Tracking your marketing emails

12 March 2009 by Martyn Hoyer  
Filed under Email marketing

When you send out an email marketing campaign, you’ll probably want to know how well received it has been. After all, that is one of the advantages of using email instead of snail mail for your marketing. How many paper mailshots have you sent that haven’t even been opened…? I don’t know either!

Using email tracking, you can not only find out who has opened your (HTML) emails, but you can also find out what links have been clicked (across both plain text and HTML emails) within your campaign.

This is incredibly useful for two reasons, firstly you will now know how many people on your list are actually opening your emails and secondly you will know what sort of links are appealing to people.

Tracking email openings

Unfortunately, you can only reliably track the opening of HTML emails. An HTML email by its very nature will at some point have to go and ask a server somewhere for some sort of data (most commonly, an image). Software can recognise these requests and use them to figure out that an email has been opened. This technique can then be cleverly adapted to put in a unique request for some data for each recipient opening the email. That way, you can see exactly who has opened the email.

You can also see when a user opened an email, which could potentially be very useful information for marketing purposes. It could be used to determine what day of the week or time of the day to send future campaigns.

There is a slight flaw with this method; if a recipient forwards their email to 20 friends, each time a friend opens it, it’ll be counted as an opening from the original recipient. A bit of common sense when analysing your statistics is required! I’m sure you wouldn’t be complaining that your campaign is being forwarded on anyway, it means more people are reading it!

Because plain text emails are simply text, they don’t make any requests to the server, so it’s a lot harder to track openings of emails. Links in plain text emails can still be tracked, so if you notice a recipient has opened a link but their opening of the email hasn’t been logged, it’s likely that they’ve opened the email as plain text.

It can be quite exciting to have a look at your statistics after sending out a campaign, however, don’t expect everyone on your list to open your email, it’s often only around 20% of your list of recipients that are recordered as opening the email (possibly less, if there are a lot of plain text recipients) but it’s still far more cost effective to send to 1000 recipients and have 200 people open it than to send out 200 paper mails!

You could use the open statistics to collect a new list of recipients, specifically the ones that didn’t open your campaign. You could then send them another one a couple of days later that starts with something like “We noticed you didn’t open your last email, maybe we can tempt you with these better offers?”.

Tracking link clicks

As well as whether the email has been opened, you can also track what links people have clicked on (and again, when they clicked them).

It’s a similar method to tracking the openings. Basically, the server sending out the email campaign acts as a stepping stone to the recipient’s route to the destination link. The email campaign’s link doesn’t go directly to the intended URL, instead it goes back to the email sending software. The email sending software recognises the unique request, logs a “click”, checks which URL it maps to, and then sends the user on to their destination… in an instant! It’s a bit like passing through a turnstile.

Finding out what your recipients are clicking on can be very useful. Lets say you’ve sent out a marketing email with three different offers in it, with a link for each to a page on your website with more details about that offer. A quick review of your link statistics could tell you which offer of the three was clicked on the most. Knowing this information can help you decide on offers for future campaigns.

Going slightly deeper, you could target individuals by using links to determine their interests, and then send them follow-up emails with details of products/services that are tailored to those interests.

All the links in your campaign are automatically tracked if you have tracking turned on, so you can also use link tracking to see if and when recipients have unsubscribed themselves from your email campaign (and if you send a lot, which campaign prompted them to unsubscribe could be useful information).

You can also see if and when users have clicked the “webversion” link.

The webversion link, when clicked, opens a copy of the recipient’s email in their browser. It is usually presented as “If you are having trouble reading this email, click here to open it in your browser” (because HTML email is extremely delicate and you need to know all of its quirks to get it to display correctly across all the incarnations of email client) and is very wise to include.

If you find that your webversion link is getting a lot of clicks, then it’s quite likely there’s a technical problem somewhere in your campaign, and something you’ll probably need to get fixed before the next one goes out!

It’s worth noting that you can make a webversion uniqe to each user, in which case all the tracking will still work even if a user is viewing their email in their browser.

A quick note about the look of the links themselves. They’re obviously dependent on the style of the rest of the email, but you probably do want them to be prominent. The standard is blue, underlined text (when using a white background), but there is a bit of leeway, it’s certainly not something to overlook!

An example…

Over the Christmas period we decided to send out an E-Christmas card to our colleagues. It was a bit of fun, but also contained a few tracked links. You might remember receiving it… if so, we hope the turkey still has pride of place on the mantlepiece!

The email in its original state can be found here. Note the code to invoke the first name into the text to give the email a bit of personalisation (you might find me talking about that in another blog entry!), there’s also the webversion link which won’t work as it’s automatically generated at the time of sending to make each user’s webversion unqiue.

After that email was sent out, we regularly checked the statistics, finding out who was opening the emails, and whether they were clicking any links. One particular recipient appeared to be opening it repeatedly, which we soon realised probably meant that they’d forwarded it around their office, which did bring a smile to our faces.

I’ve also just taken another look at the statistics, and it seems one recipient opened the email on the 9th February! That’s got to be a record for the earliest card for Christmas 2009!

Obviously, as I said, that “campaign” was just a bit of fun, but it did prove the power of tracking, and if you didn’t download the PDF, we know who you are!