Website Branding - America (Favorite) Icon

Have you ever noticed how certain websites in your favorites bookmarks have nifty little pictures next to the Link Title instead of the boring old IE symbol? This little piece of branding is the result of a nifty option available to any website, the use of a Favorite Icon. I thought it would be beneficial to talk about how to create a favorite icon, and then as an added bonus to our loyal readers, how to use this as an analytic tool.

First, let's talk about the basic of a Favorite Icon. The Favorite Icon is a graphic file placed in your root directory that IE searches for whenever a visitor using IE enters your website. If you have a favorite icon installed, IE will use this icon in place of the IE icon normally associated with bookmarks. If you do not have a Favorite Icon, then IE will use its own icon. So this means that any website may use the Favorite Icon option and replace Microsoft Branding with the IE icon with an image that brands their own website.

Favorite Icon requires that the image in question be saved in [.ico] format. In other words, the image you use for your favorite icon must end in [.ico]. It also requires the icon be 16 x 16, 24 x 24, 32 x 32, or 48x48 pixels. This begs the question, how do I save a graphic in this format? Thankfully there are some free programs that will help you change any image into an icon format image. Furthermore, the program will actually save multiple copies of the icon in as many or as few sizes as you require and store them all in that one .ico format.

My choice of programs would be AnytoIcon by aha-soft. This program allows 30 trial uses of the program before needing to purchase the software. Unless you are a web developer working on more than 30 websites, you will probably only need to use the program a few times and maybe only one time if the first image you choose is suitable to your needs. This means you can incorporate Favorite Icon into your website today, right now, for FREE! Aha-soft's program allows you to simply drag your image into the program, choose your sizing or custom options, and the program instantly creates the Favorite Icon in its proper .ico format. Then simply upload the icon to your root directory, and you are done. Every visitor who bookmarks your page from this point forward will see your chosen icon next to your link in their bookmark list!

Now, I promised I would tell you how Favorite Icon as analytic tool, but perhaps some of you can see where I am going with this. Once your Favorite Icon is active, simply use your analytics program or web logs to examine how many times the favorite icon was download on a particular day or week. Remember, Favorite Icon is only accessed when a visitor chooses to bookmark your website, so this resultant number will indicate how many visitors bookmarked your website over any given time period.

Now take the number of times the Favorite Icon file was accessed and divide it by the total number of visitor to your website over the same time period. The resulting figure is the percentage of visitors who bookmark your website! Over time you can track this percentage to determine if your content/products are appealing to customers. As you make modifications to your site's content you can measure if Favorite Icon % is increasing or decreasing giving you a quick baseline figure to see if your changes are helping you to retain and recirculate visitors!

With about 5 minutes work you can brand your website with a custom image icon, and add another useful and easily calculated metric to your analysis that measures the appeal of your website for repeat visits. So why wait? Post your Favorite Icon image today and start branding with the big boys!

SEO & PPC Metrics - Website Page Depth

When managing an SEO initiative or PPC campaign, one of the more challenging aspects is measuring the lead quality of the inbound visitors. Most people know how to measure Return on Investment (ROI) and Return on Ad Spend (ROAS), but that should not be the only metric used to evaluate the quality of the leads coming from different Search or PPC Engines. One of the best metrics for measuring lead quality is Page Depth.

Page depth is simply the average amount of pages a visitor sees during a session on your website. It should be fairly obvious that higher page depth is better than low page depth, but how to figure the metric correctly and determining a page depth baseline can be a bit more involved. FOr this excercise we will look at some actual PPC data from a major online retailer, but first let's talk about how to measure page depth accurately.

We know minimum page depth must be 1.0 as it is impossible for any visitor to see less than 1 page. Most people would think page depth is simply calculated by: (Total Number of Page / Number of Vistors). While this is correct in the general sense, page depth should be adjusted to eliminate shopping cart pages. Let's look at a hypothetical example before proceeding to an actual data set.

Suppose your website showed 10,000 pages last month to 2,000 visitors. Let's also assume that your website converted those 2,000 visitor into 200 sales. Finally let's assume that your shopping page is 3 pages long (this means that it takes 3 pages to fill out personal & billing info, and a confirmation page.

Using the tradional Page Depth formula, Page Depth would be calculated as:
10,000 pages / 2000 visitors = 5.0 Page Depth

Now let's look at a calculation for page depth that adjusts the page depth calculation to remove shopping cart pages. The reason for this is that visitors who convert are forced to view 3 more pages, 1 of which (the confirmation page) is completely superfluous. The reason I say this is that these pages do not measure the depth of a visitor's session because they happen after the visitor has converted. Page depth as a metric should measure how far into a site a visitor goes regardless of conversion or no conversion. What Page Depth should be measuring is how conducive your website is to dispensing information about products or services you offer. Also, typically service wensites do not have a shopping cart while e-commerce site do. By adjusting page depth to exclude shopping cart pages, page depth can be compared between types of sites without a shopping cart bias "helping" e-commerce sites.

So let's do the Page Depth Calculation again and exclude shopping cart pages:
(Total Pages - (Conversions * Shopping Cart Page Length)) / Visitors
(10,000 - (200 * 3)) / 2,000 = 9,400 / 2,000 = 4.7 Page Depth

Notice that page depth dropped from 5.0 to 4.7 using the adjusted page depth formula. However this 4.7 pages calculation tells us that any average visitor will typically see 4.7 pages before either leaving or converting. If the 5.0 figure were used, it would be skewed by shopping cart pages that happen after the visitor has already been converted. Also all we could say for certian is that visitors who do not convert see less than 5 pages and those that do convert see more than 5 pages, but it does not tell us what the average visitor sees whether they convert or not. If you have a long shopping cart process or if you were to convert a high percentage of visitors, the tradional formula of [Total Pages/Total Vistors] would become even more skewed. Therefore, I have found that the adjusted page depth calculation gives a more accurate indication as to how many pages a typical visitor views since we do not know beforehand if they will convert or not.

Developing a baseline page depth figure should be down over time and should be done to each engine or campaign separately. You will also have an overall page depth figure for all campaigns but by using this figure it will be more difficult to spot meaningful fluctuations as high traffic campaigns will be weighted far more heavily in the overall Page depth figure. I would say that 1 month is a good amount of time to collect data in order to calculate reliable Page Depth figures for each natural and ppc campaign. However, there are red flags to look for that can indicate a problem in less time than a month.

Let's look at the PPC campaign data from a major online retailer doing around $10 million in gross revenues. (Click on the graphic below to see full image)



Notice the Raw (Unadjusted) Page Depth Figures for Kanoodle & Enhance [Not surpisingly, this figure also happens to be the adjusted Page Depth as neither engine converted a single visitor]. Anyone who has read this blog or my contributions to various search engine forums knows that I have a dislike for these two engines and here is the reason why. Enhance has a page depth of 1.34 and Kanoodle a rather unbelievable 1.10! Kanoodle's Page depth figure says that for every 10 people that visit the site, 9 of the 10 bounce (Bouncing is a visitor who sees only the landing page they are directed to and then leave), and the tenth sees a second page and then bounces! This is the telltale sign of poorly qualified traffic or click fraud. I will get into click fraud in a future post, but be aware that the smaller PPC engines and Shopping feeds are notorious for click Fraud. Among these would be Kanoodle, Enhance, Bizrate Shopping, & Nextag Shopping. Not only do they send unqualified traffic, but many of them also charge 200-300% more clicks than any analytic program tracks. The Page Depth metric is an excellent tool for spotting these types of fraudulent traffic sources. Statistically speaking, it is highly unlikely for any visitors to convert from sources that generate a raw(unadjusted) Page Depth of less than 3.0. Why, well because most e-commerce site have anywhere from a 2-5 page shopping cart, so a raw Page depth of less than three indicates that few if any visitors from that sources are making it to or through the shopping cart. If they were converting, raw Page Depth would have to be higher due to the additional pages a shopping cart adds to the page depth metric.

Naturally, this retailer no longer advertises with Kanoodle or Enhance. Furthermore, the Page Depth figures for the new Miva Engine show signs for concern [Miva combined Find What (miva-f) & Espotting (miva-e)]. During July Miva will be monitored for Page Depth and if things do not improve significantly, advertising will be halted on Miva as well.

Hopefully this gives you an idea on how to calculate and use Page Depth as a significant metric when evaluating your campaigns. Remember, that it could be your site's architecture (confusing navigation), bad landing page choices, or your site's appearance that could be hindering vistors from seeing more pages. Assuming that there are no appalling usability issues plaguing your site, Page depth can be an invaluable tool for measuring the quality of leads coming from your different traffic generating sources. Being that it is a simple caluclation to gain Page Depth figures, take an hour or so and examine this metric to determine which campaigns are working for your site, and which are simply trying to get into your wallet.