Email & Website Optimisation

Improving conversion for over a decade

United Kingdom United States Germany

With online data it’s all about engagment


By: Mark Patron | June 24th, 2009

Database Marketing – June 2008

I believe that measuring and segmenting customers by their level of engagement with the brand will become essential for online marketers over the next five years. Measuring and segmenting online prospect or customer engagement will define an online marketer’s success. Let me explain why.

Direct marketers learnt long ago that past responsiveness is the best predictor of future response Why? Well we all are creatures of habit, past behaviour is the best predictor of future behaviour. If someone reponds they are more likely to respond in the future. If someone visits a particular content group it is likely that that will be the content they will go to on their next visit.

For direct mail there are typically only two possible outcomes, either response or no response. For email there are many possible outcomes or levels of engagement; not open, open, click through or purchase. Similarly comparing the website with the call centre, while there are more potential levels of engagement outcomes in the call centre than for direct mail I would suggest there are more typically measured and tracked on the website. For example number of pages, length of visits, interactions and how engaged they are with the brand. Obviously during a call an operator will be measuring their level of engagement with the consumer to a highly accurate level, but sadly this data is rarely available for use in the future.

Rich online behavioural data can be more easily used if website visitor activity is tracked with a good quality web analytics tool. Then different levels of engagement can be targeted appropriately. So if someone abandons a shopping basket when shipping charges are added a follow up email can be sent offering free delivery. This may seems like a no brainer but research we did recently showed that very few e-commerce sites do this. While the data exists joining it all up is not easy. A recent study by MarketingSherpa showed that 68% of attendees of the ad:tech conference in the US planned to invest in integrating their analytics, email and search compared with 50% the year before. Much online technology may be very sophisticated on a stand alone basis but it’s still not very joined up in practice.

Engagement segmentation is a good way of identifying the degree an individual is likely to do something in the future. For example, repeat visitors to a website are eight times more likely to buy than first time visitors. So use email marketing to get people back to your website. We have found that segmenting emails by level of engagement improves results. Testing different offers and creative for people who never open their emails, do open or click through works well.

I find it scary that online customers tell us when they are about to lapse but we do nothing about it. If consumers interact and visit your website fairly regularly, say on average a couple of times a week or month, then if the number visits halves it probably means they are spending the other half of the time on competitors sites. If we care to look customers are telling us when they are about to lapse. I think it’s downright rude not to use this data! Yet hardly any websites do.

To summarise, offline marketing is yes/no response push marketing. Online is incremental engagement pull marketing. In other words engagement is the online equivalent of offline response. Offline it’s black or white, online it’s shades of grey. There are many more rungs on the online loyalty ladder. This means you can convert customers online more easily if you measure, understand and segment those rungs but it’s very early days.    


Post a comment

Your email address will not be published. Required fields are marked *

*


*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>