Horvath: To Suppress or Not? Segment!
List size is a metric that is ridiculous on its own. As every Email Marketer knows, it’s not about the size of the list, but the quality.
The battle over when to suppress users is age old, commonly fought between those who are focused on the size of the list, and those on the email side who understand that inactives hurt deliverability.
This is a battle I fight daily, where my C-levels are saying ‘can we open up the inactive suppression - extend it - because the more we mail, the more we make.’ I totally understand where they are coming from. As a company with email-driven revenue, the more people we can email, the more money we will make (from their perspective).
If only it were that simple, right?
Taking it back a step, we know a few things about lists, and email engagement and the impact on deliverability.
Lists are made up of several types of emails:
- Spam traps and blacklists - lets face it, these get on everyone’s list, even with validation and hygiene and good list practices. The older your list, the more likely you are to have these as old (valid) email addresses are often used as recycled spam traps. (There are many services out there such as Webbula, BriteVerify and Fresh Address that can run your list.)
- Abandoned emails - an email address is valid, and at one point was active, but is not anymore. (Audience Point can help weed these out.) The fact that there is no positive engagement with your mail hurts your deliverability.
- Inactives - users still engage with mail from this box, potentially just not with yours. This could be because your mail going to the spam box, or the user may just not be engaging. Either way, mailing these users doesn’t help you.
- Actives - the list that everyone wants. These users want your mail, and actively engage with it.
A MBP (MailBox Provider) does not know if an email was solicited or not. All they know is what you do with it once it hits your inbox:
- - How long does it take for you to open it
- - How long do you spend reading it
- - Do you ignore the email
- - Do you move it out of the spam box
- - Do you reply to it/ or forward it
- - Do you put it in a folder and save it for later
- - Do you mark as spam
- - Do you delete without reading
- - Do you click on a link
- - Do you open it more than once?
Knowing this, we re-visit the question of inactive suppression.
We cannot make money if our mail is not hitting the inbox, therefore we need to take all reasonable steps to ensure that our mail has the best possible chance of getting delivered. While we may be able to send to more users in the short-term, the inactivity of those users will hurt our deliverability and sender reputation which will result in more of our mail being delivered into the spam box, of filtered entirely by the MBP.
Again, this is something that every Email Marketer already knows. So how do you juggle internal pressure with deliverability?
While there are several companies out there who are active with segmentation and suppression, I know there are many more who send a single campaign to their entire list with little to no segmentation, and the stats are reviewed in aggregate.
I feel like every article I’ve read searching for answers or proof one way or another all say the same thing ‘TEST’. Which is great if you have the framework for setting up tests and an environment conducive to testing. Depending on how you setup your test, that could also mean not mailing some users or delaying a campaign while the test runs, which may not be conducive to time sensitive mailings. Not mailing users is typically what creates the battle in the first place.
Setting up a test that accurately measures not only engagement (opens, clicks, read rate), but ‘issue metrics’ (complaints, unsubscribes, delete without reading) can be challenging.
Here is my solution. Ideally, you are starting from a list that has had some form of hygiene or validation already. If not, and if it is in the budget, I highly recommend doing a batch scrub. This will help reduce the risk that comes with mailing older/inactive users. Shouldering this up-front cost helps ensure that you do not take a revenue hit due to the massive deliverability issues that can come from Spamhaus, for example.
Take you list and build segments - active users, inactive for 30-60days, 61-90 days, 91-120 days etc until you get to the point you currently have your suppression set at. While I recommend keeping your segments at roughly 30 days each, this does depend on your business model. If you only mail once a month, it makes sense to open these segments up.
Take your content and duplicate it for each segment. Update the links in each to contain unique tracking, ie utm-content=30-60, utm-content=61-90 etc. The reason this is important is it gives you another point to correlate your data - you will be able to use whatever analytics tracking on your website to determine if a user does click through, how many of them came from your campaign, and if it was quality engagement.
This will mean you mail the same amount of users you normally do, just over 4 campaigns (for example) instead of 1. While it may be cumbersome, it is worth doing this for several campaigns over multiple weeks. This is worth doing on both promotional and content mailings, if only so you have solid evidence regarding user retention, total revenue, and revenue per mail (factoring in the cost per mail, if known) for each type of mailing to build your case. Doing this over a longer time span also means that you can have confidence in your results as single mailings can be skewed by the curiosity effect and external factors.
From here you review positive engagement (opens and clicks), issue metrics (complaints, unsubscribes), and revenue per group.
This data is how you can build your case - active users typically have the lowest complaint rate and highest engagement, which leads to better deliverability. Again, I know this is simple, but the better your deliverability, the more revenue you can generate from email.
As you go through each of the segments, you will typically see engagement drop, and complaints and unsubscribes increase. If you were to keep mailing these inactive users, your sender reputation would start to suffer and your overall deliverability would decrease, thus reducing your overall ability to generate revenue.
While this is what you will typically see, it may not be the case. This segmentation strategy is designed to give you the data that you need to make informed decisions that balance business objectives with email deliverability.
- Break your segments out into sections that make sense to your business and your program
- Make sure your links have unique tracking
- Review not only engagement metrics
- Review issue metrics
- Check in on your deliverability and which campaigns are causing issues
- Report on revenue per segment to build the business case
Not everyone understands the intricacies of email marketing - this keeps us employed, and often times frustrated. Internal stakeholders may not understand deliverability, but what they do (or should) understand is revenue. When you can’t exclude parts of your lists to wait for test results, use segments to build your case.