Recently at 250ok we’ve had a few conversations with new clients and partners about some of the seeding advice they received from other deliverability solution providers monitoring inbox placement rates (IPR) for their brands. When we talk about seeding, we’re talking about a variety of possibilities; brands sending multiple versions of their newsletters, utilizing a send-time optimization (STO) program, or just publishing multiple different newsletters on the same domain/IP.
We’ve seen advice ranging from instructing brands to send seeds to every large segment identified by your STO vendor, send seeds for every segment, or to seed every newsletter you’re sending. Unfortunately, though this is common advice we’re hearing, it’s just not very good. This can oftentimes result in overseeding your campaigns, causing an artificial pool of unengaged users for your messages.
Let’s run a quick calculation on how many seeds you could be sending under this proposed model:
Our opinion on the most effective seed testing plan rests on a couple key tests:
Why does our advice diverge so greatly from other providers’ advice? We believe seed testing should be used as a directional testing tool, helping you predict the success of a deployment and changing your approach as needed. Seeds give a solid indication of how mail should behave for new users, with only the mailbox providers’ (MBPs) global filtering applied to the messages. It is important to remember seed accounts have no history of actively engaging with email or reporting spam (or not spam) to the MBP. Mix in other email intelligence tools like Design, Analytics, and Reputation, and you’ll see a more holistic view of your email behavior and expected delivery rates for existing subscribers.
We firmly do not believe it’s important to seed every segment when you’re sending the same or substantially similar content to multiple groups. Maybe you segment based on time zone, but send the same copy, or use send time optimization to deploy to your lists at the most engaged time for your specific consumer. The fact remains when the actual content is that similar, it’s not necessary to seed each segment because it only works to dilute your real engagement metrics. In fact, this could cause your smaller segments to appear even less engaged than they actually are.
“Based on the client data I’ve worked with, overseeding often seems to be more of a pain than providing valuable insight. Seeding provides the best benefit when you’re able to look for patterns over time and then dive in for anything actionable,” says Sloan Simmons, solutions architect at 250ok. “From what I’ve seen, results don’t change so quickly that you won’t catch the same insights day to day, as hour to hour. Seed testing is a piece of the reputation puzzle and works best in conjunction with monitored domain/IP reputation and live engagement statistics.”
Try implementing the network health test method to give a general sense of delivery at the start and end of your day. Not only will you get better data on your email program’s health, but you won’t be signaling to the MBP that you are over-mailing to specific users with the same content, only in different segments (which the MBP won’t know).
When you consider some deliverability vendors charge per seed testing event, it makes business sense that they’d recommend over-testing campaigns. We don’t subscribe to that business model, ethically or…well, in any sense, as our customers are granted access to the 250ok platform and can test as frequently or as infrequently as they choose.
If you want to learn more about why we hold this philosophy, or to get more information on how to access our powerful email intelligence made easy (and honest), let us know.