Archive for the ‘Email Delivery Studies’ Category

The Secrets of Email Deliverability Unveiled (Part II)

Friday, February 9th, 2007

In my last blog posting, I mentioned Microsoft’s deliveribility guy Brian Holdsworth but ran out of time before jotting down my notes about his talk, so here they are:

MS Outlook and Windows LiveMail (formerly Hotmail) represent 600 million people / users in the world. Microsoft expects this to climb to 1 billion in the next few years.

Big change in Outlook 2007: it automatically postmarks emails to email addresses NOT in your address book. It delays the send several minutes per email address not in your address book. This will significantly reduce the amount of spam generated by botnets and zombies that send email through unprotected PCs (which is responsible for 80% of the increase in spam in the past year).

Spammer Trends and Tactics:
Microsoft has brought 376 successful lawsuits against spammers in the past 3 years. Part of the lawsuit process requires the spammer to come into Microsoft’s office and explain why they spammed. Recently, a 17 year-old came into to talk to the Microsoft deliverability team – he started when he was 9 years old and was now sending 25 million emails per day and made $300 per day on affiliate revenue (which breaks down to roughly $1 per 100,000 emails – not great ROI, but pretty good when considering there are no costs other than his eventual fine and jail time).

Of the 4 billion emails per day that Hotmail processes, 90% is spam. Much of this is image spam, and spam generated from botnets and zombies.

Next version of Hotmail is Windows LiveMail which is very similar to Outlook.

An Unsubscribe link is built into every email in Windows LiveMail in the Return Header.

Microsoft Goals:
1. Reduce Spam in Inbox
2. Improve deliverability for legit senders:
a. volume based reputation
b. Sender ID + past reputation (Outlook postmark)
c. Unsubscribe built into Windows LiveMail

Last note is the phenomenon of communication barriers within Microsoft. The Outlook team makes major changes/shifts every 3 years, and for Outlook 2007, they are moving to a content rendering engine built in MS Word instead of the natural choice of Internet Explorer. Microsoft’s deliverability found out about this at the same time the general public saw the press release a couple weeks ago – doesn’t know why this decision was made by the Outlook team. Maybe the next ReturnPath event can address this topic in more detail…

The Secrets of Email Deliverability Unveiled (Part I)

Thursday, February 8th, 2007

I attended Return Path’s Seattle Workshop yesterday and learned a lot more about “Email Strategies that Increase Deliverability and Response” than I expected. They even had a guest speaker from Microsoft Hotmail (now branded Windows LiveMail) – Product Planner Brian Holdsworth who shed light on Microsoft’s deliverability strategies. Apparently, over 50% of email browser use is MS Outlook, so maybe we should listen to what he has to say.

Here are my notes from the event (Return Path speaker):
Large ISPs getting significantly more email volume each day – recently showing 7.4billion emails per day.

What is Email Reputation?
1. Complaints (Informal)
2. High Unknown Users (unclean list)
3. Spam Traps (spamcop.com and other types of spam-catching email addresses)
4. Sending Infrastructure (IP address must be clean)
5. Sending Consistency (similar daily email volume)

Where you are sending from matters a whole lot more than domain reputation and email content.

Feedback loops are important – automatically unsubscribes people who complain. Added benefit of decreasing complaint rates up to 40% at Yahoo, Hotmail, when feedback loops are implemented by senders.

Strategies to Reduce complaints:

  • Welcome email upon subscription 
  • Link to sample newsletter
  • Use double opt-in
  • Always respect unsub requests
  • Make it very easy to unsubscribe
  • Content relevancy is key
  • Conduct complaint analysis

eROI Releases Q2 2006 Email Deliverability Study

Tuesday, August 15th, 2006

Home and work users have different preferences.

For e-mail marketers looking to make the biggest impact, it makes sense to pay attention to whether users are receiving their e-mail at home or at work.

According to a Q2 2006 study of e-mail marketing preferences by eROI Inc., 79% of people who subscribe to business-to-business e-mail marketing messages receive them at a business e-mail address. Just 19% of people who receive consumer-oriented e-mail marketing messages subscribe with their business e-mail address.

Those who receive consumer-oriented messages were also much more likely to use an e-mail address created specifically for e-mail marketing: 24% did, vs. 6% of those who received B2B messages.

Read and Download the Study

Please Take Our Short Email Survey

Monday, July 24th, 2006

In an effort to expand our latest study we are currently working on, we would like to invite you to participate in this survey. This survey will help us to combine some key data points we analyze over our emailROI network and provide you with the combined results in August.

Take the short survey

Sign Up for our Study Release and download past studies.

eROI Releases Q4 2006 Email Delivery Study

Friday, January 27th, 2006

It’s time again for the eROI quarterly email study on deliverability. The eROI, Q4 ’05 email study breaks down, by list size, which day is the best day for email campaigns. In general terms the Study indicates a 29 percent drop quarter over quarter in read rates, and a 21 percent drop in click rates in Q4 over Q3 2005. Additionally, the Q4 data shows a similar trend from the third quarter in that the noticeable high point in the week occurs on Friday for both reads and clicks. So, from this quarter eROI reaffirms that sending volume is inversely related to how reads and clicks react.

Q42005Chart-3.gif

For Q4 eROI sees similarities to last quarter with read and click rates generally declining as list size increases, but there are some major differences. Read rates are much more erratic between 500 and 10,000 emails. Click rates see a more consistent behavior and do not decline in nearly as consistent of a pattern as last quarter. Volume for 250,000+ lists takes a huge jump upward resulting in approximately 20 percent read rate. We believe this goes to show, segmentation works and the general B2C marketer really needs to adopt more targeted mailing campaigns. B2B mailers are starting to adopt them more and more.

For the full study and charts go to: http://www.eroi.com/eroi-email-marketing-statistics-q405-study.html

Q3′05 eROI Email Statistics by List Size Study Released

Friday, November 11th, 2005

Just an FYI, we released our new study this AM. Look for it on MarketingVox today, eMarketer Friday and others in the days coming. Read the overview below, and get the whole PDF at our site.

GettheQ32005Study.bmp

Get the PDF

In Q3 of 2005 we notice that the middle of the week is the low point, as far as read and click statistics go. Noticeable high points in the week occur on Sunday and Friday for both stats. So from this quarter we reaffirm again that sending volume is inversely related to how reads and clicks are going to react with the one exception of Saturday.

Why is this shift in behavior occurring? As we close in on the holidays, consumers are beginning to think more of how they are going to spend their money this holiday season. While we anticipate these stats to start even out more as we come closer to December and through the new year, this behavior shows marketers are getting an early start and secure the consumer awareness and solidify a place in their, um, your, wallet.

Overall By List Size
This quarter eROI takes a different look at day of the week stats. When looking at aggregate stats we see less dramatic changes in read and click statistics and they are not always applicable to large and small senders. eROI decided to take a look at day of the week statistics by list size.

Get the PDF

Testing Email Creatives in Gmail and Hotmail Email Clients

Saturday, July 2nd, 2005

by Dylan Boyd, eROI

We created a test email blast to see what the limitations of HotMail and GMail are in terms of CSS. We know that many email clients limit the use of CSS, but it in some cases it is possible to use it. If you are familair with your audience, you may look to segment lists by ISP and email client to maximize delivery. This is a real limited analysis, but we found out some guidelines that should help us in the future.

Specifically, here’s what we found:

GMAIL

Ignores all styles in “” “” tags. It doesn’t matter if they are in the “” or in the “”
Does render inline styles, with the exception of:
a. Background Images
b. Specific Heights
Doesn’t display ANY images by default. It needs to be activated by the reader to “include external images”.
Make sure that you use complete alt tags to illustrate the images with text placeholders.
Be careful as to the extent of the words you use as your competitors can
server Adwords alongside your email based on the keyword index.
When displaying external images, it does render traditional background images correctly, for example: ”

“, as well as height and width.
Ignores BODY background images. 

(more…)