Whether youâ€™re paying for clicks or just bandwidth, traffic to your website costs money. To get a better return on investment, you need to optimise your site. That means guiding more visitors to make business-critical actions. Whether those are phone calls, purchases or something else, these are the five most important steps.
- Use Google Analytics for your conversion goals
This is the first step to improving conversions. An analytics tool enables you to accurately track conversion rate. With Google Analytics, you can set up custom goals and it will calculate the conversion rate of each one.
Youâ€™ll need to decide on what type of conversion you want to optimise. If itâ€™s for purchases, you can set up ecommerce tracking. Along with conversion rate, it records other stats like revenue and product performance. Other websites are aiming for downloads or newsletter signups. Pick a goal that suits your aims and start tracking it.
- A/B test your pages
A/B or split testing is when visitors see different versions of the same webpage. By the end of the testing period, you keep the page that performed better. Ideally, youâ€™d run an A/B test like a controlled study. That means only having one variation, like a red vs. blue call-to-action button. Itâ€™s also a good idea to run the test for a long time and get a big sample size. For low traffic websites, this isnâ€™t feasible. This blog from Moz has some good tips for these cases.
For A/B tools, Google Content Experiments has been a popular choice. Itâ€™s built into Analytics, so you donâ€™t need to set up a new tool. However, Google has deprecated it in favour of . Both tools are still useable for now, but itâ€™s probably worth learning Optimize, since Content Experiments is on the way out.
- Get direct feedback
Thereâ€™s nothing likeÂ direct feedback to help understand the user journey. So while A/B testing reveals results, surveys reveal intent. You donâ€™t even have to chase existing leads by email or social media. Instead, you can embed surveys on the website itself.
You can also get direct feedback by watching visitors. Mouseflow replays user sessions, so you can see exactly what they did and where they struggled to convert. There are also some lighthearted options, like The User is Drunk. There, a user will report on your siteâ€™s design, copy, navigation, etc. while impaired. The idea is to test how idiot-proof your website is.
- Get trust signals
If you want to convert users, you have to remove doubt. According to BrightLocal, 84% of people trust online reviews as much as a personal recommendation. Not only that, 74% of consumers say positive reviews make them trust a local business more. Services like Feefo will email customers and request reviews on your behalf. When you have enough good reviews and a high overall score, you can embed them on your website.
There are also security signals to consider, like HTTPS. Google Chrome is getting more alarmist about websites not using this protocol. Users will be put off your website if their browser throws up warnings. Also look into other signals, like payment badges in the footer. PayPal, SagePay, Shopify and others have badges you can add to show users how their payments are processed and secured.
- Review your Unique Selling Points (USPs)
Push USPs like free delivery or price matching. Even better if you can make visual stuff, like a video of your quality control process, for example. When youâ€™re picking USPs, donâ€™t settle on generic stuff like â€œwe treat our customers like familyâ€. Any company can claim that.
Be as specific as possible. If you run an ecommerce site, people will want to know about delivery. So donâ€™t wait for them to email you. Here are some common examples of vague USPs along with better alternatives:
The thing to remember about these steps is that theyâ€™re ongoing. You should always test new ideas, improve your copy and source feedback. Technologies like Optimize and Mouseflow are relatively new, but the principles haven’t changed. The aim is to improve communication between the website and the user. That can be anything from a lengthy survey to changing the colour of a button.