Site optimisation

My team have been very busy developing new updates. Some have been fixes, but most are improvements and new features that we hope will encourage more activity from users and ultimately make us more money.

However we have no way of guaranteeing that these “improvements” will work, so the only logical thing to do is for us to test EVERYTHING. Or at least test what we can.

If you own a site then you should have a methodology in place for optimising it. I don’t mean SEO (which is important), but improving the website to encourage whatever activity you want from your visitors.

This is a break down of the kinds of testing we do and the tools available to optimise your site.

AREAS TO OPTIMISE

1. Page design
There are certain pages that you should be constantly optimising. Registration and payment pages are key examples which are easy to monitor as you only need to measure conversion rates. Changes like a simple text change, a new button or a completely new layout can all have an effect. We do this as an ongoing tasks constantly trialling small and big design changes.

2. Core logic
In the case of Besinful these are things like: What features our members pay for, how many messages they can send, is it free to read or send messages, how much they pay for 1 months membership etc.
Tracking and reporting on changes to core logic is harder than simple design tweaks, however it’s still massively important. We recently made 1 small change which resulted in members becoming 3x more likely to upgrade! It was obviously worth while us making this discovery!

3. New features
If you have a site which is constantly evolving like we do, you’ll have a roadmap of big development milestones. While on the face of this it’s easy to assume that more and better features is a good thing it is definitely not always the case. Site users are often attached to what they currently have, big changes might aggravate them if they don’t personally like them, or scare them away if it’s not easy to use. Facebook is a prime example of this, every time they roll out something new a small % of members go crazy! Although in Facebooks case they clearly test everything very well and based on their success it’s impossible to argue they’ve done much wrong.
We recently introduced a new user dashboard on Besinful. While we had immediate positive feedback via member emails, the real sign for us that it was a good feature was an increase in the number of pages visited by members, longer time on the site and an increase in uploads – all of which we were aiming to achieve.

TOOLS FOR TESTING

1. Google analytics
It’s free, constantly being improved and works directly with google adwords so you’d be mad no to use it! Though it’s not exact, you’ll be able to see where your members are coming from, what pages they’re visiting, how long they’re staying for, set up goals and funnels, and segment the data in many different ways, plus you’ll see all the results in some lovely easy to read graphs.

2. Google Website Optimiser
This is an extension to google analytics which is also completely free that makes it very easy for you to split test different versions of the same page with a few lines of code, and find out which version has the highest chance of success, and with what probability it has of out performing the others.

3. 3rd party solutions
There are restrictions to using google analytics. Primarily the biggest problem is that you have no link to your own database. As you’d expect there are numerous solutions out there that promise to provide you with much better insights and results than you’d see using just google. Some might offer ways of integrating with your database to give better segmented targeting or testing, even more specialised reports, or they might have a managed solution whereby they provide suggestions for improvements, run the tests and report the results to you.

4. In-house solution
Because of the restrictions of google analytics and the cost of 3rd party software we opted to build in tracking and reporting to our own system. We know what the conversion rates are for members that come in from different sources using google analytics, but by storing campaign data in our own database we know the the spending habits of members from different sources which is much more useful. No point in paying for traffic that appears to convert well, but in reality they’re only spending 1 penny over their lifetime!
We also filter by user demographics: country, region, age, gender… by usage demographics: registration dates, payments dates, last login dates…. by what features members have access to (when we roll out a new feature we don’t make it available to everyone allowing us to compare how the feature is used)… and a lot more besides!

CONCLUSION

Make use of everything you can get your hands on, especially free tools. Don’t assume that just improving your site will give you better results.
Have someone responsible for continually testing your pages – it doesn’t matter if it’s yourself, an employee, freelancer or another company.
If you’re developing a site yourself rather than using someone else’s software, build in to the specification some extra reporting tools. This doesn’t have to be any more than recording a little bit more about your members and having the ability to manually run a query on the database. You just need to decide what it is you want to know about your customers to determine what data to store about them.
Don’t delay – do this from day 1. If your site is more optimised from day 1 your bottom line will be better from day 1.

If you read this and think missed anything off let me know. We’ve applied a simple, logical and mathematical approach to optimisation, but there are other tools we could be using and other things we could be doing which I’d always be willing to explore.

Happy optimising!