Why now is the time to optimise your site for christmas traffic
It’s that time of the year. Summer is over and the days have started drawing near. You can feel an autumn chill in the air (Ok if you’re from the other hemisphere, then you got summer coming..lucky you). But more importantly, it’s the time when retailers and small businesses are focussing on the next big event in the calendar.
Yes. Christmas is coming and bringing along with it, shoppers in the droves.
Christmas is a crucial time of the year for many businesses not least online ecommerce stores.
In the US, online sales are expected to rise from $306.85 bn [£189.26] in 2014 to $349.20 bn [£215.39 bn] in 2015 and $398.78 bn [£245.96 bn] a year later. Canada’s online sector is comparatively small, but is forecast to grow from US$20.82 bn [£12.84 bn] in 2014, to reach $23.56 bn [£14.53 bn] in 2015 and $26.99 bn [£16.65 bn] in 2016. These stats come from a study commissioned by RetailMeNot and the Centre for Retail Research.
It’s easy to see that online sales are rising year on year and mobile is playing a very important role in this. There is still a hesitancy for mobile buyers to complete purchases online but the number of purchases made on mobile has certainly increased.
The time to act is now. The worst thing you can do right now to improve your website for getting more sales is nothing. A common mistake is to wait until December to optimise a website. We have seen businesses wait till the last moment to optimise quite often and it will harm your prospects of getting good sales from your website
1. You don’t want to be doing knee-jerk changes
Gut feeling is the worst indicator of what is wrong on a website. Making changes solely on what you feel is wrong is a surefire recipe for disaster. Mainly because you cannot track whether the changes you have made are contributing to an uplift or downturn in sales and conversions. If you see an uplift in sales, it could be easy to attribute it to the website changes whereas it could’ve been something seasonal that cause lots of visitors to buy from your website. There is no way of comparing the new version to the old in terms of performance.
Enter A/B testing. You must’ve heard of it. The magic pill that cures all the woes of a site. Run an experiment where you change the colour of a button and you’re laughing all the way to the bank with results.
Except that’s not the way it works
A/B testing is a rigorous process of testing your idea in real time against your existing site. This way, with all other factors being consistent (ie. the time of the day, seasonality etc), you can say whether the change you were testing had a positive impact or not. It’s the scientific method.
2. But…Real optimisation takes time.
You can’t run a test for 2 or 3 days and look at the results and make a decision on whether the test was a success or not. It will give you false results that can be as dangerous as not running any experiments at all. Why? Because it takes a while for a test to reach statistical significance. You must run a test for two business cycles at the least. (2 weeks so you can factor in the weekday and weekend trends in visitor behaviour).
The other key point to explain is optimisation is not the same as running A/B tests which is a common misconception. It is a systematic process which involves research and analysis of your data to come up with ideas that you can then test out.
Testing is risky. Even the best ideas fail in the tests or the change is not conclusive. Worse case scenario, a test that’s failing could reduce your revenue. That is why you would mitigate it by sending only a small portion of your traffic into the tests. If the test fails, you can do a post-mortem analysis and run another iteration of the idea or scrapping that idea entirely to move in a different direction.
You need focus and discipline to be able to run an optimisation programme in a consistent and structured manner.
Now, what happens if you leave this to the last minute?
You won’t have enough time to execute any of the optimisation tests on your website.
Let’s just say you start on the 1st of December. You manage to set one or two tests live. You’ve already missed out on the majority of potential buyers who had been visiting your site before that. You are also at the peak of your sales season so you’re adding a big element of risk into whatever tests you do manage to run.
When should you start optimising?
This post was published mid-September. So, Now!!
What are some of the things you can start with?
For starters, find and isolate the problem areas of your site. Where exactly in the funnel are users dropping out? Could it be that the checkout form has hidden issues that you’re unaware of. Make sure that you have the ability to view all the interactions on your site. This can be done through your analytics or your heatmap tools etc.
Sit down with your team and discuss those findings. This should be a brainstorming session where you are able to generate ideas based on the data you have collected.
Out of all the ideas you have picked out, which ones can you test quickly or with the least effort.
Start with testing those.