Malleable Musings

June 17, 2009

DM / Email / SM … the same rules apply

Filed under: Emails, Social Media, Trad marketing methods — Brendan @ 10:22 pm

I got a piece of direct mail yesterday that seemed to bring together a lot of different ideas.  My thinking was undoubtedly influenced by the #digitalleap twitterstream, which I dipped in to on Monday. Also, many of the arguments about social media being just another form of communication were fresh in my mind having listened to the excellent Six Pixels of Separation #159 podcast on the train home yesterday evening.

As an aside, personally I find JayJay’s notes about Digital Leap to be a far more detailed and coherent summary of the day than the twitter #hashtag.  To me it doesn’t matter that this is one person’s point of view and she’s espousing it away from where the main online conversation is likely to take place.  Why? Because I trust her point of view and think the permanence, readability and searchability of JayJay’s notes to be superior.  I think that there’s probably a whole blog post just on this topic – but for now that will probably have to wait.

Outside of CAFOD mailer

Outside of CAFOD mailer

Anyway the direct mail piece was from a charity that I’ve given to in the past.  What I loved about it was the language used, how on message it remained and the calls to action that were used.  It got me thinking about how similar this sort of communication is actually to the digital methods that many would say have superseded direct mail.

This is the mailing as I received it.  Immediately it had done all of the right things.

1) Their targeting data was good – and their delivery mechanism was sound. (The piece was delivered and had got in to my hands. I presume I was targeted through the Gift Aid scheme in which you provide your addresses.)

2) I recognised the sender (branding).

3) The subject line attracted me (alongside the crunch that told me there was something other than paper inside the envelope).

4) All of this led me to open the communication (envelope).  If steps 1-3 weren’t in place the mail either wouldn’t have got to me or I would have thrown it straight in to the recycling bin.

Compare this to the cardinal rules of email marketing.  The only significant factor missing is receiving the message at the right time.

This was what was inside of the envelope.

intInside of the envelope

What struck me first of all was the consistency of the message.  On the envelope I’d been warned that I would be asked to tell them what I think and the contents delivered.  This core call to action should sound familiar to the social media fundraising mantra of building influencers and friends and understanding them and their commitment to your cause as a first step.  It’s certainly no new idea, more like fundraising 101 – people give to causes they believe in that are articulated well and to their friends.

Anyway, the entire pack had been carefully designed.  I thought the addition of a Fairtrade teabag and the second line of the headline “Will you tell us over a cup of tea?” was a masterstroke.  Throughout the covering letter and the survey the language used was chatty and informal and the graphics complemented this.

Every opportunity was used to tell the little interesting stories that fit with the brand messages that the charity is waiting to push.  It informed and cross-sold, but at no point did I feel I was being pushed.

I was left in no doubt that the purpose of this mailing was predominantly a listening exercise. The paragraph requesting donations was actually the top paragraph of page 2, and it was a simple understated, “when you’ve completed our survey, would you consider making a donation” backed up by the following line which was underlined, “We really do want to hear your opinions, whether you choose to make a donation or not.”

In the UK, I think this approach fits the university sector well.  It’s actually the approach used by my alma matter, the VC wrote to me saying something about how she wanted the alumni to help keep the university honest.  A clear message that as alumni that we are the probably the most visible aspect of the brand that the university has.

In the institution in which I have worked for the past seven years, I’ve long argued that our alumni base is probably more important to us as influencers than as a cash cow to be milked.  And I also argued that setting up our alumni association from scratch one of the key things we needed to do was to locate and listen to our alumni before anything else.  What’s more we needed to listen harder to our alumni who were distance learning students who had never set foot on the university than might be necessary for more conventional students.

We built the numbers very quickly and did a lot of listening and surveying but we always had a problem with segmentation.  Our surveys were rudimentary and nothing like as elegant as the one I got in this mailing.  When I get a chance I’m going to scan and save it here as it really is an exceptionally well executed.  It will provide a tonne of information that the charity can use highly effectively to both reduce their costs in communicating with me, recruit me as a volunteer and improve their likely return by targeting me with the key messages that matter to me.

However it’s also written in language that encourages and reinforces all of the brand messages so I’d imagine quite a few people will get to the end of the survey and will feel compelled to add a donation.

Thinking about it after a while I realised that there was one extra thing that could have been done, but for me this was the only fault.  This minor thing was to separate the email address from the donation box.  Ideally they should have also used a quite statement about helping them reduce their communication costs / or communicating electronically to save paper.  The reason for this is that some people will take the charity at their word and just fill in the survey, but not realise the savings that the charity can make in dealing with the person electronically.


March 26, 2009


Filed under: Emails, iTunesU, YouTube — Brendan @ 3:12 pm

I got this email earlier today from Apple.  I wonder how many other people did too.  It’s excellent branding for the University of Oxford isn’t it.

Now your favourite destination for music and movies is also a great place to
entertain your brain. iTunes U in the iTunes Store offers free audio and video
content from top universities, famous museums and other cultural institutions.
So whether you want to learn from the world’s leading thinkers, get a sneak
peek at an upcoming exhibition or simply brush up on your Spanish, iTunes U
makes it easy. To see for yourself,
explore iTunes U.
Some products and promotions are not available in all countries.
Apple Sales International, Hollyhill Industrial Estate, Cork, Ireland.
Company Registration number: 15719. VAT number: IE6554690W.

It reminded me to have a look at the Shakespeare Found video on Warwick iTunes.

Then later on in the day I was pointed towards another university (the OU) that has a long standing relationship which dates back to the 1960’s with another media provider (the BBC).  Interestingly the OU is the only other logo that can easily be noticed

As a marketing person primarily focusing on student recruitment what I love about this sort of collaboration is the reach this type of approach provides especially for what can be quite niche courses.  I’ve long argued that releasing open educational resources through strategic partnerships and providing sharable content are key promotional routes, particularly in the distance education world.

However it also got me thinking about the sorts of changes that we’ve seen in the past few years and wondering what might be in store in the near future.  I think things like CMS 3.0 and the eventual introduction of university APIs may well provide some pointers, but backing winners is always difficult.

P.S. launched last night? Some other channels include

January 9, 2009

Email = one way communication (why Tracy Williams is a superstar!)

Filed under: Emails, International Student Recruitment — Tags: — Brendan @ 12:58 am

It’s been an odd day – that probably deserves half a dozen blog posts on a range of different subjects.  However I thought I’d write a quick reminder about the problems of relying on email for communication, especially in terms of process management.

The key reminder is:

You can’t assume that because you’ve sent an email that it’s been read or even received.  Silence is not an answer – it’s silence.

An open doesn’t equate to a read, a click doesn’t equate to a read.

You can’t know an email has been delivered, you can’t know an email has been read.

I think in the case below the issue is the wrong email addresses being used but there are many reasons why a correctly addressed email may not reach an email account.  I’ve written more about this here.

A side learning point for me is:

Start from the other person’s point of view, aggressive posturing won’t help forge a connection – look for solutions, life is all about win-wins.

The background story:

Every year, usually in Feb, we attend the Hong Kong Law Fair – essentially a graduate recruitment fair organised by a group of the big law firms in Hong Kong.  This year we’ve already booked non-refundable flights, arranged and paid for our freight, arranged hotels and informed potential students that we’d be there.

Unfortunately late yesterday we found out that we don’t have a space at the exhibition.

I don’t really know what has happened, some things you just have to put down to bad luck.  It looks like a couple of critical emails went to the wrong email address – at least one email went to an email address ending .ac and another went to instead of  – however there may have been other reminders.

However it meant that this year we didn’t receive any requests or reminders for payment.  Bizarrely though, by chance, the temporary member of staff we had covering our vacant Exhibitions Co-ordinator role was in communication with the organisers, again by email, asking about venue, stand furniture, dates etc.

I should at this point say that the organisers are a group of volunteers drawn from the law companies.

So lets think about what’s going on as percieved by each side.

Our point of view: everything’s fine but we’re still awaiting some information and an invoice in order to arrange payment.

The organisers point of view: these people keep making requests but aren’t committing to paying and are ignoring us – we’ve got enough attendees in any case and we don’t need a headache.

The solution: I had to handle it this morning (as no one else was in the office) and thought it best to phone someone direct and talk through what had happened but as you can probably tell from my side learning points, I didn’t handle the conversation that well.  However despite my shortcomings Tracy Williams who is Head of Graduate Recruitment at Lovells in Hong Kong was extremely understanding and is on the case to see what can be done. With luck I’ll find out tomorrow morning.

January 7, 2009

Why I don’t email from my work email account

Filed under: Emails — Brendan @ 8:04 am

Because I didn’t run the test I wanted to run last night I ended up running an alternative A|B test.  The summary five hours after the email was sent is shown here:

Results from an A|B test of an email

Results from an A|B test of an email

Once you’ve got your head around that the line graph (surely a bar chart would be more intuitive) you’ll see that it’s only a small test just less than 370 emails in total, sorted by domain and then split in to two segments based on whether the row number that they occupied was odd or even.

This is the first time for a while that I’ve run this test and I did so as I wanted to check that open and click through rates were still better when I didn’t use the the standard University domain to send email.  (I thought I’d better check following this unfinished post).

The emails being sent were identical and all efforts were taken to ensure that the segments were as even as possible (similar numbers of emails from the same domains).

So why is an domain better – well for whatever reason changing the DNS records on our domain is not a high priority.  So registering the domain with a third party ISP and adding DNS records myself seems to be one of the best £10 I’ve ever spent.

There’s a good primer on authentication on the DMA website.

January 6, 2009

Go ask Apple

Filed under: Emails, iphone — Brendan @ 9:21 pm

I felt that I had to blog about my very interesting experience earlier today / yesterday with Constant Contact.

I’ve signed up for a guest account with them and was going to run a test campaign this evening through them. If it had been successful I could well have been pushing our email marketing out to them.

Constant Contact has got a nice simple interface with a WYSIWYG editor and an API that seems to do what I want and to cap it off we’ve just appointed a new member of staff (started on Monday) who is likely going to be writing a lot of the email who is used to using Constant Contact.

Screen shot of a Constant Contact Template on iPhone

Screen shot of a Constant Contact Template on iPhone

However doing some testing I noticed something that I didn’t like in their footer. Look what happens when you send a message to someone using a free webmail service (e.g. yahoo or gmail) on an iPhone.

Why does it do this? Well at first I thought it was entirely to do with the code used to insert the Constant Contact branding images, which is in a footer that I can’t edit.  I noticed that there are no width attributes on the img tags. Without the width attributes the alt text will expand to fill whatever space is available, ie. in this case the entire width of the iPhone screen causing disruption to the top of the template.  However on closer inspection I don’t think it is this but I’m not enough of a coding guru and don’t have the patience to work through the nested tables and div tags to find out what is actually going on.

Anyway I used their live chat facility to ask a few questions. The full transcript is saved on their support section. I could paste it here but I don’t think it would add anything. Whilst I was having a live chat a sales person from Constant Contact rang me. Now I don’t know about you but if I’ve got the option of voice or IM contact I’ll opt for voice nine times out of ten so I excused myself from the Live Chat (the sales person apparently didn’t know I was chatting online) to concentrate on the phone call.

Side note: Live chat is great from an organisational point of view (good blog post highlighting benefits) but it doesn’t satisfy me as a customer – but that’s probably for another blog post?

The gist of both chats (online and voice) was that I could probably get the footer changed for $100 if I upgraded to a paid for account. But usually the turn around time for a custom footer is 2-3 business day. The sales lady I talked to passed me through to one of their “senior” technical people to see if he could do anything.

So I explained the situation – explained how we do a lot of work out in emerging markets in which mobile is important and how our %ges of mobile users within our site visitors are at least double or triple stats like this and how they are growing fast.  I also explained that I expected email to be even more likely to be accessed via mobile phone technology than web browsing. I then offered to send through a screenshot to explain the situation, which I did straight away.

I got an autoreply with the following (very helpful?) links:

Adding an image to an email
How does Outlook 2003 impact your campaigns?
Match graphic dimensions to template layout
Classic Wizard: Inserting extra images
Adding stock images to an email

and then an hour or so later I got the following email.

Dear Brendan,

I see that you already spoke with our of our representatives regarding this issue. I would like to inform you that most handheld devices (PDA) and iPhone have webbrowsers that aren’t fully compatible with HTML emails. This may result in your emails to not appear properly. When we create our templates we try to make them appear properly across most widely used web browsers (targeted to PC’s and MAC). However since the issue is relate to the compatibility with the browser/client in iPhones you must report the matter to Apple support directly. In the meantime I will also forward your email to my engineers to see what can be done to make all our HTML emails compatible with iPhones.

I’m very sorry this was not exactly the information you were hoping to hear; but if I can be of further assistance, then please don’t hesitate to reply to this email.

Thanks for using Constant Contact!

Constant Contact Customer Care

For me reading this email was one of those wake up and smell the coffee moments…

1. It made me think a lot about the steps we’re taking in customer service and whether we have enough mechanisms in place to really help a customer.
2. It reminded me how I want to be working with organisations that are looking to the future and are easy to deal with (N.B. despite how this post may come across I think that Constant Contact may still fit this criteria).
3. It reminded me how much testing is actually needed on things like this. My guess it’s not just the iPhone that’s impacted it might well be anything that displays less than the default 600 pixel width templates that Constant Contact are currently using.
4. It also reminded me how easy it is sometimes to phrase things badly…. e.g. it’s not that the templates aren’t compatible, with the way I think an increasing number of people will be accessing their emails, it’s that:

most handheld devices (PDA) and iPhone have webbrowsers that aren’t fully compatible with HTML emails

So what’s next – well I don’t know. I haven’t got any emails planned for a while but maybe next time I’ll give myself enough time to tinker with their CSS or I might pay $100 to get a custom template designed (if I remember that is).

January 4, 2009

More on email (unfinished post)

Filed under: Emails — Brendan @ 1:10 am

Just reviewing some stats on a couple of emails that were sent out just before the Christmas break. Interestingly I sent to the same email list a week apart, but from different domains. Email A came from our main domain whilst email B came from a domain we set up specifically for emailing purposes (i.e. SenderID and DomainKeys have been set up alongside SPF).

Email A was sent at 4pm UK time on the 10th Dec whilst email B was sent exactly a week later. Long enough to make a difference I know (no two emails are ever the same) but the closest time period I’ve ever had for emailing the same list, so I thought it would be worth a comparison.

The basic stats were:

Email A %’ges Email B %’ges
All recipients 15,495 15,459
Opens 5581 36.018% 4,455 28.818%
Not Opened 9,348 60.329% 10,453 67.618%
Bounces 566 3.653% 551 3.564%
Clicks 1740 11.229% 156 1.009%
Unsubscribes 3 0.019% 7 0.045%

Now OK the content isn’t comparable and consequently I’m not going to try and compare click or indeed open rates because of this, Instead I wanted to look at the bounce rate.

My ESP allocates bounces into various types of soft and hard bounces. A hard bounce where the message is considered permanently undeliverable, but the recipient’s server has not provided a specific bounce reason. Hard bounces are automatically removed from your subscriber lists – so hard bounces. There were 34 hardbounces from email A which were removed from the mailing list for email B.

In email B there were 14 hard bounces (two of these email addresses had soft bounced in email A). In terms of domain distribution of these 14 hard bounces, 10 were free email accounts, e.g. Hotmail / Yahoo etc which presumably had expired in the week between the two sends.

I’m not too worried about the other four hard bounces as there were opens or clicks of email B from three out of the four domains.

The soft bounces are then categorised as follows:

Email A Email B
2 Auto Reply (AR) – These are soft bounces caused by an automatic response from the recipient, for example “Out Of Office” messages. The email is still actually delivered to the inbox, and once the subscriber opens the email (and is recorded), the bounce is removed from your reports.
169 161 General Bounce (GB) – The email server could not deliver your email message, but the bounce processing tool could not determine a specific reason for the bounce. Normally that is because the bounce message from the recipient’s server was very broad. We treat these as soft bounces. Example: “Subject: Undeliverable mail”
7 5 Mail Block – General (MB) – Indicates that the recipient’s email server is blocking email from our email server. Example: “550 Message REFUSED by peer”
10 11 Mail Block – Relay Denied (MBRD) – Indicates that the recipient’s email server is blocking email from our email server. Example: “551 relaying denied”
3 5 Mail Block – Spam Detected (MBSD)Indicates that the recipient’s email server is blocking your email because the message appears to have content that looks like spam. Example: “550 Possible spam detected”
186 190 Soft Bounce – Dns Failure (SBDF) – The email server is temporarily unable to deliver your message to the recipient email address because of a DNS problem. Example: “Host is unreachable”
82 82 Soft Bounce – General (SB)The email server is temporarily unable to deliver your message to the recipient email address. Example: “Connection timed out.”
71 73 Soft Bounce – Mailbox Full (SBMF)The email server is temporarily unable to deliver your message to the recipient email address because the recipient’s email box is full. Example: “Mailbox over quota”
4 8 Transient Bounce (TB) The email server temporarily can not deliver your message, but it is still trying. Example: “Warning: message still undelivered after 4 hours. We will keep trying until message is 2 days old”

I then ran some queries on soft bounces by domain and got the following data for the top ten domains that bounced: 82 85 but one repeated – see below 20 20 15 15 8 8 6 6 6 2 seems to have improved dramatically 5 9 but four repeated – see below 5 5 4 2 seems to have improved dramatically 4 0 seems to have improved dramatically

In looking at these I then spotted that on Email B six email addresses were double counted (it looks like there was no double counting on Email A). One of the Autoreplies was also reported as SBMF and four TB’s (all on the domain) were also reported SBG whilst a further TB on a different domain was reported SBMF.

In total 18 email addresses that bounced on Email A did not bounce on Email B.

Whilst 11 email addresses that bounced on Email B (due to SB, GB and SBDF reasons) and a further 3 for spam detected did not bounce on Email A.

Having looked at various other metrics my conclusion is that Email B has probably performed marginally better for deliverability (but not by much). Howver from a design point of view Email B was always more likely to be caught by spam filters.

I’ve decided I need to investigate some of the worst performing domains and have decided to run a couple of A/B tests on the same email to see if it tells me more.

December 12, 2008


Filed under: Emails — Brendan @ 8:37 am

Have spent quite a bit of the past couple of days looking at ESP’s.

I’m not searching particularly hard and looking to make a decision next week or anything like that. But as previously mentioned we’re about to update our databases so now could be a good time to think this all through. I guess what I’m looking for is a combination of:

Access to an API – to tidy lists with our databases / or some other way of thinking this through
Deliverability – although recognise this is as much an issue for me as an ESP
Ease of use, meaning
– Features
– How intuitive it might be for others to use

I had a phone call yesterday from Brandon at Bronto (a very nice salesman) who talked me through their product and sent me a nice case study from Wofford (Kyle James says nice things). Their deliverability and the features seem good (I liked the in built SpamAssassin integration and that there were so many configurable fields) but they are more expensive by about 10-15% than the current ESP I’m using. I’d think I’d also need to do a fair bit of education of the clients who would be doing the email sending, however I recognise that this would probably be inevitable in any case.

In terms of cost, one of the major issues for me is the $:£ exchange rates. The £ keeps falling (against the $ and more especially the Euro) which is great for business but it means that if I’m having to pay in dollars my budget is being spanked. What’s worse, if all of the economists I listened to last week are right, is that the outlook for the £ is bleak over the next year. Which means I may need to be looking at solutions that are charging in £’s and currencies like the AUS$ which will probably share the same sort of fate as the £.

We’re currently not a large volume email sender – only approx 2m per year at present – however this is growing rapidly – across all three of the main areas that send email – recruitment, alumni and student services side.

In terms of other providers that I’ve looked at the list is long, but two that I investigated worked on an alternative business model based on a number of contacts that we have: Relatively cheap, based on the number of contacts but monthly mailing limits to 6 times the list size (could be a constraining factor) also I can’t see anything about an API and have questions on deliverability? similar to above, although not limited to 6 times list size – also have surveys included. Although costs could be much higher than e-shot.

In any case I think I’ll need to do a bit more research to work out what will work best for us, and as I said before at the moment there’s no rush.

December 4, 2008

Xmas greetings

Filed under: Emails, iphone — Brendan @ 8:52 am

I thought I’d just make a short post about a Xmas greetings I got from Purdue yesterday.  I’ve been on their mailing list  since I signed up at a Linden tours fair, in Singapore, a few years ago.

This is how it displayed on my iPhone which is how I’ve been accessing my emails this week.

The image on the left is the screen grab of the email I received.  The right hand image is a screen grab is the web page that the email directed me to. It’s great on a flash enabled machine – see for yourself.

I think it just goes to show the importance of cross platform testing especially as more and more people start accessing the web using handheld technology.  I guess it also shows the importance of updating and checking your database.

December 2, 2008

Sex sells – it just doesn’t email well

Filed under: Emails — Brendan @ 8:44 am

I’m signed up on quite a few mailing lists. Inevitably this means I get a lot of junk that I end up filtering, but occasionally it leads me into conversations related to my areas of interest that get me thinking about things in different ways.

A recent example was from an email I received from a self-help group that predominantly markets into UK schools.  A particularly vexed marketing manager from the Family Planning Association was asking,

As the country’s leading sexual health charity it is virtually impossible for us to avoid using words such as ‘sex’, ‘sexual’, ‘sexually’ etc in our promotional e-mails – even if I manage to not use them in e-mail message itself, our strapline is ‘putting sexual health on the agenda’ and it is embedded in the logo and therefore always goes out in every promo e-mail we do.

I’d really like to hear your views and advice on

1. how having a strapline like that impacts on the deliverability;

2. if you think a strapline containing word ‘sex’ would be worse one with ‘sexual’; and

3. if you think would it make big enough a difference if we were not to have the strapline in the logo in promo e-mails or even if we had a different strapline without any of these words.

This could almost be a textbook question on deliverability and spam filters, couldn’t it?

Plenty of good advice was given both in the form of links to other resources (one of the better ones surprisingly on the Microsoft Office site) and also verbally in relation to whitelisting, ensuring copy was as free as possible from language that would trigger spam filter rules, utilizing graphics for the logo and strapline without text, using very short snippets linking to web content, using spam scoring tools eg. Dotmailer’s spam tester, Spam-Assassin in advance of the mailing. Etc, etc.

However the best advice I think was bear all of the above in mind but test, test, test.

What the post did though was it got me thinking about what do I do to ensure deliverability.  OK I’m sending from a clean domain using SPF, where the MX records have been altered for DomainKeys and SenderID.  I usually include a link about adding our domain to a safe senders list.  I don’t segment my lists to see if we have deliverability issues with certain domains. I’ll occasionally run the email through a spam testing tool if ever I’m unsure but I probably don’t do enough to ensure that other people sending mail from this domain think in the same way.

So I guess I ought to more…

December 1, 2008

You must use our email….

Filed under: BTP, Emails — Tags: , — Brendan @ 10:40 pm

lsesweekemailrepIt’s been just over a week since I was involved in what I consider to be one of the very worst email campaigns ever*, so I thought it was a good time to reflect.

The graphic on the left shows the stats from this email – a week after the email was sent (although the open graph just shows opens on the first day).

Notice how big the mailing list is – 18,000+

And then look at that open rate  <1% – How appalling!

So what’s going on – well let me give a little background.

Up until recently the university didn’t provide email addresses to students.   These were finally introduced with a student portal system at the beginning of this academic year.  This particular email was sent to 18,000+ students at their new university email addresses.  So what the poor open rate really shows is that the student email accounts are not being used that widely yet.

We’ve probably not done enough communication with, or education of students.  Oh well – that’s probably another job for next  week!

* Why the worst email ever – well it’s wasn’t just the poor open rate.  There was also a mistake with a URL in the email (a line break somewhow crept in to the HTML) and it was only noticed that this caused a dud link after the email had been sent.  However it was dealt with very quickly and efficiently by the people at Campaign Monitor who have also promised to update their parsers to ensure that this sort of thing doesn’t happen again.

For info, I’m currently exploring whether other ESP’s (Email Service Providers) are likely to be better for my particular circumstances but at this point in time I’d strongly recommend Campaign Monitor.  They’ve just introduced a new interface which I haven’t really had much chance to explore properly yet – although the reporting is looking great and if the WYSIWYG support has been introduced as promised my life could suddenly have been made a whole lot easier.