The content you create for your customers—you want it to be engaging. Part of what makes content engaging includes showing people images they can personally identify with—people who help them envision the best versions of themselves, products that interest them, and peer reviews from voices that resonate with them.
One way you can create engaging content people can identify with: try a gender-targeted marketing strategy.
Earlier this year, Klaviyo released a new gender prediction feature that classifies profiles as “likely male,” “likely female,” or “uncertain,” which you can use when you create segments.
Let’s take a look at how some ecommerce marketers have used gender prediction as part of their email marketing strategy and explore the lift in engagement they’ve seen as a result. I’ll also share some tips and advice on how to create segments for gender-targeted marketing and show you how to set them up in your account.
How gender-targeted marketing improves clickthrough rates
To see the effect different gender-targeted email marketing strategies have, Klaviyo’s data science team worked with several customers to run some experiments.
To see if sending relevant content based on a recipient’s predicted gender provided a lift in email engagement, we sent both gender variations of the content to each gender using an A/B test. We then compared the clickthrough rates for those in the “likely male” segment who received the male variation of the message to the clickthrough rates for those in the “likely male” segment who received the female version of the message.
In our first experiment, we worked with Lucchese Bootmaker, a Texas-based bootmaker that’s been celebrating craftsmanship and an adventurous spirit since 1883. Lucchese’s experiment included both a varied hero image and a different order of products based on the predicted gender of the recipient.
In Lucchese’s email, the brand highlights two products: the “Roy,” a men’s boot and the “Gaby,” a women’s boot. With the male variation of their email, a man wears the Roy boot in the hero image and Roy boots are shown ahead of the Gaby boots below the hero image. In the female version of the email, a woman wears the Gaby boots in the hero image and then Gaby boots are shown below the hero image ahead of the Roy boots.
We saw a large boost in clickthrough rates from targeting customers with matched gender content.
|Likely male recipients||Likely female recipients||Uncertain recipients|
|Lift from matching gender in email to recipients||+30%||+24%||+13% for sending male variation|
But not all of our experiments yielded massive lift. If the difference between the content shown in the email variations is too small, it didn’t have any effect on how recipients interacted with the email.
For example, when working with Coyuchi, an innovative manufacturer of organic and sustainable linens, their first gender targeting experiment showed only a small change between the two variations of their emails, which were very similar. The only difference between variations was the presence of a man in the hero image for their male variation compared to no individual in the female variation of the email.
The results showed there wasn’t much of a difference in clickthrough rates between the two variations. The version that included the man in the hero image was clicked at a slightly higher rate for male recipients, but overall both variations performed almost identically for both genders.
Despite these results, Coyuchi was undeterred and had another idea for a campaign that could drive engagement. Their team designed an email highlighting three linen products. For the different predicted genders of recipients, they showed different colors: blues and greys for men and pastels for women. For this campaign, they saw a massive increase in clickthrough rates with the color variation strategy.
3 gender-targeted marketing recommendations
Gender-targeted marketing can be useful to brands who should use it primarily to show customers different products and images that create a personal, relevant, and compelling story about why they should buy your product.
One essential thing to bear in mind with gender-targeted marketing: it’s important to respect your customers and create gender-targeted content that doesn’t exclude customers based on inaccurate gender prediction.
Here are three recommendations:
- If your customer has already identified as a gender through a signup form, respect that gender when creating gender segments. In the next section, we’ll show you how to make segments using both predictive gender and any custom gender properties you may have already collected in your account.
- Reorder products to show male styles ahead of female styles for your men’s variation of your emails and vice versa instead of showing only male or only female products in your gender variations. Show products relevant to both genders in case a customer’s predicted gender is inaccurate.
- Show more male or female imagery in hero images, lifestyle content, and product reviews, but don’t exclusively show only one gender.
How to set up gender-based segments in Klaviyo
Setting up your gender-based segments is an advanced use case of Klaviyo’s segmentation engine, but you can use both predicted gender and any custom gender properties you may have to create segments that use both information customers have provided to and our predictive analytics features to classify your audience.
Setting up gender-based segments starts with your engaged list. Even though you’ll be personalizing campaigns to customers to create a more engaging experience, you’ll still want to send to your engaged audience to make sure you retain good deliverability.
A gender segment has three components. First, it has conditions to find your engaged profiles. Then, it has conditions to find male or female customers through predictive gender and custom properties. Finally, it has a section that excludes customers who have already identified as the opposite gender through a custom property.
Let’s start by showing a segment for a female audience:
- In this segment, we first include people who have engaged with marketing content in the last 30 days.
- Then, using an AND statement, we narrow it down to customers who we suspect are female. Inside this statement, we use an OR statement to state the two ways a customer could be female: they could either be predicted to be female or they could have identified as female on a signup form.
- Finally, we use another AND statement to specify that we only want customers where their signup form gender isn’t male. This excludes anyone who may have already told you they were male. This setup ensures that your custom property gender always overrides our gender prediction.
If you have more than one custom gender property in your account, you’ll need to use all of those properties to accurately classify your customers.
In the second statement, you’ll need to add all the different ways a customer could be female with OR statements. Then, for each custom property, you’ll need to add an AND statement to exclude customers identified as male with that custom property. Here’s an example of that:
To build the male segment, you would build the same segment except you’d simply swap the genders in each condition.
Finally, you’ll want to create the segment of your audience with uncertain gender. For this segment, you’ll want to find all people who are classified as uncertain gender and all people who haven’t been identified by one of your custom properties:
One thing to be aware of when creating these segments: the only people who aren’t included in one of the three gender segments are people who have conflicting gender in custom properties. This could happen if you collect gender in two different ways on your website. If a person selects “female” on one popup form and answers “male” on a survey, for example, then their custom properties might be different. Since they’ve told you two different things, they’ll be excluded from all the segments we created above. This case should be rare, but it is still possible.
To send an email to your gender-segmented audience, you’ll want to set up two campaigns, each with the gender-specific content you want to deliver. Include the uncertain gender segment in one of the audiences to make sure these active profiles get one version of your email. We recommend that you group your uncertain audience with your larger gender segment since that’s likely a better match for these profiles.
Creating the most relevant content possible for your customers is a critical component of your customer engagement strategy. Gender-targeted marketing can help you create highly relevant, personalized content that engages your customers with images, product recommendations, and content that resonates with them and helps you add another layer of personalization to your marketing strategy.
Curious about how Klaviyo’s gender prediction algorithm works? Learn more about it and how you can use gender-targeted marketing as part of your email strategy.Back to Blog Home