Non-holiday emails outperformed holiday-period messaging.

Retail and ecommerce-focused marketers frequently change their game plans during the shopping-centric months of November and December, often ramping up frequency and shifting message content. Does email performance differ during this period?

To help answer this question, this year’s report compares performance on a set of metrics between holiday emails – those sent during the heavy shopping period corresponding to the roughly six weeks from the U.S. Thanksgiving holiday through Christmas Day and New Year’s Eve – and non-holiday messages, or those sent the rest of the year.

There was minimal disparity when studying results across all categories of senders. However, as you might expect, some significant differences emerged in the Retail & Ecommerce sector.

The mean (average) open rate was 18.8 percent lower during the holiday season. The mean click-through rate was down 28.6 percent on holiday messages. However, the click-to-open rate varied less than 1 percentage point between the two time periods.

The findings, which you can read more about starting on p. 21, suggest that some retailers’ failure to deliver personalized content and offers is resulting in lost engagement and revenue.

2. The “frequency math effect” surfaced.

Evidence of this phenomenon surfaced in the lower holiday email open and click rates in the Retail & Ecommerce sector, but it applies across verticals. Do lower open and click rates mean you should pull back on frequency during high-volume sending periods, such as the holidays? Conversely, do lower indicators of list churn mean you can ramp up frequency without inflicting collateral damage on your email list? Not necessarily.

During heavy volume time frames, the higher cadence of emails typically produces fewer opens and clicks per message sent but more opens and clicks in total. This can be attributed to the “frequency math effect.” Think of it like this: Suppose you double your frequency, but see open or click rates going down by 15 to 20 percent from one message to the next. That’s not necessarily reason to panic because during the entire sending period you should have a marked increase in total opens and clicks.

Similarly, you can also expect to see hard bounces, spam complaints and unsubscribe rates decline when you raise frequency, but this isn’t always cause for rejoicing. Instead, frequency math is at work here, too. That’s because you’re likely sending many more messages, so your total list churn over the sending period will be higher.

To better gauge the impact of frequency variations, look at both per-message statistics and cumulative open, click-through and list churn metrics for an entire period. By doing so, you’ll be able to more accurately determine whether cadence changes have driven an overall net positive or negative impact on your email efforts.

Mobile email readership is highest in emerging economies and the United Kingdom, while Retail & Ecommerce had the lowest engagement rates.

For the first time, we compared how recipients across geographies and industries differ in how they access their email (via webmail, mobile or desktop) and how much time they spend looking at messages.

Client/device usage: Emails sent in the United Kingdom and the “Rest of World” region (Asia/Pacific, Mexico/Central/South America, South Africa and India) had the highest mobile readership, at roughly 55 percent. Higher consumer mobile adoption in emerging economies is likely driving the latter area’s higher mobile readership.

Engagement/read rate: It appears that email readers either study their opened messages or simply glance at them. Overall, 53 percent of emails tracked were classified as “Read,” meaning recipients spent at least eight seconds looking at them, while 26 percent were glanced at (two seconds or less) and 22 percent were skimmed (read more than two seconds but less than eight).

Among industry sectors, Retail & Ecommerce (47 percent) and Schools & Education (43 percent) had the lowest “Read” rates as well as the highest “Glanced” rates (32 percent for Retail; 34 percent for Schools).

Your own engagement rates could vary greatly from your industry’s average, so it’s important to track and understand your brand’s specific situation. Regardless, these differences suggest that whether you use a responsive design or a mobile-friendly approach, your layout, content and calls to action need to align with the device/email client your subscribers are using.

The email activity data also syncs with what IBM has reported in previous Black Friday/ Cyber Monday studies: Traffic is up on mobile. However, conversion rates lag behind — typically one-third of those on desktop — so make sure you’re delivering a top-notch experience across devices and email clients.

Understanding the Data and How to Use It

IBM Marketing Cloud’s 2016 survey examines messages sent by nearly 750 companies and 3,000 brands in 2015, using a wide variety of measurements to establish benchmarks on customer engagement (via multiple open, click and device/email client metrics) and list churn (hard bounces, unsubscribes and complaints).

Comparing Yourself to the Best

Most studies focus only on “average” benchmarks. However, in today’s supercompetitive marketplace, “average” has become the new bottom. So, we encourage you to set your sights higher. Compare your company’s open, click-through and list churn results to those you find in the top and bottom quartiles for each measurement as well as the mean and the median.

If you want to create a world-class emailmarketing program, compare yourself to the best performers on the benchmarks that are most important to you instead of the average players.

Understanding the Bigger Picture

This report examines “process” metrics, which measure activity on email messages. Understanding where your email program stands on process metrics is one half of the benchmarking equation.

“Output” metrics, such as revenue, leads generated, cost savings, order size or number of downloads, measure how well your email campaign delivered against your individual company’s business goals. You need both to fully understand your email program’s performance.

Benchmarking Against Others – and Your Own Program

The benchmarks established in this study set a baseline you can use to determine where your email program outperforms your peers and competitors and where you need to improve.

However, accurately gauging your marketing success also requires going beyond industry benchmarks to see whether your email program met, exceeded or fell short of your own goals while adding value to the customer/prospect relationship and maximizing conversions and revenue.

Accessing Additional Resources

At the end of this report you’ll find a glossary explaining terms used in the study along with a list of resources on the IBM Marketing Cloud website that can help you understand more about key issues affecting email marketing and program performance. There’s also a list of industry category definitions you can use to determine which vertical group you fall into for the purposes of this study.

Open Rates

The open rate works best as an in-house benchmark to track over time, because it can signal progress or problems with engagement. It’s also not a completely accurate metric because it doesn’t capture all opens due to image blocking and other factors.

[Table or chart in PDF file - Register or sign in to view]

Observations

Canada and the Australia/New Zealand region significantly outperformed all other regions on this benchmark.

In the 2015 report, Canada posted markedly lower mean and median open rates. So, jumping to a position as one of the highest-performing regions might be a result of the strict Canadian anti-spam rules (CASL) that went into effect in 2014.

While not an apples-to-apples comparison with the previous year’s results, it’s a positive sign nonetheless that mean and median open rates were slightly higher in this year’s report. Mean and median open rates for the Middle East/ North Africa and “Rest of Europe” regions were lower than other regions, which may be a result of these areas being less mature email markets.

Observations

Several verticals stood above the others with mean unique open rates greater than 26 percent: Automobiles & Transportation; Insurance; Hospitals, Healthcare & Biotech; Nonprofits, Associations & Government; and Schools & Education.

Looking at the top quartile, three of these industries (Insurance; Hospitals, Healthcare & Biotech and Schools & Education) saw open rates near or greater than 50 percent.

Continuing a long-standing trend, the Retail & Ecommerce and Media & Publishing verticals have some of the lowest mean and median unique open rates. This is likely a direct result of frequency, because many brands in these verticals send broadcast emails daily or even more than once a day.

Observations

Transactional messages have long been thought of as a powerful customer contact moment, and these statistics bear out that assertion. On average, these behavior-triggered emails are opened more than two times as much as other messages.

In this study, transactional emails generated mean and median unique open rates of 47.1 percent and 48 percent, respectively, and 73.5 percent in the top quartile. At 19.5 percent in the bottom quartile, even the weakest transactional emails are opened at nearly the same rate as the mean nontransactional email. These emails are excellent vehicles for including cross-sell/upsell content. For example, an email receipt for a yoga mat might include a promotion for a resistance band set, yoga towel and stability ball based on like-minded purchases.

Click-Through Rates

The click-through rate (CTR) is a process metric that measures recipient action on an email message. It’s a more revealing metric than the open rate but also doesn’t substitute for output goals such as conversions, revenue or order value. The click-to-open rate (CTOR), also known as the “effective rate,” sheds more light on engagement than a simple click rate because it measures click-through rates as a percentage of messages opened instead of simply messages delivered.

Observations

The click-through rate (CTR) of 9.5 percent on average in the top quartile was more than 6 times higher than the overall median CTR of 1.6 percent. In contrast, bottom-quartile performers mustered a CTR of only 0.3 percent.

UK brands lead the pack with a median CTR of 2.5 percent. However, Australia/ New Zealand brands claimed the highest mean (5.1 percent) and top quartile (15.8 percent).

Observations

Emails from brands representing Computer Hardware & Telecommunications and Nonprofits, Associations & Government industries topped this comparison of click-through rates by verticals, with highperforming rates across mean, median and top quartile.

This combination is likely a testament to strong content marketing for computer hardware and the more action-oriented nature of many government and nonprofit emails.

In addition to those two industries, Schools and Education had the second-highest top-quartile CTR at 13.1 percent and thirdhighest mean CTR at 4.3 percent. Computer Software & Online Services and Lodging, Travel Agencies & Services were two of the lowest- performing industries across all CTR metrics. Median CTRs were 1 percent and 1.3 percent respectively, a significant drop from previous years.

This decrease might have occurred because of a different mix of companies in the study. It’s also possible that companies in these industries are ramping up cadence, which can lead to a lower CTR on each message sent.

Observations

Transactional emails easily beat their nontransactional counterparts on every metric. Given that these messages are based on a recipient’s purchase or conversion activity, it’s not a surprise to see these emails continue to generate high follow-on activity.

Transactional emails have a median CTR of 4.8 percent, exactly three times higher than the 1.6 percent nontransactional messages generate. The CTR for top-quartile performers is 2.7 times the mean and nearly 5 times the median for overall findings. This suggests that smart marketers are finding ways to add value through transactional messages and driving further engagement and action, such as purchasing a related product or service.

To take your emails up a notch, consider offering an incentive for a related purchase, reminding buyers to review their purchases, inviting recipients to join your loyalty club or user community, or linking to user information, howto videos and other information on your site.

Observations

If your CTOR is subpar or merely average, try analyzing your actual clicks per link in several messages and look for patterns and trends. Which factors do you associate with higher CTRs? Does one type of link (CTA buttons, product photos or text links), location in the email or the message design drive higher CTRs than others? Combine this link analysis with the overall CTOR to make and test assumptions about content, layout and offers that consistently drive the highest engagement.

Observations

Nonprofits, Associations & Government had the highest mean, median and top-quartile click-to-open rates. Computer Hardware & Telecommunications also performed well.

Among the underperformers were Lodging, Travel Agencies & Services (9.2 percent mean/19.3 percent top quartile) and Computer Software & Online Services (10.1 percent mean/24.4 percent top quartile). While open and click-through rates each have their role, the CTOR is a better metric to uncover problems within your email once it’s opened.

Observations

Top-quartile performers see nearly 43 percent of transactional messages clicked on after being opened. While median CTORs for transactional emails are only 2.3 percentage points higher (or 23.5 percent higher) than for nontransactional, they are 13.2 percentage points higher for the top-quartile performers.

If you want become a top-performing brand, consider adding a call to action to your transactional messages. This way, your email does more than simply let your customers know that their orders have shipped — you’re giving them a chance to interact while they’re highly engaged. Suggested CTAs include offers for product usage guides, acting on customized recommendations or joining your email program.

List Churn Metrics

These metrics reveal reader disengagement and discontent. They also can provide insights into your data-management practices. Rising unsubscribe rates might tell you your email program doesn’t match what subscribers expected when they signed up. Higher-thanaverage hard-bounce rates can indicate you don’t practice good, ongoing list hygiene.

The unsubscribe rate is most useful as a trend indicator. Even the lowest performers generate unsubscribe rates well below 1 percent. However, is that rate going up or down? If your unsubscribe rate rises over time or remains constant while spam complaints increase, you might have an unclear unsubscribe process or be losing subscriber trust.

U.S., UK, and Canada-based companies have similar median and mean hard bounce rates, while Australia/ New Zealand, Middle East/North Africa and “Rest of World” companies are well above the overall average. This disparity is likely a result of less mature digital marketing practices in certain regions and a smaller number of companies in those regions, in which case some outliers increased the averages.

Observations

Mean hard bounce rates vary widely by industry, from a low of 0.20 percent for Nonprofits, Associations & Government companies to a high of 1.59 percent for Industrial Manufacturing & Utilities organizations.

In the case of Nonprofits, Associations and Government entities, there may be less pressure to build their databases and resort to the types of list-building tactics that result in more hard bounces. It’s also probable that people who sign up for email programs for nonprofit organizations are motivated, care about the cause and are more likely to use a primary rather than throwaway email account.

In addition, as described in other sections of this report, frequency likely plays a role, with industries that send less email (such as Industrial Manufacturing & Utilities) generating fewer total hard bounces but a higher percentage per send.

Observations

Mean unsubscribe rates are fairly close across the regions except for the Rest of Europe and Rest of World, with rates about half of the other regions.

At 0.444 percent, bottom-quartile performers have unsubscribe rates of nearly four times that of the mean rates. For these low performers, simply reducing their unsubscribe rates to an average rate would reduce annual opt-outs by 25 percent.

With even bottom-quartile performers averaging an unsubscribe rate well below 1 percent, you might be tempted to downplay this metric, but be sure you understand how unsubscribes affect your database growth. Improving your unsubscribe rate just 0.1 percent would save 1,000 subscribers every time you send to a database of 1 million people. That’s more than 200,000 contacts retained over a year if you send four emails per week.

Mean unsubscribe rates vary widely by industry, from a low of less than 0.06 percent for Nonprofits, Associations & Governments, Media & Publishing and Automobiles & Transportation companies to a high of about 0.20 percent for Consumer Products, Real Estate, Construction & Building Products and Schools & Education organizations.

No clear patterns exist this year between top- and bottom-performing industries. However, a general best practice for keeping rates low is providing alternatives to opting out. These include allowing subscribers to easily change email addresses, lists and frequency and update profile preferences. Most important, deliver on your subscribers’ expectations after they opt in and provide ongoing value through content and relevant offers.

The mean spam complaint (“abuse“) rate is fairly similar across the geographic regions, although Canadian (0.017 percent) and “Rest of World” companies (0.016 percent) were a notch better than the overall average of 0.026 percent.

While top-quartile performers tallied too few complaints to record, bottom-quartile companies saw roughly 920 complaints per 1 million emails sent.

Computer Hardware and Telecommunications companies had the lowest mean abuse rate at 0.010 percent, while Consumer Products and Real Estate, Construction & Building Products companies had the highest mean rates at 0.039 percent and 0.038 percent, respectively. Real Estate, Construction & Building Products companies also land in the bottom of the lowest-performing quartile at 0.137 percent.

If your complaint rate is higher than average, consider adding a temporary unsubscribe link at the top of your email templates. Subscribers sometimes hit the spam complaint button when they don’t know how to opt out or just want your email to stop. Making the unsubscribe link more prominent might stave off some of these complaints.

Holiday Retail and Ecommerce Metrics

This study includes a new section that compares open, click-through and list churn rates for the Retail/ Ecommerce sector sent during the holiday season from Nov. 27 to Dec. 31, 2015, to non-holiday email.

Why focus on this category? According to one study, the holiday shopping period generates 50 percent to 100 percent more revenue than shopping days during the rest of the year.

This also is the time of year when retail/ecommerce companies increase their email cadences significantly, often doubling the number of sends compared to the rest of the year.

Not surprisingly, mean, median and topquartile open rates for Retailer/Ecommerce companies were significantly higher during the non-holiday time of year than in the throes of the holiday shopping season. With a 3.3 percentage point difference for the mean open rate, this equates to an open rate that is 18.8 percent lower during the holiday season.

Although open rates are lower, total opens for the period will be much higher because of greater frequency. If you increase your reach by 100 percent but see open rates down by less than 20 percent, you’re still way ahead in terms of overall engagement in this period.

Mean click-through rates are 0.80 percentage points lower (28.6 percent) during the holiday season. This is a greater disparity than the difference in open rates. You might expect click-through rates to generally be higher during the holiday season because of aggressive pricing discounts, free shipping offers and urgent purchase deadlines. However, the increased send volume spreads those related increases in click-throughs across many more email sends.

Click-to-open rates, which measure how well the content and offers within an opened email drive a recipient to click, were fairly close across both time periods and all measures. The slightly lower CTORs during the holidays suggest that strong offers and calls to action are performing reasonably well but that retailers are likely leaving some money on the table. With higher purchase rates during the holidays, click-throughs in opened emails should be higher. Emails that lack personalization and undifferentiated offers are likely not motivating enough recipients to take action.

Further confirming the effect of increased cadence, the hard bounce rate for Retail/ Ecommerce companies was nearly 2.4 times higher (0.3165 percent versus 0.1328 percent) during the non-holiday season. How does that happen? Again, frequency explains it. Given that most retailers double (or more) the cadence of their emails during the holidays, they’re also processing hard bounces more frequently. This reduces the per-send rate.

Unlike hard bounce rates, holiday-period unsubscribe rates (mean, median and topand bottom quartile) are relatively close to the non-holiday period. Given the typical retailer’s increase in frequency during the holidays, that’s potentially cause for concern.

If a retailer usually sends four broadcast emails a week during the year but increases cadence to six or seven during the holiday period, and roughly the same number of people unsubscribe from each email, the total number of opt outs will increase 50 percent or more in this scenario. To reduce holiday unsubscribes, consider sending an email to subscribers before you ramp up frequency, explaining the increased frequency, what kind of promotions they can expect, and shipping and return policies. Include prominent language about frequency options and link to your preference center. In addition, revise your preference-center language and options to encourage unsubscribers to opt down or snooze emails instead of simply opting out.

Email Usage and Engagement Metrics

This year’s benchmark study has expanded to include the growing impact of mobile readership on email marketing programs and the amount of time recipients are spending engaging with emails.

One new metric included this year is client/device usage, which examines what percentage of recipients are viewing emails using desktop, mobile or webmail client. The other is engagement, also called the “read rate,” which reflects the time recipients engage with the emails. You can use this data to develop strategies for email content, design and layout that reflect changes in your subscriber population.

Nearly half (49 percent) of all emails are read on mobile devices throughout the world, though with some regions a bit higher and or substantially lower. “Rest of the World” and United Kingdom regions had the highest mobile readership at 56 percent and 55 percent respectively. Higher consumer adoption of mobile in emerging economies is likely driving the “Rest of the World” region’s higher mobile readership, with contacts often skipping over desktop computer ownership entirely.

The “Rest of Europe” region lagged at 32 percent, while the extremely low 25 percent rate for Middle East/North Africa is surprising and could result from the higher cost of data plans and culture in certain markets. Webmail (29 percent) and desktop (22 percent) make up the rest of the overall email readership. Australia/New Zealand and Canada had the highest use of desktop clients at 25 percent, while Middle East/North Africa (11 percent) and United Kingdom (18 percent) had the lowest rate.

Mobile readership across industries ranged widely from a high of 57 percent for Food Service, Sports & Entertainment to a low of 28 percent for Computer Hardware & Telecommunications. These percentages seem logical as Food Service, Sports & Entertainment users will often be both younger and literally mobile when engaging with these industry brands.

Somewhat surprisingly, Hospitals, Healthcare & Biotech brands had the second-highest mobile readership at 55 percent. The low mobile readership for the Computer Hardware & Telecommunications sector isn’t terribly surprising because those recipients tend to be B2B customers and more likely to view messages on office work computers.

Computer Software & Online Services brands had the highest desktop usage at 45 percent, well ahead of Industrial Manufacturing & Utilities (32 percent) and Computer Hardware & Telecommunications (31 percent). Food Service, Sports & Entertainment had the lowest usage of desktop email clients at 14 percent.

For webmail readership, Computer Hardware & Telecommunications (42 percent), Banks & Financial Services (39 percent) and Lodging, Travel Agencies & Services (39 percent) had the highest rates. Hospitals, Healthcare & Biotech brands at (17 percent), followed by Computer Software & Online Services at 23 percent, had the lowest webmail usage.

Engagement, often called “read rate,” measures how long a recipient reads an email at three different time intervals: “read” (eight or more seconds); “skimmed” (two to less than eight seconds) and “glanced/ deleted” (less than two seconds).

Overall, more than half (53 percent) of emails tracked were classified as read, followed by glanced (26 percent) and skimmed (22 percent). Australia/New Zealand and the “Rest of Europe” regions had the highest read engagement rate, at 61 percent, while the United States was lowest at 52 percent. “Rest of World” and the United States had the highest percentage of glanced emails at 26 percent, followed by Canada at 23 percent. The skimmed rate was fairly close across regions, from a low of 20 percent for Canada and “Rest of Europe” to a high of 24 percent for Middle East/Africa.

Insurance and Automobiles & Transportation and had the highest read rates at 63 percent and 60 percent respectively. Retail & Ecommerce (47 percent) and Schools & Education (43 percent) had the lowest read rates.

Conversely, Retail & Ecommerce (32 percent) and Schools & Education (34 percent) had the highest glanced rates. The low read rate and high glanced rate for Retail & Ecommerce isn’t surprising as the abundance of “percent off” and “free shipping” offers used in subject lines combined with the use of product photos enables recipients to quickly determine their interest through a glance.

Skimmed rates were in a tight band from a low of 20 percent for Automobiles & Transportation and Food Service, Sports & Entertainment to a high of 24 percent for Consumer Services.