormant subscribers can lower your reputation.

Want an outside view on your deliverability before November? We run Q4 readiness audits, or you can talk directly with the Deliver team.