All Collections
Diagnosing Issues
Send Time AI: Requirements, Best Practices, and Issue Diagnosis
Send Time AI: Requirements, Best Practices, and Issue Diagnosis

Looking to get the most out of Send Time AI? Of course you are. Here's where you'll find answers to both basic and advanced questions

Scott Gutelius avatar
Written by Scott Gutelius
Updated over a week ago

Quicklinks:

Send Time Optimization, or Send Time AI, can be a great way to boost engagement with your email marketing campaigns. Our patented AI modeling analyzes open data on a per-contact basis, then predicts exactly when to send your message so that it lands in your prospect's inbox when they are most likely to engage. It's pretty cool, if we do say so ourselves.

Below is a list of ways to get the most out of our per-contact Send Time AI. If you're running into any issues, this is the first place to check to see if it's something we've encountered before, and what steps to take to correct it.

If you don't see your use case listed below, feel free to reach out to us and we'll be happy to investigate.

Requirements for Per-Contact Send Time AI:

The basic requirements for configuring and executing a successful Send Time AI step are pretty simple.

First, you need to have at least seven days during which you can send messages.

Since the objective is to reach your prospect at the best possible time FOR THEM, this can often be outside of regular business hours, especially if they are in another time zone, or another country.

Sending over a 7-day period helps ensure that you’ll be reaching your audience at the time they are most likely to engage.

Second, STAI works best when there is open data for your prospects.

It's possible that your prospect will open their email at set times throughout the day, but not necessarily. Having a solid history of each prospect’s patterns over time allows our predictive modeling to work its best.

Keep in mind that with Send Time AI, it may take up to seven days for all the prospects in a segment to receive your email. So if your message is time-sensitive (like a transactional email, welcome message, or business compliance), it's best to choose a different method of delivery, such as a Simple blast.

Also keep in mind that your prospect doesn’t need to have flowed through a Send Time step previously. We’ll collect all the open data from Eloqua on each contact regardless of whether it’s a Send Time step, Message Testing, a Simple blast, or any Eloqua step. We will gather and use all that data for our predictive modeling.

Best Practices:

Here are some tips and tricks to get the most out of Motiva’s Send Time AI.

Establish a Baseline

You may have done this already, but it’s good practice to establish a baseline for your marketing campaign steps, so you’ll be able to measure Send Time AI’s effectiveness. Be sure your baseline is in place before proceeding.

Remove Inactive Contacts

This one actually applies to any marketing campaign step and any contact database. If you have inactive contacts – contacts who have stopped engaging or have never engaged – it’s best to just purge them altogether.

There are a number of reasons why this type of data hygiene is important to your email marketing campaigns, but it’s outside this document’s scope. For a deeper look at all the reasons, check out our blog post covering them.

(And if you’d like us to help identify which contacts should be purged, our Dark Pool Cleanup will do just that.)

Analyze the Send Report

In each Motiva step, you will see a sender report analyzing the results when you log into the Motiva app. This is where you can level up your superpowers.

These reports will show you at a glance when engagement is highest, based on the volume of sends and the unique open rate. It’s a quick and easy way to understand campaign performance.

Restrict the Send Window

Speaking of those Send Time Reports, look for patterns emerging in your audience when you analyze those post-action reports.

For instance, say your initial Send Time step was configured with unrestricted send times (best practice when you're first starting with STAI), and you notice your audience’s highest engagement is clustered within a certain range.

If you do identify one of those patterns, you may consider focusing the STAI configuration during those peak hours. Try removing some of the "off hours" with lower engagement, and see what effect it has on your metrics.

Possible Issues and Recommended Actions:

See something strange? Here are some tips if you see something you weren't expecting.

No Improvement in Open Rate

Scenario: You established a performance baseline, and you’ve been using Send Time AI for a while, but the numbers aren’t improving. What’s going on?

There are a couple of factors that could be causing improvement to plateau.

1. Send Window Restrictions are too strict. If you’ve tried restricting send times, leave them unrestricted.

2. No Send Window Restrictions. Basically the inverse of the above. If you've been leaving the send windows unrestricted, check out the Send Time reports and start removing 3-5 hours each day when engagement is lowest.

3. Inactive contacts. Remove inactive contacts from your database, so they don't flow into the Send Time AI. Inactive contacts can have a seriously negative impact on marketing campaigns (engagement metrics, deliverability, Sender Score, etc.).

4. Email Frequency. If you’re emailing your prospects too often, they may just be fatigued and at risk of becoming unengaged, unsubscribing, or marking messages as Spam. Check out our Frequency Intelligence reports to see what the sweet spot is for your audience.

5. Email Content may not be relevant. This one’s a bit sensitive, but don't shoot the messenger. Even if your email lands right as your prospect is checking mail, if the message isn’t deemed relevant, they aren’t going to engage.

If you suspect this may be the case, you may want to try refining your content through a series of multivariate Message Testing experiments until you land on messaging that resonates with your audience.

Sends Beyond Send Window

Scenario: You’ve configured your Send Time step, activated it, and the results look great... but email batches are still being sent after seven days. What’s going on?

1. Activation during send window. The data processing and modeling that enables Send Time AI to work like magic can take a few hours, depending on how large the audience is and how much open data is available. Be sure to allow the contacts to flow into the step BEFORE the beginning of your send time window.

2. Frequency Management enabled. Sometimes frequency management can interfere with sending, if the same audience segments are in use in multiple campaigns. If STAI determines that your prospect has already received the allotted number of emails (based on your Frequency Management configuration), it will reschedule sending for that prospect. This can result in multiple reschedules, week after week.

To prevent multiple reschedules from happening, you can either set an end date during STAI step configuration, or override Frequency Management.

Send Time AI fails to schedule or send any batches

We’ve only encountered this with Send Time AI steps which are configured with extremely restricted Send Time windows (such as just a few hours on the same day of activation).

If you are seeing this happen, double-check the step configuration to ensure that you’ve allowed a full seven days for the Send Time step to run, or you’ll be missing out on all those hours during which your audience might prefer to engage.

Or to put it another way: there are 168 possible send hours during a week. If you limit your sending to eight hours on one day, you're missing out on 95% of the possible hours that your prospect may prefer.

If you are short on time and need to get your email messaging out to your audience quickly, your best bet would be to send it as a Simple email blast.

We hope that helps solve some of the issues you might encounter when using our Send Time AI. If you'd like a little more guidance, or you're having an issue not shown above, feel free to reach out to our team for support!

Did this answer your question?