What Champions are doing differently with segmentation, send strategy, and list health

Key lessons from our first Klaviyo Community Office Hours session

For our inaugural Klaviyo Community Office Hours session, digital marketers brought real-world questions about email and SMS segmentation, engagement, testing, and seasonal planning. Champions Eve Karon, Director of Retention at Boarderie, and Tabish Bhimani, Founder of Mastrat Digital, shared practical advice grounded in the realities of growing lists, changing inbox behavior, and balancing short-term performance with long-term customer value.

Here are the biggest takeaways from the conversation.

1. Let your data inform the calendar, but take a step back and use human judgment, too

The strongest send calendar starts with the customer, not a generic benchmark.

Eve recommended working backward from key events and holidays to map announcement timing, urgency, and reminders. From there, use Klaviyo’s Claude MCP integration to analyze historical performance by day of week and time of day.

But historical data needs a sanity check. If you have never sent on Saturdays, Saturday performance will naturally look weak. To build a more complete picture, consider splitting your audience into sevenths and sending comparable content across each day of the week.

Tabish added that timing should reflect when your audience is most likely to act. For a mall, that might mean Thursday or Friday evenings and weekends, when foot traffic and weekend planning are highest. Event reminders can also work well closer to the date—including the day before.

Klaviyo’s smart send time feature can help, but treat the results as directional. Profiles without an IP address may default to 12 a.m. in the account time zone, which can skew the data.

2. Look beyond click rates as your list grows

Declining click-to-open rates are a widespread industry challenge—and they do not automatically mean your program is getting worse.

As a list grows, newer and lower-intent subscribers are added. That can bring down average engagement rates even as the program generates more total revenue. AI-powered inbox previews are also changing how people interact with email, making formal opens and clicks less complete indicators of attention.

Both champions encouraged marketers to track revenue per recipient alongside open and click rates. The better question is not simply, “Did engagement rate go up?” but “Did this send create more value without damaging list health or deliverability?”

To understand performance in context, segment subscribers by purchase frequency—0 purchases, 1 purchase, or 2+ purchases—and layer in engagement recency such as 30-, 60-, and 90-day cohorts. This makes it easier to see which audiences are driving revenue and where send breadth should be adjusted.

3. Segment by who should be excluded, too

Effective targeting is not only about identifying the right people to include. It is also about avoiding irrelevant sends.

For product-specific campaigns, exclude customers who have already subscribed to or purchased that product. You can also layer in abandoned cart or checkout audiences from the last 90 days to reach people who have already shown interest.

The audience breakdown tab in Klaviyo campaigns can help you understand how product-affinity segments perform within a broader send. Post-purchase survey emails are another opportunity to collect preference data and improve future targeting.

For long-inactive subscribers, assess whether they represent a re-engagement opportunity or dead weight. A smaller, more engaged audience is often healthier than repeatedly sending to people who have not clicked or purchased in six to twelve months.

4. Use progressive profiling to collect better data

Sign-up forms should be easy to complete. Once someone joins your list, email can do the work of gathering richer preference and demographic data.

Tabish recommended sending a simple follow-up email with a prompt such as, “Tell us what you’re most interested in.” With Klaviyo’s custom property links, each click can assign a profile property—without requiring the subscriber to complete another form.

This approach can help identify interests such as:

• Events versus general updates
• Families or specific audience groups
• Tenant or product categories
• Genre or artist preferences

Preference collection also performs better when it is framed as a personalization benefit. Instead of asking subscribers to opt out of sensitive topics, explain that their choices will make their inbox more relevant. Bordery used a content-preference checklist around Father’s Day to help subscribers choose the themes they wanted to avoid while capturing useful segmentation data.

5. Protect deliverability with smarter suppression options

Not every disengaged subscriber wants to leave forever. A “snooze for 30 days” option in an email footer can give people a temporary break without forcing a permanent unsubscribe.

Custom property links can also support this workflow. Other low-friction feedback mechanisms—such as thumbs up or thumbs down—can help you learn which content resonates over time.

For subscribers who have been inactive for a long period, consider limiting sends to roughly once a month and using those messages as re-engagement opportunities. This keeps the relationship alive without making every campaign feel like noise.

When a campaign series includes multiple messages, use the first email as an engagement filter. Send follow-ups primarily to people who clicked or otherwise showed interest in the first touchpoint.

6. Treat testing as a measurement problem

A/B tests are only useful when the results are statistically meaningful. That requires enough volume and a testing window short enough to avoid seasonal confounds.

On smaller lists, a test may run for months and still produce an unreliable result. Even identical AA tests can sometimes appear statistically significant, so marketers should be cautious about overinterpreting results.

That does not mean testing is off-limits for smaller brands. Pop-up tests may reach significance faster when site traffic is high, and some learnings—such as CTA color, image style, or signup copy—can be more durable across seasons than others.

The key is to match the test design to the size and behavior of your audience, and to distinguish durable learnings from results that may be tied to a particular season or campaign.

7. Use slow periods to build momentum for peak periods

A seasonal lull is an opportunity to improve the system before demand returns.

Eve recommended pulling back to a smaller, more engaged core audience during slower periods. For less-engaged subscribers, a monthly send can feel more intentional and less intrusive. Use the extra time to test pop-up copy, imagery, CTA colors, and gamified experiences such as spin-to-win.

Tabish suggested using teasers and interactive content—like voting on a favorite product color—to train subscribers to open and click. User-generated content and lifestyle imagery can also maintain engagement with relatively little effort. Campaigns that drive site traffic may trigger downstream browse-abandonment and other automations, helping keep the customer journey active even during a lull.

And BFCM preparation should start now. Grow and replenish your lists, test send times, and build re-engagement segments for customers who purchased only during last year’s BFCM period.

8. Start with early signals when the audience is broad

When you have many products, artists, or content themes to promote—but limited preference data—early engagement can become your first layer of segmentation.

Collect preference data through progressive profiling, and use website behavior to infer interests in the meantime. For a music publication, that might mean tracking which artist pages a subscriber visits. For a retailer, it could mean using product views, category engagement, or previous purchases.

Limit the number of emails about the same drop sent to one person. Then use the first email in a series to identify who is interested before sending every follow-up to the full audience.

The bigger lesson: optimize for value, not vanity metrics

Across every question, the same principle came through: strong lifecycle marketing is not about protecting one perfect engagement rate. It is about using data, customer context, and marketer judgment to create more relevant experiences over time.

That means measuring revenue per recipient, watching list health, building better preference data, and making thoughtful decisions about who receives each message. A broader send can be the right choice when it creates meaningful incremental revenue without compromising deliverability. A smaller send can be the right choice when it protects the audience for a more important moment later.

The best next step is to choose one idea to put into practice: build purchase-frequency cohorts, test a preference-collection email, add a snooze option, or start your BFCM re-engagement segment now.

Thanks to Eve, Tabish, and everyone who brought questions to our first Office Hours session. We’re excited to keep learning from the community—and from the Champions who make these conversations so practical.

Additional related resources:

How to strategically use RFM properties in campaigns and flows
Replenishment Flow
How to A/B test an email campaign
Email deliverability best practices reference
Getting started with Marketing Analytics
Klaviyo MCP Server: AI-powered analytics insights and marketing actions with customer data