Search, The Internal Frontier

I’ve recently taken on a project which includes a complicated “Internal” site search.  This website not only needs to be found on the web, but members can run specific searches against specific elements on the website, click on subcategories and then create filters.  Sounds a bit like an eCommerce search, right?  Well, it’s not, but eCommerce is a good starting point.  This blog article is going to explore what I know about internal website search and places they have solved it well or not so well.

First off, there are two kinds of searches online as you should know by now.  You got your World Wide Web search now referred to as a Google search, and an internal website search.  It may not be so easy to figure out which kind you are doing these days, because an internal search quite often is using Google and/or the Google Appliance, a box Google sells you that manages a “Google like” private search engine for your website.

I first ran into a hybrid search, what I call “Directed Search” when I was hired for a consulting gig at The Limited, VictoriasSecret.com.  They had no search box at the time.  The year was 2005.  I mention that, because that shows how large, successful brands, can get stuck in the past.  They required the customers to click on the word search, then click on something like a bra. Then you would have to choose your size and finally choose a style and color.  It was like 9 clicks to checkout. Well that was unacceptable.  But it made sense from a technical standpoint.  A lot of things make sense from a technical point of view.  Having a system that holds people’s hands through a step by step search can make sense, but not with bras.  At VictoriasSecret.com, later that year after I had left, the search became a search box, and finally you could put the word “Pink” in that box and get some results and then make adjustments on the left for price, color, size, etc.  Point is, that not all searches are equal.

I have been studying two well used online internal search paradigms to solve my problem, Linkedin.com and Bankrate.com.  Both websites have a complicated, MIT level issue with data, and getting customers to the right answer, without taking up too much time and without losing them in “advanced” areas.  I think both sites live and die by helping the members find their way to a person/company or whatever int he case of Linkedin.com and to a mortgage company in the case of Bankrate.com.

Basic Versus Advanced

I think we can all agree that the best searches have a limited amount of functionality at first.  If you read my article on “2 Online Marketing Segments: Insecure Focused, Secure Unfocused“, you will see that you need to accommodate the less sophisticated, yet ready to go, user like myself first and foremost.  I admit I don’t read everything, and I would guess that 80% of the population is like me.  But for that 20%, there are extreme details they want to know, and they WILL use the advanced search. So, you got to keep everybody happy.  It is not easy for Linkedin.com for instance.  You will notice they have it easy to find people, but if you start peeling back the layers, the advanced search is there.

One Click Versus Many

One UI goal and conversion goal for all searches is to reduce the clicks.  This means if something can be done in one click that’s great and lets say no clicks, that’s even better.  Remember with JQuery you can have a no click solution, where people type things in, look at Answers.com and others, and have it change as you enter your search terms.  The way that Linkedin.com solves this is by having a little drop down for the search, which means you can choose what kind of search you are starting before you enter data, people, company, group, etc.  This is really innovative and useful.

Complications & Third Parties

So sometimes a search can not be as straight forward.  There could be third parties who have ad-space making it advantageous for search to take a little bit longer (a few clicks).  Bankrate.com is a great example of searches where, in order to search customers correctly, they have to know where you are looking to get a mortgage and what type  you are looking for to even start the search.  For instance I would be looking for a 30 Year Fixed in Boca Raton, Florida.  Then they can start the search… So for some searches it is more about getting the basics in place, and then start the filtering.

No matter what kind of search you set up.  Trust me the Google Appliance is not for everybody and there are still about 15 companies out there which provide third party internal Saas search products like Omniture and Autonomy.  If you are an eCommerce company you can’t live without them.

 

 

Web Feature Discovery Process – Part 1

Not that I am any bit an expert on web features, product management… and not that I know something that anybody else couldn’t figure out, but I am somewhat obsessed with the website “feature” development process, especially when it comes to overlooked, under-estimated, misused assets. This blog entry is about how I would go about discovering features and services and solutions towards increasing website traffic as well as finding new profitable directions and increasing conversion rates. I am going to give a few examples, mainly from my experience working on a site like Whois.net (which is far from maximized; as far as I know they have just started the process). But first, before I describe the process to develop these features and new products (aka, the old product management process in a new era), first we have to understand the lexicon of the web feature discovery process

Assets

Assets are virtual and sometimes physical things that an organization or company owns and manages. Examples are domain names, websites, email addresses, segmented email addresses, unique site visitors, pageviews, sms, twitter accounts, Facebook and LinkedIn, Persona of the visitors (how can we break up the kinds of visitors based on background), programs and applications, patents, copyrights and internal resources like people. Oh yes, even people are assets. These assets have a value and if you don’t put a value on these assets, you are making a mistake.

You need to start the web feature discovery process by working with asset value and not revenue, because while most everything in marketing is revenue based, the overall value of what you are building towards (ultimately revenue) may be determined by the value of the assets, and assets may ultimately determine long and even short term revenue. When I say assets, many technology and marketing execs would often give me a blank stare, like they have no clue what I am talking about. (I can only chalk this up to the fact that asset value is not how they are compensated, so don’t blame them, blame the boardroom for not keeping up with current valuation measures!) But what is going on in the start-up world, whatwe can call the modern world of business, is a whole new world of asset valuation. And if your division created a website worth $20 million that only made a net income of $100,000 a year, maybe it is time to sell it and take the profits…

The second thing to know about these virtual assets are the more detail you know, and the more they are optimized, the more valuable the assets are worth. We are in a world where websites, domain names and other assets are sold off to make some cash. So, don’t overlook asset creation vs. revenue creation. They now go hand in hand. 20 years ago you would think I’m a loony bird, but the world has changed and companies do sell stuff. At my last firm they did sell assets, except not after we maximized them, but after they personally devalued them.

Often a firm may be collecting email addresses. I say “may”, because some out of the way places don’t. Good luck to those outfits. If most organizations just knew 10 things about these email addresses, what people were interested in, or simple things like their first and last names or where they are located, the asset value of the list may be double or triple the value. Add more detail detail like age, demographics, what they like, who they like, etc and increase the value further.

How do you value these assets? My way of valuing them is simple. If you were to take them away and wanted to get them back, how much would it cost you? For instance, and I am often going to use Whois.net as an example, because that was the last large site I worked on. The Whois domain search site got about 2 million unique visits a month. Now the company saw no real value there because they had a hard time understanding the relationship to their sales, but when I asked execs how much would it cost you to buy 2 million visits a month by buying the PPC (Pay Per Click) words “Domain Name” and “Hosting”, it would have cost them a minimum $2 million dollars a month budget and therefor, just the traffic was worth $12 million a year or $36 million over 3 years. It’s a bit of fuzzy math, but going with “remove it and try to reacquire it”, is a great way to get them to understand. The reason most don’t understand is they typically don’t sell assets and are graded on revenue…but is that really what this is all about in the end… Because if you make something worth millions maybe the asset sales is bigger than any revenue you would ever be able to generate.

So start the process by doing an inventory of assets. First day on the job and you want to make a new web features happen at your web company, start by finding out the basics. What domains do we own? What websites do we run? What are total number of email addresses? How many visitors to each site? How many segments are we catching from customers in the email area? How many members? How many orders? How many skus? How Many? What? Where? Why and How? Get this information down on paper, because this information is the foundation for new and improved services.

Champion

Leaders, execs, people who have an idea, guys in the company basement, people on the customer service lines, MBAs with a business plan or just a lonely CEO with an idea are champions of web features. This means that somebody has to believe in it and want to make it happen. A group of people may want the features to happen, but a person has to ultimately answer and stand up and say I am the product champion. Groups don’t champion stuff, individuals do. This is one of the critical mistakes made by corporations, to think that a group of people will decide by committee ultimately what a web feature will do and how it goes is a big mistake. Not that web features are made by a dictator, and if the champion is a dictator, he probably will fail. If he is a benevolent dictator and listens well, it will succeed. Funny thing is this champion can not be limited to execs with great salaries and titles like VP, Director, CEO, CIO, CTO, BFD or Founder. A champion can and should be everybody. That is what makes companies succeed, not a special group who say only we do the thinking. It can and should by anyone, including employees, shareholders, customers, husbands & wives, sons & daughters, friends and the UPS man. Champions need support and guidance and promoters from above, below and sideways. Being a champion has its risks, as I’ve learned and you can get burned by being the champion or you can get the accolades and make it happen. You can even make it happen and not get the accolades, but then you would have known inside you made it happen. So it does not matter! Money comes later, first comes action!

Little Trees

Years ago we used to pay to plant trees in Israel. And then years later when I visited Israel I got a chance to see those trees grow. In order to grow a tree, a big tree, you have to plant seedlings or small trees. This is where many execs lose their patience and understanding of the product management process. You have to test, test and test again these little trees in order to find a big one. Ok, if you don’t get the tree allegory of ideas, you may be missing the point. Where do you find these little ideas? Somebody recently said, “Dan’s An Idea Guy!” That is not true. I am not an idea guy, I am a guy who listens and hears other people’s concepts and evolves them into ideas. There are ideas all around us, if we would take the time to just listen and sort through the data. Remember, the assets… Just doing an inventory will start to flesh out these little trees or concepts or ideas. One thing I always did at these companies is walk around and chat with the various people in the business. They have ideas. They know what may or may not work and though they don’t know how to implement, they do know something they are not telling everybody. Often its something in the business that bugged them for years that they want to share.

Sharing

Never thought the web feature discovery process would come down to sharing, but learning to share, something my 3 year old has not yet mastered, is the key way towards finding those little trees. People need to get together and chat and think and find answers. These answers are something somebody read somewhere. A company environment where people don’t share their thoughts is a place that will never flesh out new concepts or web features. Look at Google, they are actually asking for the ideas and look what they have produced. If we want the rest of corporate America to be successful on the web, they better listen up and start to share. Like I said earlier, it is the champion that takes an idea and makes sure it happens, not the product management guy or gal. The product management person should be the facilitator and not the creator in the end. Listen and learn, not ignore and complain. Sometimes doing your job requires less not more of giving and taking. Learn from your childhood and share. The secret to sharing is giving of oneself. If you can not give to others, by give I mean tell people something about yourself or your ideas, you will not be able to acquire ideas. The sharing has to start with you.

The Other Guys

Now this is the easiest way to find those things that people have not thought of and get the real brainstorm on what is happening in the market. You actually need to go off and look carefully at the competitors. This is not about mimicking people, this is about concept development. You see a feature on another competitor and you grab it. Fine, but you not only have to take it, you have to own it and therefor it needs to come from you in a new way. What I ended up doing for my Whois.net thesis, which you will read about in my other blog articles on Domain and Whois Tool searches. This meant locking myself up and reading through 500 websites. From this exercise alone I ended up with a dozen new products and features for the Whois.net site. No rocket science involved here. It is simply looking and learning.

Integrative Strategic Thinking (Aha Moments)

Once you have all this data in front of you, such as the assets, the new ideas from those around you, and the competitive information, you now start to see things from a different level. At this higher level, as you look holistically at information. You can start to piece together stuff you did not see originally in just the assets. For instance, when I looked through and discovered that Whois.net did not allow international domain name look-ups, I knew immediately this was an important issue. The importance was simple enough. If you increase upon (extend) what people already like, you will probably have success. They used to call it product extensions. This is where you take a product that is already successful and you add on a new feature or extend the product to new areas. Not a high risk activity. For Whois.net, international domain name traffic look-ups blew out the traffic, automatically doubling it in six months and it tripled and quadrupled traffic over a year. Just satisfying people with stuff they already wanted is easy. However, what is easy to do, is often not seen by the blind. And when you are busy in a high end corporate product management job, you are blinded by the requests from above and from the sides.

Stats and Prioritization

What did I do with the 500 websites I viewed in my these on Whois searches? I came up with a scientific approach to figure out what feature was important. Most of you who read this probably prioritize every day. I rated each feature by value (yes a monetary value), ease of implementation, where I found it, as well as the monetary value of the websites I reviewed. This review process was not about money, it was about assets again. What I focused on, was how do I get visitors to this site, not on how do I convert. I was leaving conversion and selling at this point up to marketing and sales. That was something they knew how to do pretty well. What you need to as a good champion is to understand the data beneath the hood and how to use this data to make a point. Prioritization and hitting low hanging fruit are extremely important ways of working as well. We are in an impatient world, where execs don’t have the time or energy to listen unless they are just seeing the cream of the crop. Maybe they shouldn’t know everything till the time is right. Sometimes companies kill a product or project the first time around because it failed. That does not mean the second time it will be the same.

Learn From Your Mistakes

Organizations that learn from their mistakes and take actions the second time around to make things right are rare. Most organizations bury a concept that has failed and when it is mentioned again by an newbie, the newbie is crushed with the notion that “this has failed here”. This is a big mistake. Failure should never be viewed as a doorway an organization can not go through again. It should, however, be the shining example of how not to do the same thing, the same way. Failure should be used as a way to understand what to do right next time. Like a pyramid, building upon their knowledge, great organizations store this learned mistake information and use it positively going forward.

Leapfrog

One of the concepts I learned while working on a well ignored site like Whois.net was if you are so far behind the competition, it is sometimes worth it, to not mimic, but rather take a leap of faith and go for something greater, different, in a way that competitors would never do. Why won’t they? The competitor has already made their product or web feature decision and taken the current path. If they are leaders in the market already, it will take a lot for them to change. This leap of faith may be something like give it away for free, or combine it with something new or offer something completely different or in a way that is easy to identify but not the same. Simply cookie cutter mimicking is a nuisance on the web. Who wants to go to Bing, when Google does it so well? Why would I ever do that, other than Microsoft has figured out a way to trap me when I load the next IE Browser? Now if Bing did something so different, so incredibly well done, it would make sense. If they were better on an Ipad, sure. If they were better with voice search, sure. If they were bettter or different…it would matter. Making it matter means being different not the same. In the case of Whois.net, I was determined to make the site a competitor with Sedo, the domain auction house, except my idea was to make it a free place to buy and sell domains. Sedo is not free. This is the kind of leap that makes a difference.

Hopefully this blog entry got you interested in discovering a new web feature… I will continue this blog entry in Part 2. Click Here to go to Web Feature Discovery Process – Part 2! And it turns out there will be a part 3, which I am currently working on.

SEO in Overdrive :)

Like may things in the web everybody is looking for the edge. You can go and read the JimBoykin.com site and get some interactive agency to do an audit, but often it is the same SEO stuff over and over.

A couple of years ago, at the beginning of SEO, Search Engine Optimization, around 2001, I was tasked with getting this startup speed dating site as high up in the search engines as possible, because as you can imagine, it is always less expensive to be found naturally, than to be found through PPC, Pay Per Click. Lots of stuff out there and lots of the same stuff where ever you look, so much so, for interactive firms these have become a standard.

Surprisingly, like 90% of sites have yet to meet these standards. These are pretty much the same as when we started SEO around the early 2000’s, plus all the new media stuff like blogging, plus now webmaster tools from Google, Yahoo and Bing like sitemaps.xml files.

There there are some additional small details that you would not think of to round this off. This is what I call SEO in Overdrive. Recently I gave a short talk on what are critical technical issues in getting a site nicely found on Google. So here is my SEO in Overdrive list below. I will probably miss a few things here and there.

1. Choose either www or no www URL.
Choose either www.strategicpoints.com or strategicpoints.com. Having both URLs working on your site will cause Google to count your rankings for pagerank separately. Together, as one domain, this will increase your site pagerank, because these two URL’s pagerank power is diluted. How to do this? If you have a linux server, the .htaccess file can be used. I am pointing to a good refernce to figure this out. If you are using WordPress, then do it for you automagically, and will switch you over to either one or the other based on the choice. If you are using windows…well I am sure there is an answer out there.

2. Localize or specialize each page of a site.
The old days of one title for all 100 pages of a site are like 1999. The top areas are title and Meta Description. As I have found out recently, if you mess up your title, the listing on Google will reflect this. So make sure the title is accurate and contains the key words you want to be found by, first… The Meta Description (if you don’t have top content) will be used by Google to show the descriptive text underneath the title on a Google result page. Take for instance a situation recently where we migrated a site, and the Meta Description was removed from a page. Without the little description beneath the title on the Google Results page, the link looked like an ad, especially since it was showing up first in the rankings. We learn from our dumb mistakes. We put the Meta Description back, and wholla, the link reappeared and so did our traffic.

3. Healthly, always changing content…
The search engines love content, no matter where it exists on your site. This is why having a forum, blog or other interactive parts of your site are so critical. This content is always changing, and this is always freshly indexed on the search engines. The introduction of the blog, and WordPress, which I favor, has really given us an ability to SEO in Overdrive, because it pings out to the rpc servers that update the search engines within 24 hours. This means not waiting for crawlers to come out to your site, but telling the search engines when and where to look for new content.

4. Text Links…
Who knew they would be so important. For sites that do not provide text links on the home page, they are missing out on the ability for both crawlers to go deeper, and the ability to make these links important part of search criteria for your site. Ever notice a group of links below your search results on Google… They determine these for you initially. These are considered the top internal links from your homepage on your site. The calculate it for you if you don’t do it yourself as sitelinks, through Google Webmaster Tools. You can remove these, but you can’t add these I believe… Only having images links is a downer…, and can be a negative impact on your sites. Top sites that want to continue with image links often put <div> based text that is not visible behind the image, that state the link in real text as a way to solve this issue. Either way, always take your links seriously.

5. Images, Site Elements and alt tags.
One of the tricks we used years ago to get to the top of the search engines without violating any code of ethics with Google was a technique I call snow balling. We decided on our top search content words, which were “speed dating”. Then we renamed the /images/ directory to /speed-dating-images/. Then we proceeded to add alt tags to each image, that said the image name with “speed dating” after the name, like “party photo – speed dating”. Then we went on to include alt tags in the links with the name of the link “event xyz – speed dating”.

6. Directories and Virtual Directories.
A newer technique is to create directories or virtual directories with tons of additional terms. No need to create these real directories anymore, unless you are on a windows server. I only know the Linux apache version of .htaccess and this can be used to create virtual directories for SEO. In our case it would have been /speed-dating/, /speed-dating-parties/, /speed-daters/, etc., etc. These so called directory and sub directory structures are really the cutting edge of the web today in terms of getting found on the search engines. No need to work hard and create these directories and subdirectores anymore or even a super .htaccess file. This is why I am lovin WordPress. If you have WordPress, all this is done for you and no need to work about it. WordPress not only can create these for you with real directories, but the redirection program I mention in my enterprise wordpress discussions on earlier blog entries gets into how it works, and why to use it.

That’s it for now. I am going to add some more to this list soon, and as you know there are a hundred other blogs out there with a lot of the same information. Good luck putting your SEO in overdrive. Contact me if you have any questions about the stuff I mention today at dgudema AT StrategicPoints DOT com.