One of the interesting parts of working in custom web design is creating specific calls to action that actually produce the desired results. We definitely haven’t nailed down the process yet, but we believe it is our obligation as a professional web design firm to give our clients the best possible return on their investment. We identify key site goals with our clients through our Discover and Define process, create unique calls to action to funnel visitors, while maintaining usability and putting the visitor experience first.

A recent example is our redesign of the Julie Lawrence Yoga Center site. By pairing images of the yoga classes to develop visitor interest, with large orange buttons that guide them through the site, we hope to maintain the design aesthetic our clients are looking for, provide the portals that the site visitors are looking for and turning that visitor desire into action.

Here are some typical web reports from typical web analytics tools:

I struggled with reports like these for years. What does it mean if unique visitors (the number of people that visited your site, regardless of how many times they came back) are up this month? Is that good? What if unique visitors are up because you had a link to your site posted on a random MySpace website? Average time on the site is up. Is that good? Do we want people to stay longer? Or does that just mean that a bunch of people fell asleep at their desk while looking at your site?

Well, the answer is that we don’t really care about reports like these. They give us aggregate data for the site as a whole, which makes it very hard to draw useful conclusions about the many different groups that come to our website, and they use behavioral data that doesn’t get us very close to how we actually make money from our websites.

Below is a report from Google Analytics, a free analytics tool that we recommend to all of our clients. While this is a great tool that allows you to get beyond the basic behavioral reports that most tools focus on (visits, time on site, page views, etc.), it can be hard to find the useful reports in a tool that lets you view your data in thousands of different ways. Take a look at this and compare it to the reports above:

What’s the first thing that you notice? I’m guessing that you see that we are comparing different things on the site (pages, entry pages, search key phrases, etc.) to each other. Every one of these pages is being compared to the site average, and the pages that are performing worse than the site average need to be fixed. My two favorite metrics for these two are bounce rates and conversion rates. Pick the pages that are not performing, brainstorm about why they are struggling, and fix them: these are the proverbial low-hanging fruit.

If you are curious about how to generate this report, go to traffic sources->keywords (search keywords) or content->top content (specific pages), and select the icon on the far right above the data:

Now select Bounce Rate or Conversion Rate from the second drop down and go pick your low-lying fruit!

There are two big problems with conversion rates, however.

Problem #1 with Conversion Rates
Since a conversion rate is seldom higher than 3%, you can spend all of your time obsessing about how to improve that number while ignoring the other 97% of your traffic. There are usually a lot of steps that visitors go through before they will convert, and we have all been to the sites that won’t let us view anything before we give them our name, email, social security number, and the keys to our home…and we hated it. My general reaction to these kinds of sites is to grudgingly give them my info, get whatever it is a wanted, curse the company under my breath, and never return to their website.

So what’s the solution? Don’t forget that not every visitor will convert every time, and don’t try to make them do something they don’t want to do: they will never come back, just like the visitor to the used car lot who can’t afford to buy right now but is relentlessly to buy right now. And, pay attention to other metrics such as your bounce rate that are leading indicators of a poor customer experience. If you can, use exit surveys and other qualitative data to ensure that your visitors are having a good experience on your site and that they are left wanting to return in the future.

Problem #2 with Conversion Rates
Not everything can be easy to measure. Let’s say that I want to measure the number of leads that Synotac generates from our website. What if someone sees the site, loves it, but when they want to call us look our number up in the Yellow Pages online? How do we track that? Or, let’s say you want to measure the number of visitors who use your resources section and successfully resolve their problem without calling or emailing your company. Is it good or bad if people spend a lot of time on a page? What if someone looks at a lot of pages?

The solution to problem #2 is, not surprisingly, to get very clever about how you can measure these things. In the digital world that we all live in now, there very little from the web that we cannot track (one huge advantage we have over print, tv, radio, and all other forms of media). Here are my top four tips for getting the data you need to measure your website’s conversion rate.

Tips for better measuring your conversion rate
1. Get a different 800 number for your website. This can be done for $10-20 a month with a good service, and now you can track all of the calls that come from your website.
2. Rebuild any email links on your website to be email forms that can be tracked.
3. Add tracking code to your e-commerce success page, contact form, newsletter sign-up form, case study download link or whatever else you might be trying to track.
4. If you are trying to measure customer success in resolving a problem without using email or phone support, add a page-specific survey to ask if the page was useful in fixing the problem. If the answer is no, ask them for details as to why.

A while back we wrote an email newsletter that talked about bounce rates. Due to the clamoring for more information, I have decided to expand this into a post.

First, what is a bounce rate? Depending on your analytics tool, bounce rates are defined slightly differently:

1. Any visitor who stays on your site for less than a certain amount of time (usually 10 seconds).
2. Any visitor who only looks at only one page on your site before leaving

While the definitions are slightly different, the end meaning is exactly the same: you got absolutely nothing out of those visitors. Zilch. They don’t even remember your company name or what your logo looked like.

So why does this matter? Well, bounce rates represent an incredible opportunity: it’s quite hard to get more traffic, and it’s generally pretty easy to get more out of the traffic that you already have. Bounce rates are also one of (if not the) best metrics for actually giving us insight into what our visitors think about our website. Page views, visits, unique visits, and time on site don’t tell us anything about WHY the users actually do anything. A high bounce rate indicates that your visitors are allergic to that page. You can take a look at it, come up with a hypothesis as to why it might be having troubles, and make a change (or better yet run a controlled experiment so that half of your visitors see the old version and the other half see the new version). Wait a month and see if things get better or worse, and now you have feedback on your hypothesis. If you were correct, you now are getting more results out of your website without having to pay for more traffic AND you know something about your visitors. This kind of data is hard to get, and invaluable to helping your visitors to have a great experience.

What do you all think about bounce rates? Let me know any interesting experiments that you have run using them or feedback you have about how to best use it.