Yahoo! Implements Big Filtering Change

Y! Mail Logo.pngAccording to this article by MediaPost, Yahoo! has confirmed something we have been tracking at Responsys for the past several weeks. They are now applying more weight in their filtering to subscriber engagement metrics. These metrics include whether your subscribers are opening, clicking, or otherwise interacting with the email you send them.

What Does This Mean For Marketers?

For marketers, this means it has just become more difficult to reach a large portion of your subscribers.  With Yahoo! typically comprising up to 40% of a consumer marketers’ email list in North America, this has major implications for big brands.

Back in February this year, Gmail stepped forward with new filtering rules that add focus to how engaged your email audience is.  If you’re sending email to a bunch of people who never open or click, this is a signal to the ISP that people really don’t care to receive email from you.  Now ISPs are stepping in and routing this unwanted email to the spam folder, saving the recipient time and effort.

What’s a Marketer to do?

Yahoo! has been around as a webmail provider for a long time.  If you have people on your list who subscribed years ago and rarely or never open or click, this would be a good time to cut your losses.  No longer can you simply keep sending to the unengaged masses and expect the ISP to keep delivering this unwanted email.

You can see how engaged your subscribers are by reading open and click data.  If a subscriber hasn’t opened or clicked in months, they’re not interested.  Cut back on your send frequency to the less engaged, and stop mailing entirely to the unengaged addresses on your list.

This segment is not reading or buying from you anyway, so you have little to lose.  On the other hand, continuing to send email frequently to the unengaged will trigger filtering on all your email, and you’ll lose the opportunity to reach even your engaged audience.  This will gut the effectiveness of your email program.

The definition of active/inactive subscribers is different for every mailer depending on your business model.  But in any case, the activity window is measured in months, not years.

Industry Implications

With Gmail leading this industry shift in February this year and now Yahoo! following suit, engagement filtering is here to stay.  From an ISP’s perspective, it works well.  Expect to see more of this from other ISPs in the coming months.

Metrics

Deliverability monitoring has become more complicated than in the past.  Popular seed monitoring tools are not designed to measure engagement-based filtering.  Seed accounts by definition represent unengaged subscribers – seed accounts don’t open or click.  For this reason, seed monitoring tools have a tendency to overstate delivery problems at Gmail and now also at Yahoo!

The real deliverability metric to watch closely is your Unique Open Rate.  If you see a sharp decline in open rates, or a relatively low open rate for one ISP compared to the others, this is a strong indication your email was filtered out and did not reach the inbox.  Everyone’s open rates are different, but if your Unique Open Rate at Gmail is 2% or less when other ISPs show much higher open rates, your email is landing in the Gmail Spam folder.  Some people actually do go into their spam folder and open mail – but not many.  If seed reporting is showing low inbox rates, but open rates look healthy, your mail is reaching the inbox.

Don’t forget all the other factors you learned about ISP filtering.  You still need to keep spam complaints low, avoid mailing to spam traps, keep bounce rates low, etc.  None of that went away.

ISPs are raising the bar on entry to the inbox.  Marketers need to watch their open rates more closely and not assume everything is fine because your send volumes are way up.  Develop a smart and refined activity targeting plan based on individual subscriber open and click data.  If you implement this plan consistently and watch your metrics regularly, you can have good delivery rates at Yahoo!, Gmail and other ISPs.  You will actually see your open rates increase over historical norms.


Y! Mail Logo.png

Like what you've read? Click here to subscribe to our newsletter.
This entry was posted in Email Marketing on by .
Daniel Deneweth

About Daniel Deneweth

Daniel Deneweth heads up a team of Email Deliverability Strategic Consultants at Responsys. He shows clients how to maximize the ROI from email through improved inbox placement. Prior to Responsys, Daniel held a variety of roles at the deliverability firm Return Path. His tenure included managing the Sender Score Certified program, where he collaborated with ISPs and helped senders implement email best practices. Daniel brings this insight and in-depth deliverability knowledge to help clients maximize their inbox placement rates, and accelerate the ROI of their email channel. Follow Daniel on Twitter where he focuses on email marketing and deliverability.

One thought on “Yahoo! Implements Big Filtering Change

  1. Tami Forman

    Dan – Thanks for spreading the word about how Yahoo is looking at engagement to determine inbox placement. As we wrote in MediaPost on Aug 1 (http://www.mediapost.com/publications/article/179994/summer-2012-when-engagement-began-to-matter.html) this is pretty big shift that marketers need to pay attention to, especially at the big ISPs. In working with a bunch of clients we have found that the most accurate way to assess inbox placement rates is with a combination of seed data (which give you a really good indication of how the overall spam filtering technology at a given ISP is treating your email) and panel data (which tells you how your email is delivered at the individual subscriber level). Unfortunately, open rate (unique or otherwise) is really not a good proxy for IPR. The open rate metric is plagued with problems anyway (image suppression, for example) and there are too many variables in play that can cause your open rate to be low. Also, open rate is not what ISPs are using to assess engagement. They are looking at metrics on *THEIR* side of the server (like read rate, how often an account is accessed, whether or not messages are deleted unread). It is entirely possible for a marketer to have a low open rate and high engagement (and vice versa). Using tools like those provided from Return Path (which, of course, are readily available to Responsys customers) continues to be the best way to assess deliverability performance.

Comments are closed.