The SparkPost Inbox Tracker platform takes a multi-sourced approach to gathering and displaying your deliverability data. In this article, we will cover:
We recommend reading this full guide. However, you can also watch a 15 minute explanation on seeding and best practices here:
What are seed lists?
In short, a seed list is a set of email addresses that shows the deliverability performance of a campaign after they are sent to.
At SparkPost, we have two different types of seed lists: a traditional seed list, and a virtual seed list. Both lists can be downloaded in Settings (see the screenshot below).
The traditional seed list is a list of static email addresses used purely for testing email delivery and inbox placement. A seed address will never engage with email, which means it looks like your most unengaged recipients to the ISPs. Our seed list covers 72 ISPs providing a high-level view of inboxing to assist in identifying email authentication issues, IP performance, and domain reputation.
Account Managers can set which ISPs they want traditional seed coverage on in Settings > ISP Configuration:
For optimal results, it is best to configure your traditional seed list to only include ISPs that you send to on a regular basis. We'll elaborate more on this later in the guide.
Our virtual seeds, also knowns as IntelliSeeds™, are virtual users who have behavior attributes modeled after real subscribers from our panel. They automatically subscribe themselves to email lists, read and delete at a certain cadence, and even go on vacation. Since IntelliSeeds™ look like a regular subscriber on your list to an ISP, having your sends to the virtual seeds to match your regular subscribers will result in the closest supplemental datapoint to panel available.
Why should I use seed lists?
Regular seed sends increase the amount of visible deliverability data on your domain's overall inboxing performance. Seeding is especially useful for senders with smaller list sizes and limited panel coverage.
Another key benefit of seeding is that the initial send will show where ISPs are placing your mail to a brand new subscriber.
Before you start seeding, we first recommend investigating how much data you have with panel alone. This can give you an idea of what type of campaigns may require more seeding than others.
To see your overall panel coverage and how it varies on a campaign-basis, head to the Deliverability tab in Inbox Tracker and click on Campaign Tracker:
The top of Campaign Tracker has a key under Sources where you can see where we are collecting data. If you are a new customer and haven't sent to our seed lists, our panelists are the only datapoint we'll have to measure your deliverability.
To see overall panel counts, on the top lefthand side of Campaign Tracker you'll see My ISP Deliverability. Click on Volume twice to sort the panel counts from highest to lowest volume. These numbers reflect how many panelists we detected in the time period you have set:
To see panel counts on a particular campaign, scroll to the bottom of Campaign Tracker:
The above two campaigns have differing panel counts and have not sent to our seed lists (since the only Source is showing the panel icon). The gray bars under AOL indicate that no panelists were found for this ISP.
You can see how many panelists were detected for each ISP by clicking on the small chart icon next to the inbox percentage:
If possible, we recommend a minimum of 10 datapoints per ISP (especially for Gmail, Yahoo, AOL, and Outlook). In the example above, this sender should incorporate seed lists into future campaigns to get deliverability coverage for AOL and other ISPs that have little to no panel coverage.
Campaigns that use a combination of our panel, seeds, and IntelliSeeds will eventually look like this:
How do I implement the seed lists?
Because our two types of seed lists are inherently different from the ISP's perspective, sending methodology also differs between the two lists. Read this section carefully to understand how best to incorporate both lists into your program.
Traditional Seed Implementation
Remember, the traditional seeds are a list of static email addresses that do not engage with your emails, and are equivalent to the least engaged segment of your mailing list. Because ISPs look to engagement to determine inbox placement, there's a risk of oversending to the traditional seeds and skewing deliverability results. Follow these best practices to optimize data from the static seeds:
1. The entire traditional seed list should be included on a send. If you are sending to the traditional seeds, our system expects every seed to receive the campaign. This is why it is important to change your ISP Configuration to only reflect ISPs that you mail to regularly.
For example, if you only send to North America and you are including the .fr traditional seeds in your deployments, your ESP may be configured to suppress those .fr addresses and they will not receive the send. When this happens, the platform reports those deployments as a Missing:
These missing percentages count towards spam placement on the dashboard pie chart and can skew your inbox placement results. It is best to reduce the amount of undelivered deployments to the seeds as much as possible.
If you are seeing a regular Missing percentage on all your seed sends, we recommend spot checking the delivery status to a specific seed address that is reporting the missing percentage.
In the example above, since "Free.fr" is showing that all six seeds did not receive the campaign. One of those six addresses is "firstname.lastname@example.org"; check your ESP to see if they are suppressing the address for any reason and if you can override that suppression to the free.fr ISP. If the seeds are bouncing, please notify our team at email@example.com.
2. The traditional seeds should be deployed on a weekly basis. Again, these seeds do not engage, so including these addresses on frequent sends will cause the ISPs to quickly place your emails into the spam folder. Weekly sends are a safe way to have additional data without damaging your reputation to the seed addresses. Think of a campaign in your program that deploys weekly, and simply include the traditional seeds (in their entirety) on that send for best results.
3. Strive to include the traditional seeds in a live send for optimal results. If you can't include them in a live send, have the send be exactly the same as it is to your regular audience. Avoid using the words "seed" or "test" in the subject line; ISP algorithms detect these words and they may alter the results of the send.
Virtual Seed Implementation
IntelliSeeds are email addresses that we own and manage that provide additional deliverability data. Their uniqueness comes the fact that they engage with mail as if they were a real subscriber. Since IntelliSeeds look like real subscribers to an ISP, we recommend developing a sending strategy to them that emulates how you treat your regular audience.
How to Segment IntelliSeeds to Represent your Real Audience
Let's pretend that the entire IntelliSeed list makes up your whole mailing list. In your normal email sends, do you include every single subscriber on all email sends? Likely not; most senders have different segments of their list that influence the cadence and content that is sent.
Although it may be tempting to include all the virtual seeds on your sends, here are they key benefits of splitting up the IntelliSeed list:
You do not want to overwhelm your genuine panel data with our virtual subscribers. The panel is there to help you understand real subscriber experiences with your campaigns; including too many virtual seeds will overshadow data from your actual audience. If possible, we recommend striving for a 1:1 balance per campaign between panel and IntelliSeeds.
Since the virtual users are programmed to interact at a certain cadence, the less-engaged IntelliSeeds will begin to exhaust, or show spam folder placement, more quickly. If you stop sending to users that haven't opened in over 30 days, then apply that same logic to the IntelliSeeds that match that criteria.
Breaking up the list allows you to have extra "untouched" IntelliSeeds to be used later on. Remember, one benefit to a first seed send is that it tells you where ISPs are placing your mail to a brand new subscriber. Having some untouched seeds throughout your term will continue to provide valuable insight as to how a new subscriber is experiencing your sends.
Now that you understand the benefits of segmenting the IntelliSeeds, it's time to develop a plan for how you want to incorporate them into your mailing list.
To start, we've created two Google Sheets with some pre-segmented IntelliSeeds. Go ahead and download either the North American IntelliSeed list or the International IntelliSeed list (or create your own!). Please note that we have to grant viewing access to these spreadsheets.
You'll notice that the spreadsheets have five segments of 20 IntelliSeeds, as well as a "Remaining IntelliSeeds" tab. There are two ways you can approach incorporating one or multiple segments into your program:
Have each segment represent a portion of your subscribers that receive a certain message stream. For example, perhaps one segment represents users that have made a purchase in the last 30 days. Another segment may portray subscribers that have signed up for emails in the last week. Or maybe you noticed your Newsletter campaigns have very little panel coverage; have a segment receive those campaigns on a regular basis.
If your email segmentation is mainly based on engagement, another way you can approach the IntelliSeeds is to take each segment (or a few segments), and take them through your welcome series as if they were a new subscriber. That way the IntelliSeeds will naturally fall into your different engagement buckets to represent each portion of your engaged (or on-engaged) audience.
Want help with developing a seeding strategy? Click here to arrange a Basic or Advanced training session with our Product Specialist Team!
You can also reach out to our support team via Intercom chat or by mailing firstname.lastname@example.org.