Want a Superpower? Beat Your Competitor’s Conversion Rate : Seedcamp

This guest post is written by Will Critchlow, Founder and CEO of Distilled; an award-winning online marketing agency. Will has consulted on digital strategy for some of the world’s most influential companies and believes tomorrow’s biggest brands will be built online. Will recently came in to lead a masterclass focusing on results-driven digital marketing. This article follows on from that session, focusing on how optimising a website’s conversion rate can be like a superpower.

Setting up an effective website is tough. It takes time to figure out what works, and you quickly lose an outside perspective on how your customers experience each page.

However, it’s worth the battle. Having a higher conversion rate than your competitors isn’t just useful, it’s pretty much a superpower.

Achieving a higher conversion rate is particularly important to startups as it not only brings the obvious benefits of more customers or sign-ups for any given level of website traffic, but less obviously it also:

Proves that the market wants what you have to offer – the elusive “product-market fit”
Helps every bit of marketing spend (and hence investment) go further
Unlocks new marketing channels that aren’t cost-effective to your competitors
Helps with raising money by improving the metrics that investors care about

Sounds pretty good, right? Welcome to the world of conversion rate optimisation (CRO).

CRO is, essentially, about getting more people doing what you want on your website. Ask yourself: for every 100 people who visit, how many people do you convert? Now, how many people could you convert? (The ‘conversion’ may be a sale, someone downloading a guide, or whatever it is that defines success in your business.)

So let’s delve into the ‘how’. A senior member of the Distilled team, Paddy Moogan, outlines a five-step process:

Step one: Gather the relevant data

Qualaroo

There are certain key areas to look at here, which can roughly be broken down as follows:

What are the goals of your company? Why were you set up in the first place?
What objections are often raised about your product? Here you’ll want to speak to people across the company. Find out what customers typically complain about or misunderstand.
In terms of the website, what is the sign-up or sales process like? What are the particular touch-points between the customer and your website, plus any other marketing materials such as email? Look for places where customers ‘drop-off’.
What’s the traffic breakdown? (If you don’t already track web activity in Google Analytics, now is the time to start doing so.)
What technology do customers typically use? Look at how your website appears in all relevant browsers. Also, think about devices. Do many people prefer mobile to desktop? Are they often using more than one device across a purchase journey?
How well do you know your customers? What do they look for in a product or service? Surveys such as Qualaroo can be handy here.

Step two: Outline your hypotheses

After gathering the above information, it’s time to work out what to test. You need to come up with specific hypotheses, based on your research. For example, if your research indicated that customers like your money-back guarantee, you could make this selling point more prominent on your product page to see if a higher number of people start the checkout process.

You also need to pin down who you’re testing – what types of customers – and which web pages.

Step three: Wireframing

Balsamiq

If you need sign-off on your tests, it’s time to start showing exactly what they’ll look like. Good wireframing tools include Balsamiq and Mockingbird.

Step four: Implement the design

At Distilled we use Optimizely to run split tests, redirecting a proportion of traffic to the test pages. Whatever you use, remember to test your new designs across various browsers as you don’t want to distort your tests by having broken pages on some platforms.

Step five: Review

Were your hypotheses correct? Before you can answer this question, you need to make sure you’ve reached statistical significance, ie. you have enough information available to form a reliable conclusion. To get more in depth on this, check out this article.

If your hypotheses were wrong, it’s a case of returning to your research stage to dig a little deeper into what’s preventing or discouraging people from converting. None of your previous tests have been a waste – it’s all about building up an understanding of the process and this inevitably involves some leg work.

Soon enough, however, you’ll start figuring out the weaknesses of your website. It’s then immensely satisfying to remove the bumps in the process and see the conversions speed up.