DMA Email Marketing Council Blog

July's Issue of Infobox

This month we have a focus on metrics and results. You can see when marketers are actually sending emails; how to overcome three common barriers to adopting metrics and how to make the most out of your own list.

We also have news from the DMA Legal & Best Practice Hub on its definitive guide to email metrics.

UK marketers do not optimise for time of day; send times are based on sender convenience rather than subscriber preference Dela Quist, Alchemy Worx

Email marketing: overcoming three common barriers to adopting metrics Riaz Kanani, Silverpop

Lists: it’s not the size; it’s how you use it! Tink Taylor, dotMailer

Metrics, metrics and even more metrics Jonathan Burston, Chair of The Legal & Best Practice Hub (alias: Do Good Things Hub)

Read the full newsletter

And if you haven't already signed up to receive Infobox in your inbox, then subscribe here!

Customer data – knowing where to look

You might say that I’m bound to say this but...email marketing is a bit of a no-brainer for tough economic times – it is highly measurable, can be highly targeted and exerts less budgetary pressure than most other DM disciplines. The smartest brands have grasped this and are devoting more time to the channel but, while they are recognising its importance, many are still failing to target their mails and segment their databases effectively. Recent Forrester research showed that 19% of email marketers are still sending a single generic version of an email to their entire contact list while 26% will only divide their list by a single dimension. Put simply, by doing so these marketers are systematically reducing their chances of delivering high returns.

Segmentation is a word that can often provoke fear in the minds of those that hear it – seeming as it does to carry connotations of all sorts of impenetrable IT and data processes. The truth is, however, that with just a little time marketers can segment their email distribution lists into target audiences using data they don’t even know they’re capturing. The key is knowing where to look. Broadly speaking, there are six places:

A customer’s e-mail behaviour:
The most easily available information is the number of opens, clicks and link-levels your emails get. These simply indicate which subscribers are most interested in your message enabling you to market to them individually, targeting specific topics of interest.

A customer’s profile:
In addition to asking them directly, you could determine your customer’s profile information (such as location, name, and professional title) from other data. A favourite store can give you an idea of their geographic location while a purchase history of childcare goods can be a pretty good indicator of a parent/family.

A customer’s preferences:
Successful email marketing is all about allowing the audience to dictate the details of the communication. The best way of doing this is simply to ask how often they would like to be contacted and what content is most relevant to them.

A customer’s online behaviour:
Simply identifying those customers that have recently purchased, browsed or abandoned a shopping cart can help you generate additional revenue and build relationships. Emails alerting customers to items they left in their shopping cart or a goodwill ‘Thank you for your purchase’ messages can be highly effective.

A customer’s offline data:
It is vital to account for offline customer behaviour and purchase data. This also enables you to provide personalised information in future campaigns such as special promotions tied to a local store, preferred airport, or destination.

Data from Poll/Surveys:
These are useful tools to gauge what different audience segments want. You can use data from polls or surveys to inform decisions on making a change or assembling an offer. For example, ask customers ‘what new feature would you like to see in a monthly newsletter?’, ‘how would you rate the company’s online programmes?’, or ‘how often would you like to receive promotional emails?’

Some or all of these different sources of data are available to most marketers but, of course, simply gathering the data is not enough. Once you have identified what your customers actually want it is essential that you can make this data ‘actionable’, in other words ensure you are actually able access it quickly for timely execution.

Simone Barratt, MD e-Dialog 

Life Story or just an email address?

Having spent more time than is healthy thinking about how much or how little data is the "right" amount to collect, here are my tips on how to gauge what is right for you.

1. Why am I collecting data in the first place? Consider what you are going to do with the data and make sure that the consumer is made aware of what they should expect when they submit their details.

2. Collecting a very limited amount of information will probably increase your sign-ups, but think about how useful this information will be. Even if you’re just starting out, think ahead. Profiling and segmentation will be a lot more straightforward if you have collected a postal address. Also, consumers are not as scared as you may think about giving their details as long as they can see what they will get in return and how you will look after their data.

3. Permissions – I’ve sat in protracted meetings discussing permissions and I concede that they are dull, however they are vitally important. Most channels are opt-in, although postal is still opt-out of course. Does it make sense to ask people to opt-in for all channels? Test it…does positive postal opt-in result in higher quality, more responsive data than relying on failure to opt-out? Am I future proofing my base by doing this? Is one simple multi-channel permission, followed by access to a preference page, less confusing than a jumble of little tick boxes that may confuse the consumer?

4. Consider a "creeping data strategy". Collect a limited amount of information up front, then use this to communicate and build brand trust with the consumer. Over time, collect additional (but only relevant) in subsequent communications.

5. Where do I put the information I’m collecting? It might sound daft, but is there actually a field in your database where you can store the answers to the questions you are asking? It's all well and good collecting "interesting stats", but infinitely more useful if you can add depth to your database and use the new criteria for targeting.

As you can see there isn’t a right or wrong answer, but be forward thinking, transparent and honest and you’re already half way there.

Rupert Harrison

Chair Best Practice & Data Hub, DMA Email Marketing Council and Data Planner, News International

"Email Equity" or "Spray & Pray"

£9.11, £20, or at the risk of sounding like Bruce Forsyth - Higher / Lower? 

At the start of January 2009, an article that was published in Revolution magazine put the value of a permission based email at £20. This proposed a new updated value from the £9.11 put forward by the Direct Marketing Association (DMA) National Benchmarking report in this post in 2007.

Two things struck me by this:

1. Whilst both these figures can be criticised as being only "finger in the air" - how many companies can, even if they choose not to say, actually place a value of their own asset (excuse me if I don't put email list, database etc but lets call it what it is)

2. Given that they can value this asset do they treat it as such i.e. regularly measure it, manage it, and through a structured process aim to improve it !

Over 2008 I monitored the number of emails I received from 24 companies (I registered but then did not open). Have a look at the monthly figures but the difference was stark.

• One company emailed me 105 times in the year!! This suggests that they are either very persistent, very optimistic or not listening to the feedback they should be getting ?

• Others emailed hardly at all - I opted in and they failed to deliver !

"The time has come when advertising has in some hands reached the status of a science" - Claude Hopkins wrote in 1923, but my experience over 2008 suggests not all hands !

If companies do not go beyond a "never mind the quality feel the width approach" i.e. How big as opposed to how valuable, they will be unable to have communication that is relevant both to their customers and their corporate goals.

In discussion with various delegates at a recent Ready Steady Email event, feedback suggested that people are starting on the route to email equity but it still has some distance to travel. However the sooner companies start the sooner they should get there!

I'd appreciate your thoughts.

Stefan Elliot
Six Serving Men

What has the most impact on revenue generated by your email campaigns?

Is it the time of the day? Is it the day of the week?     No it’s the week of the month!*

Of all the things that you can test about your email campaigns it is the Alchemy Worx view that time of day and day of week testing provide the least valuable insights.

A far greater number of people on your list are likely to buy because they have just been paid than the number of people who buy because you sent the email before 10am on a Wednesday.

Several things have brought the old questions - what is the best time of day and/or day of the week to send email campaigns out – back to mind recently. The first was a recent piece of research into consumer behaviour by Forrester. Their “email opens index” reveals that consumers’ personal motivations exert more influence than the time of day a message is sent. When asked “How much did each of the following influence you to open the emails you receive from companies?” Only 9% of consumers sited the time and/or day they received the message as having any influence.

By comparison the top four were:
Your interests – 59%
Your needs at the time – 51%
The sender – 48%
I like the company – 46%

This did not surprise me at all.

Exact Target also released the 2008 Channel preference survey which looked at usage patterns in email by time of day, it identifies an “email prime time” of maximum engagement between 8am and 11am.

One of the most prevalent mantras in email marketing is TEST, TEST, TEST & TEST again; so I can imagine there will be hundreds of marketers out there planning to roll out day of the week and time of day tests at this very moment. WRONG!

Last but not least, Skip Fidura of OgilvyOne in a recent blog points out the inherent flaws in the way most people test for the optimal time of day or day of week to send a campaign. He also said that I have never been a fan of [standard] day of the week testing. That’s putting it mildly! Like him I think finding what day of the week is best for the average customer is meaningless.

In my September post on optimising send frequency I shared the fact that here at Alchemy Worx we have found if you really want to optimise send frequency; it is vital to target all behaviour not just the average, because not everyone on your list wants the same thing. This is also true of send time.  If you conduct a standard day of the week all you will find out is the best day for your list and not what is best for any given individual or groups of individuals. For any given day of the week or time of the day that you could choose to send your campaign out, including 3 am Sunday night, there will be at least one person on your list who thinks that is the perfect time or day and one person who thinks that it is the worst time or day.

I have already stated that we are not fans of day of week testing here at Alchemy Worx unless of course you decide, as Skip suggests, to test and analyse multiple segments over time. Done properly this will require a significant amount of resources dedicated to the test and it will be several weeks or months before you get any meaningful results. To get around this problem we have developed a very effective shortcut; we analyse the effect the time and day of the week a message was received, has on the performance of every message sent over the previous twelve to twenty four months.

Over that time period most companies will have sent messages on every day of the week most times of the day - often without prior intent to test.

This historical data is a goldmine! At the very least you can confidently make your test plan more manageable by focusing on the most likely looking days or days where you do not have any data; best of all you have a benchmark to work from. Our golden rule is this; before you test anything ask yourself if the answer (or part of the answer) is right under your nose in your historical data.

One of the benefits of conducting a forensic audit of your historical data is that it can yield unanticipated and valuable additional benefits. In setting out to discover the best day of the week or time of day for our clients to send their campaigns out, we discovered that which week of the month a message was sent had a far greater impact on revenue than other temporal consideration!

The best week of the calendar month for generating revenue is week 5!

Followed by:
Week 1
Week 4
Week 3
Week 2

Don’t take my word for it, test this for yourselves.

The reason for it is simple, most people feel richest just after payday and in the UK the majority of salaried people get paid between the 25th and 30th of each month (I checked this with the company that handles our payroll).

So if your campaign is monthly aim to get your message out as soon as possible after payday - weeks 5 or 1 depending on the month. If at all possible try to avoid week 2 and to a lesser extent week 3.

If your campaign is weekly you may want to consider lower discounting or adding less value in weeks 5 and 1. Conversely if you need revenue in weeks 2 and 3 you will need a much stronger offer.

So back to my first point; if you also think about send time/day or week in terms of size of segment i.e. how many of the people on your list are likely to feel the same way about something or be in the same situation; a far greater number of people on your list are likely to buy because they have just been paid than the number of people who buy because you sent the email before 10am on a Wednesday.

*In the UK at least

Know your subscribers...and let them know that you know them

I recieve emails from an networking event company informing me of their continual array of excellent events and high quality speakers. However, when I opened their email this morning I was disappointed to see the words 'Were you there?'.  The email read as follows:

"Were you there?

The fabulous atmosphere at our September event was electric to say the least! Were you there?"

As it happens, I wasn't there (mainly because I was in London, UK and the event is in Sydney, Australia) - but if I had attended, I would have been even more disapointed to know that whether I was there or not meant so little to them that they couldn't be bothered to segment attendees from non-attendees and send them different versions.

What I would have prefered to see would be something along these lines:

Email to Attendees:

"The fabulous atmosphere at our September event was electric to say the least! It was great to have you there and this month we have ......"

Email to Non-Attendees:

"The fabulous atmosphere at our September event was electric to say the least! We missed you - but hopefully you're able to come along to next month's event with guest speaker....."

Simple...not a huge amount of extra work, but a little bit of personalisation and segmentation (and dare I say it...relevancy) lends so much to the relationship with your subscriber....and goes towards building loyalty.

If it is absolutely impossible for you to know who attended and who didn't -then don't advertise the fact that you don't know and leave out obvious lines such as 'Were you there?'

Kath Pay
Managing Director
Ezemail

Maximising ROI without overmailing - how to determine the right send frequency for your email campaigns

The main difficulty in answering the “how much is too much” question is the fact that an email database contains people and what is too much for one person may be too little for another. The problem is compounded when you add relevance into the equation, because relevant and timely message are always welcome.

The key to optimising mailing frequency is to analyse customer behaviour rather than campaign performance. In other words, focus on the people and not the list.

Let’s say your monthly offer mailing has an average open rate of 25%, click rate of six% and generates £20,000. It would not be unreasonable to consider whether you could increase your revenue by sending out 2, 3 or even 4 campaigns a month. So which of these mailing frequency options is the right one?

The correct answer is all of them and none of them!

The problem with campaign based statistics like those in the example above is that they tend to make you base your decisions on averages. What you end up doing is trying to identify what the optimum mailing frequency is for the average person on your list.

What we have found here at Alchemy Worx is that when it comes to finding the balance between the need to maximise revenue and minimise email fatigue; it is vital to consider all behaviour not just the average.

If you crunch the numbers differently – as we do – what you will probably find is that hidden in the averages are a very small number of people doing most of the work.

Between 1 and 5% of the list open, click and buy almost every time they receive a message; in some cases they are responsible for as much as 20% of the revenue generated by that campaign.

40 - 60% of the list will interact with the offer mailing once or twice a year and 35 - 55% of the people on the list will not have opened or clicked, let alone bought off an email you sent in over a year; very few of these inactive people will ever open another message from you.

Armed with this understanding you could argue (relevance aside) the following:

• Too few - The first group is being under mailed and might actually appreciate being sent weekly offers

• Enough - The second group is probably be being sent about the right amount of messages, however it would be worth testing the impact of a higher frequency especially if the additional messages are lifecycle or event triggered messages.

• Too many – The third group is being mailed too much and a mailing frequency of zero might be appropriate

Relevance and timing have a huge impact on how an individual regards any given messaging frequency; getting this right (a topic in its own right) requires a deep understanding of customer profiles and the use of event triggers and lifecycle marketing techniques in your email programmes.

Don't forget wherever possible let your customers decide by giving them the opportunity to choose the frequency that best suits them at sign-up and unsubscribe via preference centre.

Dela Quist is CEO of email marketing specialists Alchemy Worx Ltd and Chair of the Events and communications hub of The DMA Email Marketing Council

The Email Holy Grail: Targeted, Relevant Messages

Are you emailing like its still 1999?

You are, if you think sending a single message to a rented database of email addresses is going to get you the high returns email is famous for.

Today, "batch and blast" email is out. Email users are long over the thrill of receiving commercial messages in their inboxes. They know what kinds of email they want to get, they know what they don't want, and they will do whatever it takes to keep unwanted email away.

Here's what they don't want: email that comes from unfamiliar senders, that comes too often, looks suspicious or contains nothing that interests them. When you become a bother, you also become history.

If you're lucky, they'll just unsubscribe. If they don't trust your unsubscribe or understand how to use it, they'll either ignore you or click the spam button in their email programs. Too many spam complaints, and the ISPs will either route your emails to junk-folder purgatory or block them completely.

How can you avoid this fate? By understanding that email isn't about you and what you want to tell your subscribers. It's all about them and what they want to receive.

That's why we say targeted, relevant email is the Holy Grail of email marketing. It means your recipients not only welcome you in their inboxes, but they anticipate you and they're more likely to open and act on them.

So, if targeted email is so great, why isn't everyone doing it? Because it takes more time up front to structure. It's not as simple as loading a single message into your list software, matching it up with the right mailing list and punching "send."

You need two things to create more targeted and relevant email messages: information about each of your subscribers (email preferences, interests, purchase history, how long they've been on your list and how active they are on it) and list-management software that lets you slice and dice your database into relevant segments that match that information.

So, you might need to restructure your email program to accommodate this more precise messaging. But, the effort is worth it, because targeted, relevant messages outperform batch-and-blast emails. Here are just a few statistics to chew on:

• December 2006 JupiterResearch study found marketing campaigns built around targeted email messages had conversion rates that were two to four times greater than untargeted campaigns.
• Open rates are on average 20% higher on segmented campaigns, according to MarketingSherpa's "Email Marketing Benchmark Guide 2008."
• Click rates in segmented campaigns were 5 times higher the last 90 days of the year (also MarketingSherpa).

Now, how to create this higher-value email? I won't kid you: It takes more time upfront to create targeted messages and to divide your database into relevant segments. However, once you restructure your email program this way, you won't have to do it all over again each time.

These strategies will help you reach the new Holy Grail of email marketing: targeted, relevant and timely emails:

1. Let subscribers decide what emails they want to receive from you.

This is the beauty of opt-in email marketing and publishing. Your recipients have already told you they want to receive your emails. However, not everyone wants the same message. Give subscribers a preference page where they can customise their email experience on these factors:

• Content: Some want your regular newsletter. Others just want promotions, product information or company news. Some want everything you can send them. Others want news or offers only on certain products of yours. Give them a checklist and let them choose what they want to receive.
• Frequency: Too much email is the single biggest turn-off email users report every holiday shopping season. Give scheduling options (2-3 times weekly, once a week, bi-weekly) instead of mailing whenever you have something to push.
• Format: These days, almost every email reader displays HTML email (with colors, graphics and pictures, often linked to your Web pages) correctly. Still, you should offer a text version too, for people who don't like HTML email or who read email on mobile devices that don't render HTML well or at all.

Use this information to create mini mailing lists within your database. Then, send custom-designed, messages to these mini lists. These emails will resonate better than a mass message mailed to your entire database.

Example: You sell pet products through weekly email newsletters and a Web site. Your preference page lets subscribers check off the kinds of pets they have and the information they want to receive.

A basic list would include all the animals you cater to: dogs, cats, birds, fish, reptiles, exotics, rodents, etc. If you really want to go wild and create the most personalised experience possible, ask the subscriber to give you the name, age, breed, etc. of each pet.


2. Send email messages that reflect your subscribers' relationship with you or their actions on your previous emails.

This is related to the first strategy, but it goes beyond merely letting subscribers tell you what they want. Here, you send emails triggered by their actions (or even inactions). If someone buys a product from one of your emails, or from your Web site, treat that person differently from someone who doesn't even open your messages.

Example: You email a message to all of your dog-owning subscribers, offering free shipping on a case of dog food. Subscribers who click on that message and then buy the product might be interested in other products in that line.

Target those customers with related special offers or, because it's a product that will eventually run out, send reminders that it's time to buy again.


Summing Up: Targeting Takes Time But Pays Off

Targeted email campaigns do require more effort up front than "batch-and-blast" messages. But, they also bring in higher returns from a happier email audience, which makes the effort worthwhile.

Kath Pay

Managing Director

Ezemail

Email Marketing: Get Personal with Your Customers

Aberdeen Group (a US based research company) has published a new report entitled ‘Email Marketing: Get Personal with Your Customers’. The report covers the how, why, and what of email marketing personalisation along with the challenges and the benefits.

The report is well written and supported by a case study from within the retail sector and many charts and very useful nuggets of information spread over its’ 25 pages.

The report can be accessed free of charge until 29^th August 2008, here.

Latest research suggests your subject line test strategy needs a complete overhaul!

Last week infobox, the DMA Email Council’s newsletter featured the UK launch of a free white paper on subject line length published by my company Alchemy Worx.

For those of you who haven’t had a chance to read it yet the main findings of the research were pretty startling and run counter to conventional wisdom.

The research suggests that although subject lines with 60 characters or less make more people open your message (the traditional view) these people are less likely to then go on and click on content or offers within the message than people who open an email with a longer subject line. More opens = less clicks! There seems to be an inverse relationship between opens and both click and CTO rates.

As you might expect, we monitor a large number of UK email campaigns, from a cross-section of sectors and companies including British Airways, figleaves, Apple, Amazon JD Sports and Reuters; so I thought it might be interesting find out what subject line lengths email marketers are using and was astonished to find out just how are following conventional wisdom!

Out of 700 subject lines we monitored in the last 90 days, the vast majority—87% of them — were under 60 characters in length. A further 7% fell into the ‘dead zone’ between 60 and 70 character where neither opens nor the CTO rate is optimized, and only 6% of the subject lines were over 70 characters long and therefore likely to optimize click and CTO rates.

Does his mean that everybody out there is only interested in opens and doesn’t care about clicks?  Perhaps we could conclude that email marketers, having extensively and regularly tested longer subject lines, know for a fact that they don’t work?

Somehow I think not.

What’s more likely to be the case is that as an industry we’ve done such a fantastic job of believing the hype that we have stopped testing outside of the accepted norms.

Our whitepaper also found subject lines with a higher word count also optimize clicks and CTO rates. So how do the numbers break down when it came to word count?

The numbers are equally amazing. Only 13% of subject lines monitored contained above 10 words—where clicks and CTO are optimized. 60% fell into the ‘dead zone’ of between 6 and 10 words, where neither clicks nor opens are optimized; and 26% of the subject lines contained fewer than 6 words, and therefore optimized open rates.

What I have learned from this exercise and would like to share with you all is that email marketers need to completely overhaul their subject line test strategy:

● Subject line tests should be more granular—long and short just isn’t good enough. Subject lines need to be broken down into more character groupings (1-10, 11-20, ...91-100).

● Introduce word count testing. Words are a much better way of conveying meaning than characters.

● Assess the impact of the number of propositions contained in the subject line on your campaign performance.

● Finally, open rates are just a small part of the story. Your tests should assess the impact of subject lines on clicks, CTO rate and conversions, as well as sales.

My greatest fear is that the people reading the whitepaper will be looking for a simple answer such as “when it comes to email subject lines, short is best”, when in fact the central message is keep searching, keep optimising and keep on challenging assumptions.

Dela Quist
CEO
Alchemy Worx