The Conversion Chronicles, resources for improving your online conversion rates

The Old Farmerís Mascara

Finding Your Audience in the Age of Web Analytics

Marketers have always grouped their target audiences into different types of people based on various criteria. For instance, they might group people based on their age, gender, or geographic location. This is so that they can begin to figure out who the best type of person is to market their product to. A young woman living in Paris will obviously buy cosmetics more frequently than an 80 year old man living on a farm. The marketerís job is to figure out the differences between the people she is marketing to and position her product accordingly.

What Iíve begun to notice is that businesses are still not grouping their audiences effectively because they havenít changed their thinking to match the way current Web analytics technology can help them.

How Web Analytics Can Help

Most Web analytics tools can segment or filter a websiteís visitors based on certain criteria, at a general level Ė much the same as marketers do it now. For instance you can set up your Web analytics tool to view how visitors from a certain country or city act on your website as compared everyone else. You can (by using registration data) determine gender ratios, product/service interests, and actually record behavioral differences between different types of audience.

Returning to our cosmetic example, if our Parisian lady had logged into her favorite cosmetics website and given them all the data she had been asked for (name, address, age, interests, survey information etc), you as the website owner could aggregate the data across all your registered visitors and use your Web analytics tools to segment them into groups. You could then present offers to the right people. Our lady might be interested in a certain type of perfume along with thousands of other registered users to your website. It means you could send a tailored email to them because your Web analytics tools have told you that these particular groups of users buy a lot of this kind of perfume.

This is very valuable information to know but also the hardest type of information to get.
Registrations require that the visitor gives you the information at some point in time and itís often difficult to obtain. It works though. Itís why companies like Amazon are very successful because they know so much about every customer or registered user that goes to their website Ė and can present products which match the interests of each visitor

What If You Have No Registration Data?

Most websites donít have this kind of registration data, or if they do they donít know how to integrate their Web analytics tools with the user information so that they can start marketing more intelligently. Itís a very common problem issue and is often expensive to address. Asking businesses to spend big on something unproven is usually impossible. However there are some practical and easy to implement things you can do with your Web analytics systems which will give you much more insight into what works and what doesnít. By doing this you might be able to start to see some quick wins Ė which can give you much more credibility when you ask for investment.

1) Segment by campaign

If you have a banner or search engine marketing campaign filter the visitors arriving from the campaign source. By comparing how each campaign works and how the visitors behave on your website you can easily see where to spend your money. If for instance you find that the Ä10,000 a month youíre spending on banners has a 1% conversion rate with very low user interaction (low page views per visit & high bounce rates) as compared to the 3% conversion rate from search marketing, it makes sense to spend more on search marketing than banner ads. Itís a quick win because it saves you spending money on poor campaigns and allows you to spend bigger on more successful ones.

2) Segment by referral or affiliate

You should look to find which your best affiliates and referring websites are. Is there a website which is driving high quality traffic with high conversion rates? If so it may be worth offering incentives to get better exposure on that website. On the other hand if youíre getting a lot of traffic from an affiliate and the traffic is converting at a low level, then it might be worth pulling your resources from that affiliate and concentrating on something else.

3) Segment by location

Most tools these days can actually see where the visitor comes from by looking at the visitors IP address and referencing it to see which city the IP address comes from. Itís not 100% accurate (Web analytics tools very rarely are) but it will help you to see for instance if Paris is a better area to concentrate your efforts on than Nice.

4) Segment by search engine

I already mentioned campaigns, but search engines are very important. Knowing how people use them and more importantly how people convert from them means you know which search engine to spend your money on. Is Google better than Yahoo or does Yahoo convert better? This will allow you to gauge where to spend your search marketing money.

5) Segment by on site behavior

Many visits to your website may be Ďuselessí visits in that, for instance, they arrive at one page, find that they have arrived at the wrong place and go back to a search engine. It is worth filtering these out in many cases to see if your useful visits are finding what they need. It might also be useful to segment visits by the areas of your website that they visit. Are the visits to your shopping cart performing better than the general visitors to your website? Do you need to do something with your cart to improve it?

6) Segment by system variables

This might sound like a weird thing to do, but weíve found huge differences in onsite behavior in someone who has a low screen resolution in comparison to someone who has a high resolution. Similarly look at operating systems and connection types. Your audience may have low technology and you have a high tech website which they find very difficult to use. Are you catering to for them well enough?


Of all the uses of analytics tools segmentation is one of the least used and most insightful in terms of what you learn. In most business websites you have a large number of visitors coming from a wide variety of different sources all of whom have different agendas and goals. If you look at them all as one bunch of interested visitors itís like comparing perfume with lipstick. Both are cosmetics, but one makes you smell nice and one makes you look good. How then do you know to sell the perfume Ė and not the lipstick Ė to the Parisian lady? If, however, you have at least figured out that the affiliate website from Paris is driving a lot of good quality traffic, and these visitors have a tendency to buy perfume, there is a good chance that you have found her.

More from this months issue | Archived chronicles | More from this author
Steve JacksonAuthor: Steve Jackson, Editor - Conversion Chronicles

Steve Jackson is the Editor of the Conversion Chronicles, a website conversion rate marketing newsletter dedicated to improving website conversion rates. In 2003 he co-founded Aboavista the first web analytics consultancy in Finland and now a wholly owned subsidiary of Satama Interactive. Satama Analytics unit is a web conversion and web analytics consultancy based in Finland with offices across northern europe. You can get a free copy of his e-book sent to you upon subscription to the Conversion Chronicles web site.