Apple is one of the most visited ecommerce sites. They have an image of being slick and user friendly. I had high expectations of their website – which they meet for the most part: their product pages show off their products wonderfully and it is easy to shop. However their search is terrible. I was searching for the technical specs of the iPhone, typed in “ipone” and got no results. They don’t have any spelling correction.

I then typed it in correctly and was gobsmacked when I saw no results.

Finally I got some results when I searched for phone:

Even when I searched for an old product the search results were disappointing. The top two results are press releases (one from New Zealand of all places). They don’t have any images, prices, sort options, and only very limited filtering options. This is not what you’d expect from one of the top Internet retailers.

So how big was Cyber Monday? I took a look at the search traffic on a handful of our retail customers’ sites and compared their search traffic on Monday November 27 to their average search traffic for the 3 months before Thanksgiving. On average there were 90% more searches on Cyber Monday compared to the average, 70% more than on an average Monday.

Note: A couple of the customers showed no significant increase. One customer saw a 400% increase! The regular variation you can see here is the typical pattern of there being less activity during the weekend. But as you can see, the holiday season is well and truly upon us.

Lots of people criticize the reliability of the Alexa toolbar – see the comments on Matt Cutts’ blog or more recently from Om Malik. I would like to offer some qualified support for the Alexa toolbar – with some data to back it up.

At SLI we run the site search for lots of sites that have a wide range of Alexa rankings. Assuming that the number of site searches on a site is roughly proportional to it’s traffic we can use the site search data to test the accuracy of the Alexa data. This graph plots the Alexa rank against the number of site searches performed over a 30 day period for about 50 sites (randomly selected from our customers).

As you can see, there is a strong correlation between the Alexa rank and the number of searches that happen on a site, although there is a reasonable amount of variation.

We use the Alexa rank to get an order of magnitude estimation of the number of searches for a site, and it’s normally pretty good.

The assumption that number of searches performed on a site is proportional to traffic isn’t going to be always right. Some sites lend themselves to searching – particularly larger sites.

Some sites have completely unreliable Alexa rankings – particularly those that attract more (or less) people that have the Alexa tool bar installed. Case in point is the SEMPO web site. As others have pointed out, a higher proportion of search marketers have the Alexa toolbar installed. The SEMPO site has an Alexa ranking of 19521 (today) and had less than 5,000 searches over the last 30 days. This model would predict around 150,000.

I think Alexa is a useful service, providing that you use it carefully. I think there are a lot of people out there who have dismissed it as flawed and may be missing out on some valuable data.

We continuously try to improve our best practices around site search. Recently we’ve been looking at layout options. Specifically, which is the best layout grid or list.

The grid view is more compact and so allows you to show more products above the fold. It is potentially easier to browse through many results using this format, particularly if the images adequately summarize the results. On the other hand, the list view gives you more room to show additional information about each product. In particular you can show descriptive text, related information and add to cart buttons. The text gives you an opportunity to further entice the user to click on the result. If you highlight the search term then it provides a strong information scent indicating that the result is relevant for the given term.

It’s interesting to look at how the distribution of clicks varies between the two different options. This chart shows the number of clicks that occur at each rank when the results are ordered in a list view.

As expected the top ranked result gets the majority of clicks and this falls away fairly quickly. This particular client has 10 results per page. This is reflected in the chart in that the 11th result gets significantly less clicks than the 10th result because a lot of people don’t visit the 2nd page of results. You can also see that the 10th result gets about the same number of clicks as the 9th. I suspect this is because it is at the bottom of the page and easier to find than if it were the 10th result in a long list. For some customers we see the result at the bottom of the page get more clicks than the result 2nd from the bottom.

Contrast this with results that are shown in a grid view (in this case 3 rows of 4 results):

Here you can see that the number of clicks doesn’t fall away as fast as they did in the list example above. Interestingly the 5th result (which is below the 1st result) still gets less clicks that the 4th result (which is at the end of the first row). I had though it might get more if a reasonable amount of people scanned the list vertically. However this supports the premise that the results are scanned from left to right.

Our customers are reasonably evenly split between showing the results in the grid and list views. Many of them allow the users to switch between the two. Our analysis indicates there is no significant difference in the conversion rates of the two formats.

So which should you use? The charts above summarize the activities of thousands of people. My advice is to ask some of them and if possible test the alternatives. Personally, when I’m searching I prefer to have more information on the search results page because that makes it easier to make a decision without having to click through. That saves me time.