I like to say that Conversion Rate Optimization is the “gift that keeps on giving”. By which I simply mean that an increase in conversions now will still be in effect making extra profits for your company for years to come.
Sometimes I find that people dipping their toe into the conversion optimisation waters give up after a couple of small experiments because they aren’t getting any big increases in sales.
Firstly, to find big increases you have to test big differences. The shade of yellow on your “buy now” button isn’t going to jump your conversion rate from one percent to ten. If you want dramatic changes then you need to test dramatic things. And you need to be prepared for dramatic falls in conversion rate too.
That said, small changes can still lead to meaningful differences to your profits.
Consider this chart that shows a site with a steady 10,000 visitors per month. We start converting at 1%, getting 100 sales. If the sales are for an average of £20 then the control month (0) has a turnover of £2,000.
By making small changes each month that push the conversion rate up by 1% each month by the end of two years the site is now making an additional 24 sales, worth an extra £480 per month. Almost a quarter more than when started.
Surely out of the of the hundreds of things on a page – words, images, layout etc – you can improve something by at least 1%?
With our clients we ask that they invest 20% of their improved sales back into Pay Per Click marketing. If we had an average cost of 25p per click, lets see what that does to our figures:
Instead of 24 X £2,000 = £48,000 we’ve earned £54,000 an extra £6,000 or three months sales.
Now, we’ve made an extra £2,580 in the final month and instead of £48,000 we’ve made £54,940, £6,940 more.
Another aspect of pushing more visitors to your site is that you are able to get test results quicker. Lets say we are able to make improvements that net a 2% raise in the conversion rate each month.
We are still talking moving from 1% to 1.02 percent. Nothing incredibly dramatic and completely feasible for any small business trading on the internet.
In the final month, we’re making £1,540.00 more than in the control month! We’ve also improved the total sales from £48,000 by £16,400 to an impressive £64,400. 77 more sales than the control month.
Just for fun (and inspiration), lets have a look at increasing the sales by 3% each month:
Now we’ve raised the turnover for the 24 months to £77,700. An increase of £29,700 or more than a years takings on the original conversion rate.
So how possible is a monthly increase of 3%? In our experience very easily possible. When you consider here that the end conversion rate is still only 2.03%
In recent testing we’ve raised conversion rates from under two percent to more than five over a couple of months for one client. They now have the bug and are testing everything in sight. In a few years I fully expect them to be converting near ten percent and take a site that was making a few thousand a month to turning over more than they used to in a year in a single month.
Another aspect we haven’t yet touched upon is the number of clients your business now has. It is always easier to sell to people you’ve previously sold to. By increasing your pool of past customers you’ve a lot more people you can market to again with special offers and so on especially for past customers.
With 100 sales per month, we’ve a potential pot of 2,400 past customers at the end of two years. In our last example, we’d have made 3,885 sales over two years. An increase of 1,485 sales. Giving you a much larger pot of past customers to market to again.
The obvious question now is what to test and how to do it.
The how for most websites is fairly easy. For the majority of websites then Google Website Optimizer (GWO) provides everything you need to start testing.
GWO lets you test both as split tests and as a Multivariate (MV). Multivariate just means you’re testing several things on a page or pages at once. We pretty much always do MV tests.
A split test is useful if you want to test one page against a dramatically different version. A MV test lets you test different sections of the page against one another and comes up with the best elements to use.
What to test is a harder question to answer.
If it was easy everyone would be doing it. Thankfully for you, most people aren’t. By testing your site you’ll be ahead of 99% of internet marketeers and e-commerce sites that are currently out there.
Most sites get more business by pumping more and more traffic into themselves by buying contextual adverts from the search engines, and perhaps improving their performance in the search engine rankings with Search Engine Optimisation. They might “test” different product ranges, prices even advertising but for most businesses the site’s themselves are left well alone.
Whilst I encourage all businesses to look at Pay Per Click (PPC) advertising to try different keywords and to get more visitors, by improving the conversion rate of your website you’re going to be paying less per lead and make more money long term.
When someone wants to start testing we always make sure that they are prepared to invest in PPC advertising and have strategies to improve their site’s SEO. When you have three elements in place:
- Pay Per Click campaigns
- Search Engine Optimisation
- Conversion Rate Optimisation
Then the site is set up to snowball. In that I mean that the site will earn more per visitor as more of your visitors convert, so you’ll be able to afford to spend more on acquiring customers (advertising) and the advertising itself is more effective. By making more sales and having a wider customer base you’ll be able to improve SEO faster (more links from satisfied customers). It’s a positive feedback mechanism, with every factor helping improve the business more than the individual element. It’s win, win win.
But we still haven’t covered what to test.
Take a few moments watching someone new to your site and watch them stumble about finding products, trying to work out how much they cost to ship and wading through the endless checkout & login pages, and you’ll soon come up with multiple ways in which the process on your site can be improved. Write up your findings and have your developer create the alternative version. Where possible test the two versions in parallel so you can see if visitors really do buy more from the revised edition of your site.
From there on there are the easy and obvious things to test – fonts, sizes and colours.
Being easy and obvious means there’s sadly little value in testing them other than to improve what would otherwise be really bad choices. For instance, if your site isn’t already black on white, then make sure to test test black text on white background. Any thing else is going to be hard to read for a number of people and will hurt your conversion rate.
If you’re using a “clever” font, then test something simple like Arial or Verdana. If you want serif then try Times or Georgia. Yes your designers may cry over their private font collections. Let the conversions do the talking. Sometimes the designers are right. The customer always is.
Here’s a brief list of some elements to test:
- Placement of the guaranty
- Placement of return policy
- Third party security badges and seals
- Call to action Colour
- Wording (buy now, add to basket or cart, trolley, bag…)
- Size
- graphic or plain text
- Arrows pointing to button
- Other iconography such as a basket or shopping cart
- Test having the stock availability near the buy now button,
- Customer service links
- Dispatch information
- Delivery costs
- Shipping options
- A telephone number
- Customer reviews
- Product comparisons
- Test the product imagery – Size
- The on-page positioning
- With or without a close up
- With a person holding it or using it
- Does it click through to a bigger version
- Does the bigger version “pop up” or display inline
But the real beauty (and increases) usually comes from the sales copy.
Take a look at this site. It has nothing to do with me I just happened across it one day. The product copy there is crafted. You’re taken on a little journey as you read the product descriptions and are giving a glimpse of what it’s like to have those items already. The copy makes you want them and to have that life.
Some pieces will work for your site, others wont. But take a moment to look at each and try and work out why. Use that knowledge when testing your next idea.
Testing isn’t always easy, but it is very rewarding. Just look at that graph showing sales of £2,000 creeping ever up and up and up.
Liam Delahunty publishes a free conversion optimisation newsletter and occasionally writes about at conversion and search optimisation at Online Sales.
Related posts:
