By Paul Morris
In a previous job as MD of an innovation division (we launched everything from property portals to give away Search Engines) part of my role was heading up a team that sent 5+ million emails per month to a database of circa 500,000.
I learnt a lot in that role about perceived spam and about turning around declining engagement and deliverability rates via contextual and behavioural ECRM. As I have not blogged about ECRM before I thought it about time to dig into the recesses of my tiny mind and give you some pearls of wisdom.
Please note that I am not preaching; I got a lot wrong in my time emailing the database however lessons learned will hopefully prove useful for others. Most of the below are headline points hence if you would like to discuss anything in more detail please get in touch via the contact form on the right hand side.
Get the basics right. Unfortunately the basics are not always basic however these hygiene factors are key to successful email marketing. Hygiene factors include getting permission to send solicited emails, make sure you optimise your sender reputation so you are not blocked by ISP’s, engage with ISP white listing, clean your database regularly optimising soft and hard bounces, etc. Of key importance is partnering with a great email marketing provider as they can help you with all the above and stop you simply going in to the spam box.
Add Value to the user. Your email value proposition is key. Is the recipient getting true value from your email? If not then don’t bother sending them an email! Equally the process of data reciprocity is key. I.e. if you are asking for additional preference information to tailor emails there needs to be a real benefit to the user; particularly nowadays when people are scared of surrendering personal data. In some cases it is also a good idea to give the recipient the control to tailor their own email preferences via an online login.
Linked with the above; when you have obtained the customers preferences or ‘just’ figured their mindset out via past engagement behaviour (you are what you click) or contextual/ personal data then make sure you do something with the information. A great example is Amazon. When I click on a few watches on Amazon.co.uk I funnily enough receive an email later that shows me different types of watches, special offers, etc. A little complex to enact (you need pre set templates and copy, a CRM system that feeds into your customer click analysis database and then an email system that sucks up the data and spits out an email to pre determined segments) yet worth the hassle. Essentially you really just need to think about your user base and how you can apply contextual and behavioural targeting methodologies in a cost effective manner and then segment accordingly.
Do not spam! It is far too easy to send great generic emails with great generic offers in to your database on a regular basis. “Wow we just doubled the number of emails sent out and pretty much doubled our income – let’s keep doing that!” Unfortunately this attitude will impact your list as people become bored with the untailored and start to disengage, unsubscribe or mark you as spam and that’s when the problems start! It’s a big mountain to climb to get back from that and I would simply advise you not to go there. Treat your email database like you would want to be treated (personalised, highly relevant, great content/ offers, timely, etc) and you will not go far wrong. Oh and DO NOT rent out your list. You will make more money in the short term but make less money in the long term.
Test. You should test everything that is important to helping you exceed your goals. You should instigate a testing matrix so you have a rough plan of what to test and when. You also need to store the results somewhere and make use of the data. Things to test include the format of the design templates, call to action messages, subject lines, what appears above the fold, offers, content, etc. As an example if I selected a 200,000 ‘segment’ from the 500,000 list we used to mail then I might send 5 different version of it out to 5,000 people per version. I would then wait 1 day for the results to come back and then use the winning version (make sure you have your KPI’s nailed for this as the winner is not always as obvious as you might think) for the rest of the segment. Testing can even go down to some quite granular detail. E.g. Geeky male tech folk generally engage with emails from a hot female name more than from a boring bloke’s name (seriously!)
Publish your newsletters online. Not only do people sometimes prefer/ have to (depending on their user client e.g. Blackberry users) read an online version, if you optimise them well they can also help towards your SEO activity.