That’s a great question @Brianna!
Smart send time (aka Send time optimisation) is definitely worth exploring and is a great feature that can help you target your right audience (based on your segmentation strategy) at the right time. Timing is great way to optimise open rates and leveraging tools like smart sending from Klaviyo can be of great help.
When it comes to your core engaged audience, make sure you exclude MPP users and send only to your right audience using smart sending.
Hope that helps.
Regards,
Mohsin
Thanks, @inboxingmaestro. Do you know what other engagement metrics are factored into the Smart Send Time determination? Is it solely based on opens? Or are clicks included? I understand what you’re recommending about excluding MPP users from our sending segment, but I am still curious about the metrics used for Smart Send Time determination, because Apple users aren’t the only ones that can be recorded as a proxy open. Thank you.
Hi @Brianna, glad to see you back here in the community. This is a great topic so I’m glad you brought it up here. So a few thoughts on this:
- Smart Send Time is applicable to a specific audience of your choosing. In fact, you may create different Segments (or Lists) and have slightly different Smart Send Time results. One idea is to test Segments you know who aren’t subjected to Apple Privacy Opens. Take a look at this: How to identify iOS15 Mail Privacy Protection opens
- From the many many Smart Send Time tests, my results varies by merchant and audiences. For some brand/audiences, the timing of the email is pretty crucial. One example is brands that cater to folks during the working hours (e.g. think Office Supply, or Corporate Gifting Brands, etc), finding the right time to send was helpful and drove comparative higher engagements.
- For other brands, the data shows higher open rates at certain times, but they aren’t dramatically greater than other times - maybe only 1 or 2% higher. But in the competitive landscape (and noise) of marketing out there, I’ll take a 1 or 2% improvement on Open Rates, why not?
- I’m not sure if Smart Send Time optimizes on any other factor other than open rate (maybe someone can tell us!) - it seemed like its design and purpose is centered around opens. But I think it leads to the basic principle that if people aren’t opening emails, then the downstream performance are probably going to be impacted too. I suppose, there are chances where a certain time might get better conversions (sales) but have low open rates, so you should probably do some analysis on what time your most orders occur to do your own comparative testing against the best time to send for open rates.
Curious to hear others perspective on this!
---
Joseph Hsieh // retentioncommerce.com // twitter: @retenion
@Brianna - Smart send time factors primarily on open rates. Here’s a great article written by Klaviyo data scientist (I believe 
) - https://klaviyo.tech/why-does-klaviyos-smart-send-time-optimize-for-open-rates-4259beb49455?gi=5fc2411c3ec5
And when we talk about MPP, it’s not just Apple users (iCloud domains) but it also include users who have setup their Gmail or Microsoft accounts within Apple Mail environment. So, relying on just open rates would not be helpful, you’d need to look beyond opens and focus on clicks / click to open rates for measuring email performance.
However, AFAIK, when Apple email clients generate false opens and opens at times that don't match actual read times, smart sending IMO will become less accurate and less effective unless Klaviyo inculcate secondary engagement metrics like clicks or conversions and create a weighted analysis basis whole engagement. I can further deep dive into this and get back to you. 
(I assume smart send algorithms will use clicks and other behaviors to validate the best send times, which will override false open signals from Apple.)
So, like I said, it’s still good to run few experiments around segmentation and segregate your audience between MPP vs non MPP and target non MPP users using smart send time feature and monitor the performance closely and its impact.
Regards,
Mohsin
