Aaron Wall spotted Google showing Related Searches below search results. This is his image.

I really like these - we’ve been generating them for almost a decade. They made their first appearance on the original snap.com (which later became NBCi.com) using our GlobalBrain technology. We now show them on most of our customer’s site searches. Here they are on the Travel Channel:

Barry Schwartz on Search Engine Land asks how Google generates these. I’m guessing that they are popular search terms used to find that particular URL. As Barry pointed out, if you search for one of the suggestions, get more links, you’ll see the URL from SEObook ranking top.

This type of related search is reasonably popular on the site searches we host for our customers. Typically we will see about 5-10% of people who search will click on one of these links.

Last week we announced that Lighting by Gregory, one of the largest lighting and ceiling fan distributors in the U.S., has seen some impressive results from using our Learning Search site search and Site Champion user-generated SEO solutions. In fact, they’ve seen a 40% increase in their online conversion rate since deploying our technology.

With more than 150,000 products in their e-commerce inventory, the lighting retailer needed a robust, intelligent search solution that would give visitors an easy way to find what they’re looking for and not deliver items with little or no relevance in the search results. The company also needed something that returned results lightning quick and that could help customers further refine their searches. Their previous site search from Google just didn’t cut it. (Don’t get us wrong, we like Google and think their Internet search engine is a powerful tool (as does practically everyone else!), but site search is not their sweet spot. It IS ours, as Lighting by Gregory’s results show.)

Lighting by Gregory initially underwent a 30-day free trial, a standard offer for any new customer, and saw how quickly SLI went to work for them. They signed up immediately after the trial ended and have continued to see positive results.

As Lighting by Gregory learned, site search users usually bring in more revenue on average per visitor than those who opt to simply navigate the site. If your site search isn’t delivering the results it should, we’d love to go to work for you too.

We recently released an autocomplete feature that’s been picked up by many of our clients. This uses previous user’s searches (in aggregate) to automatically create keyword suggestions, which are then displayed in order of popularity. It is really fast too. Head over to surveillance-video.com if you want to see it in action.

Prior to the release we had trouble estimating how often people would use it. We couldn’t find any usage stats out there we could use as a baseline. While it is still early days, it is currently being used 20% of the time, and on average people are saving themselves from typing about eight characters each time. Saving a few keystrokes is great, but the real benefit appears to be helping people use search terms that are hard to spell – spelling suggestions are required much less often on sites with this technology. This means more people are finding what they want the first time, which has to be a good thing.

We strive to continually improve the service that we offer our customers. One part of that is ensuring that the search and navigation pages load quickly. The graph below shows how the average time (to serve the complete search page) for one of our customers has improved over the last 12 months.

Even though this customer has a complex search, we thought 2 seconds was a little slower than we would like. We’re much happier with it sitting below half a second.

The changes that you can see are improvements to our algorithms and the configuration of our servers. They are examples of improvements that all of our customers receive the benefit of (with no additional cost to them).

Today we’re announcing that Spencer’s and its sister company, costume retailer Spirit Halloween are using our Learning Search and Site Champion services. Spencer’s sells humorous gifts and Spirit Halloween sells costumes.

The two e-commerce sites (the company also operates brick-and-mortar locations around the country) see their biggest traffic around Halloween and Christmas. It’s important that we are able to handle these seasonal spikes to ensure that their customers can always find just the right products.

Jay Greenberg, director of ecommerce for Spencer’s, had been an SLI customer at his previous position at Franklin Electronics, and was so happy with our service that he advocated using us at his new company. We’ve seen this several times - when individuals move companies. For me this underscores the importance of ensuring that our customers receive the best service that they can.

You can read more details about Spencer’s and its choice of SLI here: Spencer’s and Spirit Halloween Improve Customer Experience on the Web Using Learning Search and Site Champion From SLI Systems.

Steven Seiller wrote an article outlining the 11 things he hates about site search. He makes some good points and captures the passion that people feel when site search isn’t there, can’t be found, or doesn’t work.  You need to have site search, make it highly visible and it should be flawless. This is not easy to do and the disappointing thing is that if you do it well most visitors won’t notice it (people notice things that don’t work properly) - but they will be happier and will get more from your site.

I may sound like a broken record here - but if you want to improve your site search we would love to show you what we can do. Contact us and we’ll try to dazzle you.

Avi Rappoport did an excellent post on search log junk. This is the set of queries that aren’t very useful  when you’re analyzing search logs, mostly from crawlers, hackers and spammers. We see a lot of these types of queries and constantly trying to block them and remove them from our reporting - so we can extract useful information. I hadn’t seen anyone talking about this before - but it is something else you need to be aware of when you’re trying to extract the valuable information that is in search logs.

Last week Coremetrics announced that they making free ecommerce data available to their customers for benchmarking. Their announcement included some sample data including the fact that 15% of consumers used site search, resulting in a conversion rate of 5.6%. The conversion rate is interesting when compared to some of the other conversion rates in the sample data: from natural search traffic the conversion rate is 1.66%; from comparison shopping engines and affiliate networks it is 1.97%; and from direct load it is 3.29%.

This is yet more evidence showing that the conversion rates for people who use site search are relatively high.

As to the question around the value of the benchmarks - I think they’re still worth having even though it is hard to compare because of the variety of ways of measuring metrics like conversion rates.

Last week we sent out our latest edition of Telescope on Search, our monthly email newsletter on; site search, seo, search engine marketing and usability.

The March 2008 Newsletter contains:

• Black Forest Decor Discovers the Importance of Site Search Size and Placement - An article that highlights the benefits received by increasing the size and improving the location of the site search box.
• Recent conferences where we are exhibiting, including some special discount offers.

The image below is how the site looked before the site search box was changed.

The image below is how the site looks after the changes to the site search box.

To read the complete newsletter click here. You can subscribe to our newsletter at our newsletter page.

I spotted this post touting the benefits of mining your site search. It’s a point I’ve made before but it’s valuable and worth repeating - your site search logs are a fantastic source for keyword research. It’s worth looking at these to get keyword lists for PPC & SEO. And you should be looking at them to make sure that you have good results on your internal search for the terms people are actually using.