When was Quintura launched? What gave you the idea to launch it? What problems were you trying to solve by launching it?

Quintura was founded in August 2005 and released its first search application in November of that year. One year later, we launched a web-based search. It was based on visual context-based search concepts that the founders had been developing since 1990s. Quintura was founded to solve several fundamental problems inherent with today's search engines. Those problems include too many irrelevant search results returned, no one reads past the first page of results; inability to manage or tune results by defining context or adding search scope; no means for users to graphically visualize search terms or manage their relationship/relevance. Quintura is designed to make it visually simple for searchers to find what they are looking for, and to make it easy for web publishers to expose the content their visitors are looking for.

You guys have got a lot of great press from tech bloggers. On the marketing front what are some of the biggest and most successful surprises you have encountered? What have you found to be hardest when marketing your search service?

The simple fact that there is a tremendous amount of interest in our technology and service, in spite of the large field of alternative search engines on the market. We've invested most of our time and efforts in research and development. Our biggest challenge has been in getting our first marketing message out, which is we're in the process of expanding now to mainstream media.

How do you guys generate the keyword clouds?

That's part of the magic behind the Quintura technology. At the heart of our technology is a semantic-based 'neural network' algorithm. The cloud is literally a depiction of those search terms laid out to show their contextual relationship. Since the graphic depiction is dynamic - (you are interacting with the search in real time) one of our design goals has been to develop the widget to be extremely responsive. Through the past year, we think we've reached that point.

Quintura is popular as a keyword research tool amongst many SEOs (I use it all the time). Have you thought about combining your service with search volume data and/or competitive research data to create a formal premium keyword research (or competitive research) service/tool?

We've been asked that several times, but for now, our goals are to provide the best consumer site search services to the market and to provide our search widget to as broad an end-user audience as possible.

Quintura makes boolean search easy to visualize. Do you think searchers will eventually start using advanced search operators more on general web search engines, or will most only use it when it is presented in an aesthetically friendly way like Quintura does?

The question is whether users want to become adept at boolean logic or would they prefer to have that hidden in the background. From our experience, users would prefer to focus not on the math but on the search itself - finding the most relevant results in the least amount of time. By laying out search terms contextually and graphically, Quintura helps users manage their search and be in control of their search.

When partners sign up for Quintura you guys create a custom index from a crawl of their sites. How many domains can be part of the same index? What sort of sites does Quintura's visual search work great on? Which ones are not as strong of a match?

There is no limit. We're glad to work with large web publishers directly to assure that we are indexing all important content as part of our site search solution. The publisher of several web-sites can create a “vertical” search engine based on the Quintura search cloud. Quintura works well with all web-sites that we have worked with to date including numerous amount of blogs. Though, our first major site search clients were lifestyle portals and lifestyle magazine web-sites.

Do you see the face/interface of general web search changing drastically in the coming years? How might it change?

The web is getting more visual. So is search interface. That’s the trend. We are enabling our content-publisher customers to be more creative through customization of the widget itself. We're also looking at ways to make the search results even easier to see through the use of even more graphics.

Does Google have general search locked up? What competitive positions might allow people to build out a strong competitor that can take marketshare from Google?

General search is mostly locked up with Google. In my opinion, the best way of taking a marketshare from Google is not by building a better search destination site, but by changing the paradigm – give reasons for users not to make a decision to go to a search engine. Because when the think search engine, they think Google. Essentially, what Quintura site search does is creating environments where users keep exploring the passions, their interests, their information needs from where they are on the Web. People go to search engines when they can’t find what they want where they are.

Chitika has created a fairly large sized behaviorally targeted ad network by targeting ads to the search query prior to people landing on a page. Your site search strategy seems like it could be a rather powerful strategy for building a strong network. How has growth been going? Do you have any interesting success stories from the publisher or advertiser standpoints?

Quintura currently powers site search for a monthly audience of 8 million site users. The tests are underway on various U.S. sites, including two major men’s lifestyle sites and an educational publisher. We plan to reach the audience of 50 million in 2009. You can see Quintura search widget on lifestyle sites Maxim.com, Passion.ru and Cosmo.ru; technology news sites ReadWriteWeb.com and Compulenta.ru, business community portal E-xecutive.ru; web-sites of consumer magazines Hilary Magazine, Russian Newsweek, ComputerBild, luxury news site LeLuxe.ru, in addition to hundreds of smaller web-sites and blogs that joined our affiliate program for site search. We have also approached several online advertisers including security software vendor Kaspersky Lab to advertise on our search widget network of sites.

What types of ads work especially well with a service like Quintura? Which ones are less strong?

We have tested both contextual search ads and display ads. We are going to blend search ads with display ads for more visual appeal. Plus, can target those contextual graphic search ads with much greater precision because of our context-based algorithm. Ads from companies with established brand logos benefit from our ability to graphically display their logos in the search cloud.

What areas does Quintura have a lot of inventory in?

It is in lifestyle and technology areas.

Many search engines (Google, Yahoo! Search, Live), large content & commerce sites (Amazon.com, eBay, Wikipedia), and browsers (IE8 Beta 2, Google Chrome, Firefox 3) are now adding search suggestions in the browser via the search box and/or address bar. Do you see this eventually evolving into a Quintura-like service?

It’s a helpful feature that is mostly based on search statistics. We go a step further by offering contextual suggestions. One of the greatest aspects of our display cloud is that it shows contextually-related results, and to depict them with a graphical element. Can you imagine a shoppng experience that lets you see related items in real time?

Quintura is currently powered from the Yahoo! index. Do you guys ever plan on creating your own web-wide search index?

As a matter of fact, we are already creating our own web index from individual indexes of web-sites where Quintura powers site search. Quintura site search on those web-sites is powered by search results from Quintura index of those sites.

How many regular users does Quintura.com have as a search destination? Do you guys intend to become a consumer search destination, or are you more focused on providing search for third party sites?

We focus on providing site search, analytics and monetization platform for web publishers and content owners. As a search destination, Quintura has less than 1 million users per month. We will continue operating and developing our search sites to provide the benefits of our search technology to users. For example, Quintura.com will evolve into an online research tool where registered users will be able to save and share their searches online with the other registered members.

You guys have a vertical search service for kids. Is that seeing good adoption? Do you plan on coming out with any other vertical search engines?

Children are far more graphically oriented and can grasp contextual depictions easily. It was a natural extension for us to offer a search engine designed specifically for children - Quintura for Kids. It's also a great test bed for us to further evolve search technologies while giving kids a hand. The search engine is used mostly in the elementary schools and public libraries in the U.S., Canada, Australia, and New Zealand. Since its first launch in March 2007, several hundred school and teacher web-sites linked to Quintura for Kids. According to site statistics, the search engine has 70 percent returning visitors. 75 percent of visitors come to the site directly from a browser. In June 2007, Quintura for Kidswas ranked the highest among search engines for kids by Search Engine Watch.

We evaluate additional opportunities including licensing our technology to intranets and major search engines.

For now, our hands are full with upcoming site search product enhancements and monetization as well as  with our growing site search customer base.

I'm reading a book called "Predictably Irrational", by Don Ariely. It's about the hidden, irrational forces that shape our decisions, and it's a great read.

There are a few interesting case studies in this book that can be applied to web marketing. I'd like to look at two aspects which might help those of you involved in e-commerce.

Relative Pricing Structures

The first experiment looks at relative pricing structures. How do you structure your prices in order to achieve higher returns? Often, it can be a simple case of making an offer no one in their right mind will accept.

Huh?

Here is an example.

The Economist presented readers with the following subscription offer:

1. Internet only subscription for $59
2. Print only subscription for $125
3. Print and Internet subscription for $125

Notice something odd about option 2 and 3? Why would anyone take up option 2?

They wouldn't. And that's the point.

It turns out humans rarely choose things in absolute terms. We work out how much things are worth based on what other things are worth and compare them. In the above example, the "Print and Internet" offer is better that just the print offer. The "Internet Only" offer might be better than both, however there is no point of relative comparison for that offer. The relative comparison is made between offers 2 and 3, which makes option 2 look poor, and option 3 look like a steal. Ariely ran real tests to measure take-up, and sure enough, most people took option three.

To illustrate how powerful this pricing method is, let's remove option 2.

1. Internet only subscription for $59
2. Print and Internet subscription for $125

In this example, people are faced only with a cheap option and an expensive option. The point of comparison is largely about price. You can guess what option most people chose. They selected the cheapest option, as price becomes the key differentiator.

So, try splitting your offers. Create offers that are valuable compared to other - deliberately substandard - offers.

The Effect Of Expectations

In another chapter entitled "The Effect of Expectations", Ariely asks why the mind gets what it expects, and not necessarily the reality of a situation.

For example, Ariely conducts an experiment whereby researchers offer students a free cup of coffee, along with some rather unusual condiments, such as cloves, nutmeg, orange peel, anise and sweet paprika. Not the sort of thing you'd likely put in your cup of coffee! The students were asked to rate the taste of the coffee, and specify the maximum price they were prepared to pay for a brew.

From time to time, the researchers made one subtle change. They placed the condiments in a range of containers, from rough styrofoam cups, through to beautiful glass-and-metal containers. The condiments were never actually used, however the mere appearance of the serving bowls had a curious effect. When the condiments were placed in luxury containers, the coffee drinkers were more likely to say they liked the coffee, and whats-more, they were prepared to pay a lot more for it.

If people thought the coffee was upmarket, they convinced themselves the coffee was upmarket. The reality was that the coffee never changed. The coffee was of the same quality throughout the experiment.

Self-evident, right? So, can this theory be directly applied to web marketing?

Essentially, we're talking about branding. There is the logical first step of using upmarket web design in order to help convince people your product or service is more desirable. There is a trap, however, and this is the reason I think this case study needs to be adapted for the web environment. When it comes to e-commerce, upmarket, glossy sites do not necessarily result in higher sales. There are various reasons for this, but I think mainly it has to do with the level of trust. A slick website can sometimes feel impersonal, and people crave a personal feel on the web.

Trust, not slick graphical design, is the equivalent of the elaborate serving bowls.

In order to raise expectations, consider raising the level of design, but only if you do so without losing trust. Achieving a fine balance between excellent usability, trust metrics and excellent graphic design is a great target to aim for.

Consider the converse. Have you bought from sites that are unusable? Plastered with over-the-top Adsense? Such sites are less desirable as expectations are set low, primarily because of the low level of branding. The buyer is expecting "cheap". That's probably the only reason people buy from such a site.

Such sites are the web equivalent of broken styrofoam cups, compared to the elegant serving bowls.

Have your say

What do you think? Have you got any "irrational strategies" to share?

I've been trying out Google's Chrome browser. I like it. I really do.

I like Chrome mainly because it is fast. Faster than Firefox, anyway. However, I'll be alternating between the two browsers, because Firefox has a plethora of useful plug-ins that Chrome lacks.

Like many Firefox converts, I haven't looked at Internet Explorer for some time now.

Microsoft have recently released IE8, so I thought I'd evaluate it in terms of search, and contrast it with the functionality and positioning of Chrome. Many in the internet community have speculated that Chrome is going to eat Microsoft's lunch, and not just in the browser space, but with the ushering in of cloud computing. Is this plausible?

Let's take a look.

You can download IE8 Beta from here. As usual, you'll have to sign your soul, and those of your yet unborn children, etc, etc over to Microsoft, and then reboot.

Goodbye Google Toolbar

You run through the inevitable setup screens. The first search-related issue I noticed was that Google's toolbar wasn't compatible with IE8 beta, and asks me if I want to disable it. Is a bug, feature, or a market position? ;)

Next up, IE8 asks you if you want to use "Express Settings", which means that the search provider will default to your existing default, and just about everything else defaults to Microsoft products or services. Internet Explorer also wants to become your default browser. At this point, you can opt for Custom Settings, and modify each setting individually.

Pretty flexible, really. If you want to opt out of Microsoft services, you can do so easily.

The Search Wars

My main reason for looking at IE8 is in terms of search. What functionalities do you get, and how is this browser positioned against Google?

Search Suggestion

One feature, called Search Suggestions, offers, naturally enough, search suggestions. Like the equivalent Google feature, IE8 will try to guess what keyword you are search for a prompt you with suggestions as you type. This feature works with many different search providers (Google, Yahoo!, Live) and large ecommerce and content sites (Amazon.com, eBay, Wikipedia), which makes the search box a nice keyword research tool, but nothing new to most of us, I'm sure.

Note that this type-ahead feature, like on all browsers offering type-ahead suggestions, will send your search queries to your search provider, even if you don't hit send. Matt Cutts, perhaps sensitive to the privacy concerns aimed at Google, makes the point in this comment he posted on GoogleBlogoscoped that " if "Suggested Sites" is on, "your web browsing history is sent to Microsoft, .... the addresses of websites you visit are sent to Microsoft, together with some standard information from your computer such as IP address, browser type, regional and language settings.....".

How Will This Affect SEO?

An aspect SEOs need to consider is how the widespread implementation of search suggest is going to affect SEO. In this post, Aaron talks about how search suggest is likely to force a consolidation around the most popular terms. This has implications for those going after the long tail, but also provides new SEO opportunities, especially if you have a brand that incorporates popular search terms.

Explorer also allows search suggestion from any provider, which can be a useful SEO tool, in itself.

Visual Search

IE8 also offer Visual Search, which provides pictures to help you select a result. This didn't seem to work for me, but I did notice that a search on "Seattle Weather", the search term suggested by Microsoft, did bring up a page featuring advertisements for Australian outdoor sportswear suppliers. Reminds me how far other providers have to go in this text ad space in order to catch up with Google. It wasn't until I dug around a bit further that I discovered that you need to install search providers. Even then, it wasn't playing well, giving me a string of error messages.

Still, problems are to be expected in a beta release.

Other improvements include search history matching, a useful "Find On Page" button added to the instant search box, and the ability to drag the search box in order to change the width. A few nice touches.

Forced Search Provider?

On the Microsoft global-domination conspiracy front, far from locking you in, Microsoft have made it rather easy to configure IE8 to incorporate your choice of search provider. It wants to default to Live Search, but you can easily select Google, or other services. The pull-down search box provides options to add more. So, good marks in terms of flexibility.

There are various other features, including InPrivate browsing, which supposedly blocks ads and prevents people tracking you across the web. As it isn't search related, I won't review it, other than to say it is good that the user has to jump through a few hoops to enable it. Love 'em or hate 'em, web ads enable the production of a lot of "free" web content. If ads were turned off by default, many sites would simply cease to exist, or start charging for content. Full marks to Microsoft for leaving this option to the power users.

IE8 Vs Chrome

Now, contrast these features with Google's Chrome.

Did you find Chrome noticeably faster than your existing browser, be it Firefox or IE?

I did.

Speed was the deciding factor for me. On the internet, speed is (nearly) everything. IE8 didn't strike me as being any faster than Firefox, and certainly a lot slower than Chrome.

In this respect, IE8 feels like an update to an existing product, as opposed to a game changer. Chrome feels like a game-changer, even though, when pushed, I can't put my finger on exactly why this is. I think it may come down to the usability gains of extra speed, especially if your day to day use orients a search function. IE8 is adding functions, desktop application-stylee, while Google is busy taking features out in order to simplify and minimize.

If cloud computing is to take off, then the browser is going to need to need the speed of an application, and it is going to need to be simple and transparent in order for people to bother migrating.

Application-Centric Vs Web Centric

Chrome explains itself better. The Google information pages tell a cohesive story, whilst Microsoft's story appears scattered and a little confused. I'd liken Chrome to an Ipod. It lacks features some users might demand, but it works right out of the box for most people. Microsoft IE8 is, well....Microsoft. It feels more application centric.

Perhaps that says something about the web strategy of the respective companies. Google wants to pull users out of their existing habits, and into the Google web, whilst Microsoft needs to integrate existing application users with the web.

A subtle difference, but there nonetheless.

Have Your Say

What are your thoughts? Have you tried both new browsers?

Webmasters are often faced with the problem of how to approach SEO on websites which have a country-specific focus. As you may have noticed, the search engine results pages on Google's geo-targeted search services frequently display different rankings than those you experience on Google.com. 

If you run a few queries on, say, Google.com.au, you'll soon notice distinct regionalization patterns. In order to make search results more relevant to local audiences, Google uses different sorting methodologies than those used on Google.com.

Here is a guide to optimizing sites for the different regional flavors of Google.

Country Specific Local SEO Tips

1. Get a local domain extension:  Google places a lot of weight on the domain name, so it is important to get the appropriate country-code domain extension. If you compare results across the different geo-targeted flavors of Google, you'll notice the weight given to the local TLDs. There are exceptions, but the local TLD tends to trump .com when it comes to local result sets. Different countries have different registration criteria for domain resitration. It is fairly easy to register a co.uk or a .co.nz, whilst a .com.au can involve setting up a business entity in Australia. 
2. Specify your country association in Google Webmaster Tools: Google Webmaster Tools offers a facility whereby you can specify a country association for your content. You can do this on a domain, sub-domain and directory level. More detailed instructions can be found on Google's Webmaster Tools Blog.
3. Include local contact information: Specify a local address, business name, and local contact phone numbers. Whilst not critical in terms of ranking, every little bit helps, and by including local information, the site becomes more credible to a local audience. 
4. Local hosting: Depending on who you ask, you'll get different answers as to whether the geographic location of the web host makes a difference in terms of ranking. I have .com.au, .co.nz, and .co.uk sites, hosted on US servers, and they rank well on the appropriate local versions of Google. Other people feel that location-based hosting is a must. Still others say the location of the name server is most important! It's fair to say that if you have a choice between hosting locally and hosting offshore, then it might pay to host locally. It certainly can't hurt, and there might be additional benefits, such as increased download speed. If you go this route, one thing to check is the servers physical location. Often, web hosts have a local office, but their servers are located in a different country. Use an IP lookup tool to determine the exact location of a server. 
5. Spelling & Language: Ensure you use the appropriate spelling for your chosen region. There is a difference between "optimization" and "optimisation". Keep in mind that searchers will use the local vernacular. For example, if you are optimizing a travel site in the US, you might use the term "vacation". However, searchers in Australia, the UK and New Zealand, amongst others, tend to use the term "holiday". 
6. Tone: Copy that works well in one geographic location may not work in another.  For example, the sales language used in the US is usually more direct than that typically used in the UK, Australia or New Zealand. Familiarize yourself with local approaches to marketing, or engage local copywriters.     
7. Inbound links: Seek out local links. All links are good, but inbound links from local TLDs are even better. Approach your local chamber of commerce, friends, suppliers, government agencies, business partners, and local industry groups and ask them for links.
8. Local directories: Get your site listed in local directories. Local directories still feature well in geo-targeted search results as the depth of content, in terms of sheer volume, isn't as great in the local TLD space as that published on .com. Obviously, you stand to gain from the local traffic that the directories send your way, and any local link juice the directory may pass on.  Here are some top local directories:

• The local Yellow Pages i.e. Yellow Pages Australia, Yellow Pages New Zealand, and Yell (UK). Keep in mind that some of these directories may not pass link juice, however you can weigh this factor against their value in terms of local reach. You could also seek listings in the regional sections of the following global directories: DMOZ, Yahoo, and BestOfTheWeb.

Recommended regional directories:

• Scoot.co.uk is a prominent UK business directory.
• Webwombat.com.au is a comprehensive Australian directory.
• Te Puna is a government run New Zealand directory.

Have Your Say

If you have some additional ideas that have worked well for you, please feel free to add them to the comments.

When I first started blogging I tried to learn from and emulate 3 of my favorite bloggers: Seth Godin, Peter Da Vanzo, and Steven Berlin Johnson. A large part of the success of this site was learning from those guys. Recently I was lucky enough to hire on Peter Da Vanzo to help do some of the writing on this site. He has been blogging about search since 2002 on Search Engine Blog, which officially makes old school.

Please give Peter a warm welcome to the site!