We’re back with BrickHouse Security’s Director of Acquisition Ryan Urban. After introducing us in the part one of our interview to a tiny computer surveillance device that surreptitiously records everything on a computer, we’re back to focus on how he maximizes online revenue and whether the future of site search is visual.
Let’s move to technology and the nut and bolts of operating BrickhouseSecurity.com?
We host with Yahoo! Stores. We moved to that platform in late 2005 and the reliability has helped us become an Internet Retailer Top 500 site.
What are some of the vendors you use?
Yahoo! for hosting. Google, Omniture, and Yahoo! for analytics. We use Nextopia for site search.
How important is site search to your company?
Fifteen percent of our revenue comes from the search box on our site. A lot of our site visitors do not land on our home page. Site search really helps our visitors who arrive looking for specific things. They want to navigate directly to the product rather then click on links. A lot of people want to click on the product images that we display as part of our Nextopia AutoComplete module.
How do you integrate site search into your website?
We’ve been using Nextopia’s EcommISearch module for the past year. It integrates very well into the Yahoo! Stores environment. Nextopia has also spent a lot of time ensuring that it integrates well with Google Analytics.
I can see every search term typed in, which search terms converted and exactly where they converted. I see the traffic sources that are directing people to our site, and our revenue per customer. I can easily see whether we need to do a better job of optimizing our landing pages.
I’m sure you spend hours looking at logs, Do any particularly weird search terms jump out at you?
Strangely enough, the most common search term is nothing. Lots of people don’t enter a search term, yet still click submit. We just direct those searchers to the new products page.
Long tail searches on our site can reveal some really “interesting” interests that people have. Queries relating to monitoring or recording cell phones are pretty common. Unfortunately, that’s illegal so we can’t help them. Lots of people are interested in detecting bugs or are suspicious about the possibly of hidden cameras lurking in their homes of offices. They think their smoke detectors might contain hidden cameras, and many times they are right (see Erin Andrews).
Two other interesting product categories are covert surveillance and GPS Tracking. We have taken products normally sold to the government, and made them consumer friendly in terms of use and price.
Let’s talk specifically about Nextopia’s AutoComplete with Images technology.
AutoComplete with Images has really helped us. Let me explain – we first started off using Nextopia’s AutoComplete module which just displayed keyword suggestions to our customers. We later upgraded to their affordably priced AutoComplete with Images module which extends keyword suggestions to includes top converting keyword and product names and thumbnails that link directly to product detail pages. Think of it as a guided search box experience.
How does it work?
As soon as a customer begins typing into our search box, they are automatically presented with both keywords and products relevant to their search. The display instantly changes as a searcher inputs more characters. It starts predicting as each character is entered and dynamically changes the display of suggestions. Visitors end up seeing related queries and products right away, bringing them close to the point of conversion. So, customers can either click on a suggested keyword and proceed to do a search on our site OR they can click directly on a thumbnail of relevant products and be directed right to a product detail page.
Nextopia's AutoComplete site search technology on BrickHouseSecurity.com
Ex. A customer searching for the term ‘usb’ is “autosuggested” 5 top keyword suggestions for search and 6 top converting products with ‘usb’ in the product name. With each keystroke, suggestions dynamically change to display the most relevant set of suggestions.
What issues were you trying to address by integrating it?
First of all, we wanted to direct people to what they want and need. We found people often don’t use the best search terms so a system that suggestively prompts them is much better. A lot of people search for uses of products; not necessarily the products themselves. We wanted to enhance this process. We’ve found that product images are the best way to direct people. We now put them immediately in contact with the product.
We wanted to present historically strong sellers and direct people to those pages. The Nextopia technology pulls data from our sales history and changes as sales rise and fall over time and incorporates this data into their algorithm. We also spend a lot of time optimizing these pages. We ensure that our product pages include cross-selling information, recommending other products and coordinating products.
Another issue for to think about is this…if your product pages aren’t very good, then you might be better off putting them in contact with a product list than giving them a choice of which ones to select.
What was the AutoComplete implementation process?
It is real easy since we were already an existing eComm|Search customer and were already submitting tracking codes for each product SKU. That was it. Nextopia’s developers did the rest.
What kind of metrics are you seeing with the AutoComplete with Images module?
Upgrading to Autocomplete with Images was a great move as it converts twice as well as the base AutoComplete module. We started showing four images and then expanded to six images per query. Half of our searchers click on the product images in the search now. Overall, revenue is 8% higher. We’ve seen close to a 2% rise in overall site conversion, which I think will jump to 3-3.5%. That is very big for a site like ours.
Any other comments?
Nextopia’s AutoComplete with Images module is worth thousands and thousands of dollars to a larger website. In addition to the revenue it creates for us, it makes it really easy for our staff to find products. In fact, pretty much everyone in sales and support uses it to quickly find products in our catalog, even if it wasn’t designed for them.
I’m also experimenting with the number of pictures to display. I have six images showing now. If I want, I could have 8-10 pictures. I’m going to keep testing. I know that if I put it 10 images at a time, I think I could lift sales another 15%. It’s pretty tempting to fill up the screen and quickly give direct people to the highest selling products.
Finally, we’ve found that more people who use site search arrive organically than through PPC ads.
That’s great info Ryan. Thanks for your time.
Thanks to Internet Retailer for highlighting ToolFetch’s experience (“
