Blog Box

We create a lot of emails for people, brochures, newsletters, once-off invitations and more.

Usually these have a unique look and feel and require some kind of customisation, but with the world of html emails, consistency and display reliability is a hard thing to achieve thanks to the wide variety of archaic email clients being used.

This article here provides a handy little checklist of CSS items that are available to you, but it doesn’t not list the new hotmail “Live Mail” and as there are differences in the way CSS is handled we must press on.

The best article we’ve come across is written by the fantastic guys at Campaign Monitor right here. What are some of the main points we took away? We’ve distilled it way down for you anyway.

As with many things coded for cross-client compatibility the weakest link becomes the greatest determiner in what you can and can’t do in an email design. Especially as it’s extremely difficult, nigh impossible to segment your recipients based on what client they’re using. So, when it comes to sending through those shiny HTML emails Gmail and Outlook 2007 become those weak links that imform the way most of your design is hard-coded into the email. Question: So what do you do?

Answer: Table based inline css, and check the list of inline-options that you wish to use are supported (do what we did and printed off the Campaign Monitor’s article and stick it on your wall/foreheards).

Sure, you may have to spend a few more minutes adding code and feeling like a Neanderthal coder, but at least it’s going to deliver consistent brand integrity and professional results to your valued publics. Onward Ho!

~Rob

P.S. I just found this super article if you want a more comprehensive checklist when writing those HTML emails. It’s pretty old, but gives you good info on how to satisfy GMail and has some good practices in there. You know, just in case you had the itch for some helpful extra reading.

When Rob and I are creating websites, presentations and marketing material, or anything that needs to be printed we deal with a lot of artwork. Sometimes an organisation or company can send through all the photos, logos, images, illustrations and they don’t have time to sift through all of them making sure they are the best possible quality. That may even include their own logo! Long ago in the midst of time - their logo was created by a graphic designer. Handed down from person to person it has become a horrible mess of pixels and has started to look highly unattractive. The problem? Well no one has the good looking logo anymore, the graphic designer is long gone and you don’t have time to re-create it (that’s if you even know how!).

So what do we do?

Well there’s this guy we know. He’s fast, efficient and great on the drums. Another thing he does well is drawing, he draws logos for a living. Apparently his specific drawing skill is called ‘logo redraw’. Not because he has to draw it more than once to get it right, but because he can take those ugly, pixelated-like-a-criminal-on-tv logos and make them look as good as or better than new. It doesn’t take him long, in fact he said to me once he loves getting things done as quickly as possible and then moving on to the next job. Maybe efficient was the wrong word.

We’ve known Marty for a while and we decided to stop keeping him secret and tell people. We highly recommend his work - he has done some fabulous stuff for us in the past and because it’s so quick we know he will get you out of tight corners if the situation arises.

One of our clients regularly sends out tens of thousands of emails to their subscription database. We recently conducted an a/b split test (multivariate testing) on one of their emails to see if we could see any variances in the behaviour of people reading the emails.

A/B Split Test or Multivariate Testing?

The the email system that Brown Box employs to send out html email campaigns for our clients, we are able to monitor a few things. Some of the important statistics include the read rate, unsubscribe rate and link rate (which links in an email are clicked on). When an email is sent out - we want your read rate to be high, and your unsubscribe rate to be low. Apart from making sure that the content is useful to your customer and relevant there are few things you can do to see how useful your email actually is.

One thing you can do is multivariate testing. That is, sending out slightly different versions of the email to your database. Each email generated is either version A or version B of your email. That way you can measure the possible differences between things like link text, email titles, images and copy. In this case we changed just one thing: the email title.

Are Email Titles Important?

Naturally an email that has great information in it will be captivating and hold the reader’s attention, but to get the reader to open the email is itself a challenge - especially when you are sending out emails about your product/service on a regular basis. With this recent email we had two titles.

1. Petbarn : Pet Poem Competition; or
2. Petbarn : Poetry in Motion

Version 1 was quite obviously a poem competition email whereas the title of version 2 didn’t immediately let on about the contents inside. Would this make a difference?
With the email send being in the tens of thousands, here are some percentage statistics to show you the difference (if any) between the two emails.

1. Pet Poem Competition
   • Read Rate: 14.69%
   • Unsubscribe Rate: .44%
2. Poetry in Motion
   • Read Rate: 15.2%
   • Unsubscribe Rate: 0%

Quite simply - there are more reads for Petbarn : Poetry in Motion and less unsubscribes than the version titled Petbarn : Pet Poem Competition. That is easy to see and quite a simple test to conduct with a good number of emails sent for testing. While the read rate for both may speak more about the company itself and the willingness for its customers to accept marketing material, we think that the lack of an unsubscribe rate on version 2 may be related to the title. The hypothesis at this stage is that keeping the customer interested from the start may be a factor in keeping their long term attention (and thus their long term value to the company sending emails). We will continue to develop this idea over a series of emails to gain a true impression of whether or not this is correct.

What do you think?

Interested in getting your message to your customers effectively? Brown Box can deliver your email marketing with extremely competitive rates and great personal service. We can even advise you on the best way to setup and manage your campaigns. contact me today to find out how

Bad publicity can be hard to overcome - especially if your business relies on reputation for 50% of its income.

In December 2007, Kangaroo Island had some pretty horrible publicity: widespread fires on the island (and media coverage to equal the fires). Now that the fires are out - so too are the tourists. The ABC reports that tourist bookings have dropped dramatically. With the economy of the island being 50% tourism this is a big issue - so is the reliance of the summer period to sustain through the winter down period. Hard times may be ahead for some.

The usual response to this kind of publicity is to communicate your story with the public. With more and more people researching online, your website is an important place to make sure this happens. Kangaroo Island promoters have indeed done this. The ‘official’ Kangaroo Island website have put a notice on their front page:

‘Coming to KI and concerned about the fires? Don’t be. Download K.I OK Map here (PDF 467KB)‘

Download the pdf and you’ll see a well prepared map with fire damaged areas highlighted and the reassurance that everything is ‘OK’. I think this is a good thing - checking the site for up to date news was the first thing I did when I heard about the fires. However I think there are a few things that could be done here to maximise the effect of the communication efforts:

1. Make sure it’s widely visible. One link on the front page is not enough - people will enter the website in other ways and may still have this issue in mind. Under the menu is an appropriate place to insert a site wide link/text/graphic.
2. Combine the power of the Internet. Ozone Hotel is a website of a business on the island yet there is nothing addressing the issue there. Ensure that as many other businesses are on board with you promoting your good news. A simple link on other websites promotes both the freshness and unity of all businesses involved. Both are important in helping people make a decision on travel.
3. Show the reality. I don’t see how showing people what they will get is a bad thing. If people go to the island not knowing there have been fires they may be disappointed. If they expect everything to be they may be disappointed.
4. Tell stories. Allow the personality and people of the island to tell a positive story. A personal response can communicate a greater deal of trust, and persuade those who doubt that the island is still a great place to go. This is done to a degree on the pdf download - where a ‘personal’ appeal is made, I believe a widening of this idea could help.

In the end - bad publicity is still going to affect the island and people may have their minds made up without even looking at Kangaroo Island related websites. However there are those who can still make up their minds. Don’t let them make their minds up without hearing your side of the story - ensure they can hear it. When/If this happens to your business - the important thing to do is to use the news. Be loud and clear and let people hear your side of the story.

Our ad was “clicked out” within 2 hrs of starting the campaign. So much for free ads on Facebook!

The Last few days

Over the last five or so days we have continued to monitor the ad (you can see it to the right). From the start we had our budget set at $5/day - that stayed the same throughout. However each day we gradually lowered the budget of our price per click from the begining $1 per click to now around $0.30 per click. Each time the ad was clicked out within a couple of hours of it starting again. We were disappointed to say the least - that our ad was so ‘popular’. We were really hoping to have a fair amount of time between the ads being displayed and the ad budget being filled. Anyway let’s get to the results.

• Each day we reduced our budget, our ad had a higher number of clicks;
• Our click through rate was between .05% and .02%;
• On average, the ad had 48931 impressions per day. The highest being 74976, the lowest being 22578.
• The average cost per thousand impressions (CPM) was between $0.22 and $0.08

For the purposes of testing how you might get thousands of impressions without many clicks - this might seem like an okay result. The lowest click through rate (CTR) was 0.02% - virtually no clicks compared to ads served. The only problem was that the ad was up for just 2-3hrs before the budget had ran out. We were hoping for a lot more time than that. Even on the days of $0.30 per click our ad budget was filled within a couple of hours. To mitigate this, I had moved the CPC budget to $0.25 for a few hours. This again had its own problems with hardly any ads being served (maybe 2000 in a few hours) but still clicks on the ad.

Visits to the site

Another hopeful result of advertising Rob’s desperation to get a girlfriend - was to get some ‘freebies’ on his site. No I’m not being coarse, what we wanted was people visiting Rob’s site without clicking on the ad itself. Basically free ad conversions without clicks. We certainly did get them I’ve put them below - you can see the higher number visits to Rob’s site compared with clicks on the ad.

The highest day of clicks also coincides with the highest day of visits. On average the site draws 3-4 visits a day (no traffic) so to see it go to 46 visits (28 ‘free’ visits) was an encouraging sign. Though this cannot be only related to the ad itself. Word had spread that the ad was up - so some people may have not seen the ad but still visited Rob’s site.

Conclusions

• Although being written from a first person perspective and following advice from other observers on its image - the ad still did not receive a high CTR. We were hoping for this anyway - so this can be seen as a good result. However from other people’s advice we were expecting a higher CTR. Most likely taking the option to pay per 1000 impressions is going to cost you a lot more than the potential result you will get from paying per click.
• The main problem we had with this ad was achieving a high impression rate with a low CTR over a long time. Either you don’t get seen and the ad is up for a lot longer, or you do get seen and your ad is gone very quickly. You might get relatively free ads served up - but only for a short period of time.
• We are yet to see if this actually works. That is, that Rob gets a girlfriend. He has had a few contacts because of the ad - let’s see what happens.

~Andrew (email me andrew at brownbox dot net dot au).