CVHBBA Trade Show - 2007

Today I am going to talk about the most important part of any marketing campaign - measuring the results.

When you are making the decision to market online one of the first thing you need to consider is, what do you want your website to do? There are two broad categories for business websites - online sales and support of an existing business. Most business sites are a mix of these general categories ranging from 100% sales, such as Amazon.com, to 100% support, such as for a local restaurant.

To see how it works, let's start with an imaginary business, 'Mike's Widgets'. They sell one product ? 'The world's best widget'. And, they want a website to support their business which is located on 7th Ave in Courtenay.

The first step is to assign some realistic, measurable goals for the site. There are a number of ways to set these goals and each business will be different. In our case, Mike's sales staff is spending an average of 1 hour a day answering phone queries about widgets and another half hour a day on the phone answering customers question about how to use the product. Mike found in most cases they are answering the same general questions.

His goal is to reduce the time staff spends answering these questions by half. He decides to put the product information and a FAQ online.

Now we have some measurable goals. Also, Mike has some idea on the value of the website, so he can set a reasonable budget for the site.

Mike's web developer has included Google Web Analytics as part of the web page. Google Web Analytics is a free analytical toll from Google. It monitors all the traffic on the web site. And that is what we will mainly be looking at measure site performance.

So let's look at Google Analytics. Link to Google analytics - slide 1

I am not going to go into how to setup the account. Google has made the procedure very simple. Your web designer has probably included Google Analytics or some other analytical tool as part of your site. The information we will be discussing is fairly generic, and, except for the display, can be found in most analytical tools.

Since Mike's Widgets is a fictitious company, I am going to be using the stats for the Comox Valley Home Based Business Association to demonstrate Google Analytics.

This first slide is what you will see when you log into your account. At the top is a chart of the daily visits for the previous month. What you can't see on this slide, is as I move my mouse along the chart it will give the total visits for the individual day.

Next we have a line that says ?284 people visited this site?. And, I am going to come back to that and look instead at the next number - the '677 Visits'. To understand where this, and other numbers, come from, you need to understand how traffic monitoring works.

Greatly simplified. when you visit the website that is being monitored, the monitoring program gives your browser a cookie. Which is a ?teckie? name for a packet of information that is stored by your browse. The cookie records, among other things, the last time you visited the site. Each time you view a page on the website, the monitoring program checks the information stored in the cookie. In this case, if you haven't requested a page from the site in the last 30 min. it adds one 'Visit' to the total.

This is why the information from an analytical tool is much more useful than the raw data, it knows something about what happened in the past.

The next piece of information is '284 Absolute Unique Visitors'. Which is also the number of people that visited the site. What that means is during last month 284 people visited the site at least once.

Why did I say 'at least once'? Because we know that the total number of visits was 677, so some people came back more than once.

The next four numbers are where things start to get interesting. 'Pageviews' and 'Average Pageviews' are related. Here we are starting to get the information you need to know to evaluate your site. Keeping in mind that this is a summary report, and we will be looking at more detailed information later, visitors to the site requested 3,425 pages or and average of 5.06 pages per visit. This means visitors were actually getting information from the site, not just staying on the first page. In fact they were spending an average of 6 minutes on the site.

Which leads us to the next number 'Bounce Rate'. This is the percentage of visitors that left the site right away. Why did they leave? Don't know yet, but just under one third of the visitors to the site left without really reading the page. We will look at possible reasons in a minute.

Finally, 'New Visits'. We already said that of the 677 visits to the site, 284 new visitors in the last month. But how many had never visited the site before? 36.93% or 250 visitors had never visited this site before.

So from this summary we can get some idea of how our site is doing. We know in the last month 250 new people visited the site, that many of them visited more than once and that they were looking at more than the first page.

But, wouldn't it be nice if I could compare that to how the site performed last month?

And ? you know I am going to show you that next. Google lets you compare ranges of data ? in this case we are looking at the last two months. Link to Google analytics - slide 2

I am just going to quickly go over these results. 'Visits' and ' Absolute Unique Visitors' are both up but 'Pageviews' , 'Average Pageviews', and 'Time On Site' are all down.

You'll notice Google colour codes the results so while 'Bounce Rate' is down, it is shown in green, because we want a low 'Bounce Rate'.

While we now have a general idea of overall site performance, we still don't have the kind of detailed information we need to analyze problems with the site.

I am going to show you some more detailed information, then take a break from Google Analytics.

Going back to the first page for a second. Clicking on Content > View Full Report we come to this page. Link to Google analytics - slide 3

Here we have a detailed breakdown of the what pages site visitors are viewing. At the top is the './' this is the same as the index page and it will usually be viewed the most. We want the next most viewed pages to either be our 'goal' or pages leading to our goal. For the CVHBBA site, our goal is to get visitors to view information about member's businesses. This information is on two pages that either display by category or alphabetically. And, as you can see they are two of the three most visited pages.

We can look at some of the other figures. Time on page is much lower for the alphabetical listing, indicating the visitor know the company rather than looking for a type of company. Bounce rates are roughly the same. One new figure '% Exit', the percentage of visitors that left the site after viewing this page.

Going back to the 'Bounce Rate' of just over 31%. These are people that came to the site and left right away. The usual reason people leave a site is because it is not what they were looking for. Link to Google analytics - slide 4

Looking at the referrals from Google we can see that searches for 'home based business' accounted for 25 visits. The CVHBBA is result number 6 for this search term. It also ranks well for the term 'home business'.

Lets to go back to our fictitious business. Mike would be looking at the usage information product page and the FAQ. He would track that against the staff time spent on answering questions to see if he was meeting his goal to reduce that time.

And ? he is meeting his goals.

Mike is happy with the website and now wants to use the site to attract new customers. Since he still wants to sell from his store, he needs some way of tracking how many visitors to the site actually purchase a widget. Mike's web developer says she will need to rewrite the front page with a focus on sales, add a price list and suggests adding a discount coupon to track buyers.

Lets take a little break from Google Analytics and try and put some of this information into context. And look at the whole online marketing process.

Online Marketing is form of direct marketing. We can assume that most people viewing the information have some interest in your product, since they selected to view that site. And, as we have seen, results are easy to measure.

Contrast this with broadcast media, such as newspapers, radio and television. In those cases, we can assume most people are not interested in the product since there is no overt selection process. Results, particularly of a single advertisement, are hard to measure.

However, just having a motivated viewer does not guarantee sales. Business site performance called the 'conversion rate' is the ratio between the number of visitors to your site and the number that actually reach the goal you have set for the site. It may be a simple goal, such as reading a particular page or it may be an online sale.

Using the figures from the CVHBBA site and our goal on having the visitor read one of two specific pages, we have a conversion rate of approximately 40% - the number of 'Unique Pageviews' divided by the 'Absolute Unique Visitors'. Since viewing a page doesn't require a commitment by the viewer, we would expect a high conversion rate. Typical conversion rates for online sales are much less between 1 and 2 percent. Which means for every 100 visitors you can expect 1 or 2 sales.

There are a number of tests we can perform on the website to improve the conversion rate. Results are easy to measure and changes can be made very rapidly. Contrast this with other media where results are hard to measure and changes are expensive to make.

One of the first test is 'usability testing'. Ask five or six people, that have never been on your website, to try use the site. Then watch what they do. Don't prompt them ? just watch. This way you can determine problems with instructions or even site layout.

The other common test for websites effectiveness is called AB testing. For those with a newspaper background, this also known as a 'split run'. It means showing half your visitors one version of the page and half a different version. Make small changes, even a simple change, such as moving the position of a button can effect site performance.

Again ? easy to do and easy to measure.

So ? how do we get visitors to the site.

Google Analytics will tell you exactly how visitors reached the site. However, most new visitors will come via a referral from a search engine. The 'Big Three' search engines Google, Yahoo, and MSN have some what similar algorithms for conducting searches. The exact algorithms are heavily guarded secrets. However, Google has published general outlines on how pages are ranked and guidelines for developing web pages.

Their page ranking system is called ' Page Rank ', which is named after Larry Page, one of Google's founders ? not web page. The scale is 0 to 10 and is generally thought to be logarithmic, but nobody, outside of Google, knows for sure.

The concept of Page Rank is very simple each link to your site acts link a vote for that site. The more links the higher the Page Rank. The implementation is more difficult. For example a link from a site with a Page Rank of 6 is worth more than one with a Page Rank of 1. And, one from a page on the same topic as your page is worth more than one from a page on a different topic. If you don't know your Page Rank, your web maintainer can tell you, or you can add the Google tool bar to your web browser and see the Page Rank of any page you visit.

Having said all that don't obsess on that number. First it is only updated every three months, and second, for most business sites, it is not all that important.

That may sound like heresy, since there is a whole industry called Search Engine Optimization, or SEO , dedicated to improving Page Rank. However, good SEO doesn't just focus just on Page Rank but also looks at the position on the Search Engine Results Page, or SERP . This is generally referred to as Search Engine Marketing or SEM.

The SERP is the page that is returned by the search engine. The search engine is trying to match the search terms being entered with what it 'thinks' your page is about. To get noticed your page needs to be on the top three pages being displayed. In the case of SERP a low number is best. Making the content of your page match the search terms will, in general get your page lower your SERP.

A word of caution. It is going to be very hard to get a good SERP for some general search terms. For example a search for 'real estate agent' yields over 9 million entries. It would be very hard to rank well for that term. However, you can expect to be able to achieve good results for the terms 'real estate agent courtenay' .

Pick what the search terms you want to target carefully; optimize your page for those search terms; then work on getting links to your site. All other things being equal, the page with the higher Page Rank will have the lower SERP.

As you may have guessed ? some of these items are not exactly compatible. If you design exclusively for a good Page Rank and low SERP, you may wind up with a page that is hard to use.

As most things in life, it is a balance. However, by setting goals and measuring your sites performance you can achieve that balance.

Lets go back to Mike's Widgets.

Mike is thinking of introducing his widgets in a brand new colour. Setting up the new colour will be expensive and he is not sure how popular it will be. His web developer suggests an online poll. She creates a graphic of the widget in the new colour using Photo Shop sets up a poll asking visitors to vote for the new colour. As an incentive, people filling in the poll are entered in a draw for a week end at Crown Isle. --- Turns out nobody likes the new colour. Spending a few hundred dollars on the poll has save Mike from making a BIG mistake.

Finally, Mike is seeing a lot of traffic to the site, but the number of coupons being redeemed is below what he expected. So he goes back to Google Analytics to try and find the answer. And, he finds it in a place he hadn't looked before. Link to Google analytics - slide 6

Over half of the visitors to the site were from outside the Courtenay/Comox area. Mike is considering online sales.

On-line marketing is an iterative process of constantly measuring and improving the effectiveness of the message. Everything I have mentioned, including on-line coupons are available for free on the Internet. By using these tools and working with your site developer or site maintenance person you can achieve your sites objectives.