So far in this series of articles, I have described dealing and identifying Assets and dealing with People, the two ingredients needed to build any web site feature.  The next step is information discovery.  The first article talked about Assets.  Assets, as I describe in the first article, need to be understood and evaluated in an open and free environment, not restricted by management that has already indicated the final outcome.  Knowing the exact, final website product on day one is ok as a goal, but to know the exact way it will work is a not just a mistake, it can create a terrible work environment for those who see the mistakes and can’t fix them.  You start with a general idea of where you want to go in development of web features, and great spec’s can make this possible, but there is a transformation that needs to occur between concept and final product, and that is where a lot of websites go awry by simply never fixing what needs to be fixed.

Information Discovery

Information discovery is all about looking through the data, which leads to idea discovery.   But even before you go through reams of data, you need to figure out what you are trying to accomplish.  This is a bit of a conundrum, like which came first the chicken or the egg.  Let’s say you have a web page or a web site that is already doing its job.  Your site could be a simple page or 2 or be as large as a 20,000 SKU online catalog.  I’ve built both sites in the past.  The question is how do you figure out that you need to add a form to collect information or to place ads on the site?  How do you know if the site should be a marketplace or a straight e-commerce site.  This is not just about data here.  This is about business and business models.   If you have worked on developing websites, and at this point in time, and a few million people have, based on the fact that millions own domains, you probably know a thing or 2 about making your own website.

In my first post on this subject, I described assets, where you have to look at few key pieces of information.  Is there any anomalies showing a possible opportunity, or as Google in their web analytics product calls it, Intelligence?  Is this site getting tremendous numbers of visitors?  Is this site already collecting email addresses?  One site I know of, because I am a part owner, gets an email address added to the system every 10 minutes during the day.  This asset is important, because email addresses can be, should be used properly, by properly, I mean when people give you an email address, they are expecting an immediate response.  The value of the email starts to go down when you wait 6 months till you email them.   The fresher, the more valuable, and that’s because people have a small period of time to read that email you will send them.  This has lead to the concept of email sequences.  First time I had heard about this was from my business partner, who pointed out a company called InfusionSoft, that is big on it.  It’s really a simple concept.  You have a series of emails that get sent out to a person that signs up that go out in phases, like one a day, one a week, perhaps growing in length of time.  Each email has a different message that is part of an overall strategy.  You don’t need InfusionSoft to do this, but they are great at it.

Let The Wind Tell You Where To Go

Often data is telling you something.  For instance, on Whois.net, we noticed a high amount of international visitors, yet we did not offer the ability to look up international domains.  So we solved the customer’s quandry by offering what they were looking for.  So one clue is looking at the key words people search on to get to your site in Google Analytics or Adobe’s/Omniture’s Site Catalyst.  This is a simple task and 90% of online marketers know this.  What they don’t look for is the missing link.  The missing piece is what ties information to a new potential set of features.   Another good example is on checkout of most registration sites, there is a term we now call “Co-registration”.  Co-registration means the customer was here to signup for x, and you added another potential thing for them to get at the same time.  We have considered using this in the event business website I am a partner in.  Another interesting anomaly I noticed in the dating business recently, is that people are using Iphones and Android apps to signup, in significant numbers.  This is where you have to brain-storm, not about features, but about assumptions.  You can confirm these assumptions, through research.  My assumption that people are using smartphones, specifically at work during the day, because it offers more privacy, and the employer can’t track you specifically.  This is one of many reasons, but the end result is we need to have an app for the Iphone and Android…  That is simple detective work.  This is the big breakthrough.  It is finding a new channel.  The big question is having the resources to capitalize on this new channel.  It may not be a new channel to you, but to a lot of executives out there, who don’t know how to deal with this channel, it is a strange new world.

In the retail side its called Cross-Sell or Up-Sell.   Just sign-up for Godaddy and you will get your complete lesson on up-sell and cross-sell.  They are the masters at this.  I used to have people say that is not what I would put on the checkout, because it does lead to Friction, one of the key points in the MarketingExperiments.com formula, where it will actually causes less conversion potentially.   This all depends on the site.  On a dating site, yes, it can slow the process down, especially if the person is not ready to convert.  On a retail site, however, when people have made up their minds to buy, only a broken, poorly designed web page can stop them, especially if the deal is an amazingly good value.  So most features are figured out as extensions of what you already have in place.  If you build on your success, you will succeed even more.  There is no need to be radical, as you will find out the hard way…

Slash And Burn

If you let the status quo dictate what you do, or worse, let the current sales and marketing team make all the decisions, you could end up with a situation I call “slash and burn”.  I came across the concept of slash and burn, while working on a few sites.  The analogy of slash burn comes from the military tactic of burning the crops as you retreat, so your enemy will not have the luxury of food.  It refers to the, sometimes unforeseen, consequences of making a decision to kill one part of a website in order to enhance another.  This is a human decision process, typically driven by revenues.  The best example was on this Whois.net site, where, when I arrived at this company, the previous people who managed it, had attempted to drive all the traffic through links on the site to hosting sales on another site, because that is where the money was for them.   This was a major mistake in my mind.  Yes, they had driven people to where they make money, but that is not why they came to Whois.net.  They were there to look up domains that were available, find out who owns them, and potential buy them.

Consistency

Another very important point I learned from MarketingExperiments.com, and I will discuss them a lot here, is that consistency is real important in the process of building out web features.  You have to start at the beginning of the customer path.  The beginning is when they are sitting at their computer on Ipad.  Let’s say they have not even turned it on.  They are interested in South Florida Real Estate, as an example.  They open the computer and type those words into Google.  Google presents them with relevant results.  Let’s say your site is in those results.  When you click on one of the links, and let’s say its one of the top paid links, you are delivered to a web page.  The words you put in should arrive right in your face at the top of the page in big bold letters.  If not, you are not getting a consistent experience.  This is true for many of the paid links, because they are using something called Long Tail, which has a few meanings.  To me, it  means extending the search words they are buying out to more obscure words, to pay less and get more traffic.  This would typically mean they are looking to buy “Real Estate”, but it was cheaper to buy “South Florida Real Estate”.  This is great for smaller sites and pages that are meant to be for finding this exact stuff, but when you arrive at a generic Real Estate Seminar, you are disappointed.  This is not Google’s fault, it’s about people trying to get traffic, and the result is a lot of inconsistency.  Your consistency is critical in making your site found well and sticky, a term these marketing guys love to use (meaning they stick around).

Don’t Make Me Think

If you have not read this book, “Don’t Make Me Think”  and you work in the web field, you should get it, read it, and live by it.  The book’s simplicity, and I will hopefully sum it up here, is that people have developed common methods of experience online.  This means that they are expecting the same words, in the same spot, each time they arrive at a web page.  A good example is the word “About” or the word “Search”.  The book basically says if you say “Quick Search”, you are going to confuse people, and therefor making them think, like is really “Quick” or different.  More importantly, if your site is missing any of these common elements, such as the words “About”, “Contact” or “Search”, there is a customer disconnect.  And you need to understand that “Search” is a feature.  So for the basics, you need to make sure the customer experience is not so different they run away.  That also means that placing the word “Search” in the upper right of the page is preferential to let’s say the bottom left.  From the web features perspective, deliver to customers at least the minimum they are expecting on the site.  More and more, customers are expecting a web form on the home page and sometimes many site pages, where they can put their email address in, maybe with their first name or some additional data and get on that companies’ mailing list.

Short Form or The Long Form

One of questions is whether or not to put a short form or a long form onto a home page.  So, inevitably with these kinds of web forms, information collection forms, you have techniques that are learned over time.  You can learn this stuff by observing and taking the best of breed (they call it) and do things like this.  One of the methods that is recommended is you ask for a small amount of information up front, like just an email address or email address with a first name.  On the second page, you would then ask for additional information, saying that they are now on the list and they can further tell you more information about themselves.  This two step process like everything, reduces friction on the first step and allows the website visitor to make intelligent choices on the second step.  One thing I ran into when working for a hosting company was a situation where they required a domain name when buying hosting.  This was a system requirement.  It turned out to be a very costly financial requirement.  The reason is, and like everything, it came down to how people react.  The visitor, in many cases, had not decided on a domain name, so often they would just pick up and leave to figure it out…

This Web Feature discovery process article is 3 in a series of 3 so far.  The articles about Web Feature Discovery will continue in a new article next month.

Today I noticed this new Google Analytics feature in Beta called In-Page analytics.  Very simply it is the ability to pull up a specific web page and visually see all the clicks, sales, conversion rates and some other stuff by every link on that web page.  They have had this around in the past, but not at this quality of usability for zero dollars.

This is not a new feature, in fact, its just Google Analytics catching up with all the enterprise analytics companies in the market such as Omniture/Adobe, Coremetrics/IBM and Webtrends.  Is this good news?  Great news for users of analytics, because it makes life a lot easier to analyze stuff visually.  Well its not good news for the old guard of web analytics as Google chips away at features that used to cost thousands per month.   This is basically another step towards devaluation of analytics tools.  Not sure how many companies are out there still doing analytics.  I remember at a conference when Mr. Stern declared that there were 83 web analytics companies at the time around 2003.  And I have seen new ones join the fray in the last few years while old ones like Hitbox have become a thing of the past, being gobbled up with Omniture, which was feasted on by Adobe.

Why Is This In-Page Analytics Different?

Well this in page analytics is not actually as you surf your website.  That is a major difference with competitors and how they run.  Typically they almost all worked like this.  You turned on some application that is local or browser based, so either you download something or click on a new add on per a browser and then surf the site.  This required a lot of effort to surf around and honestly the in page analytics I used by Omniture/Adobe and Coremetrics/IBM hardly worked.  In fact even their staff steered us away from it.  This new Google In-Page Analytics has really created a great feature, where it grabs the page from the website and it all seems to work… Nice.

How To Use In-Page Analytics

Quite simple to use.  Just login to your google analytics account and click on CONTENT in the left side bar and you will see a red BETA next to In-Page Analytics.  Once you click on In-Page Analytics you will notice that your home page shows up.  Notice that links have a little div/verbiage above them with a stat like 1% or 50%.  That stat and others are showing you what percentage have clicked through, purchased, and how much they have purchased.

Hitting It Out Of The Park

So this proverbial Holy Grail of web analytics seems to do all and suffer none when changing to use revenue, goals, or any other analytics.   Yes, the bigger enterprise solutions do have more bells and whistles and flexibility when it comes to in-page analytics, but you have to admit this is pretty much an amazing tool.  You could quickly look over your home page or any page and make a determination visually what link is working well and what does not work well.  This turns web analytics into a real science and brings this cool method to the masses.

The fact they are free is the amazing part, and to the chagrin of several hundred developers at competitors like WebTrends, Omniture (now Adobe) and Coremetrics it is just one more dink in the armor, especially when these types of enterprise features are considered something you would typically pay for.

Ok, so what it is… Well actually if you don’t have much traffic you won’t notice much when you click it.  Since I work for a large corporation and have quite a few small, medium and large Google Analytics accounts for my sites and other clients, I can see the disparity in what you get.  For bigger corporate clients there is data showing up.

2 kinds of data.

One is called  “Automatic Alerts”, the other are “Customer Alerts”.

Just assume the automatic alerts they decide for you, while you have some control on your own.  Weird thing is the actual types of control are very general with a simple low to high sensitivity bar.  Without a lot of information about this tool, I am going to take a stab at the relevance and how to use it…

Automatic Alerts (Variance Reporting)

Like I said, if you have some data coming through that is a bit more than just 10 visits a day, you will see something.  What I am seeing is any form of data that is above a specific variance curve showing up.  If you have had basic statistics.  I suffered through it twice in two MBA programs, you have the basic within like 2.5% of the mean at both sides of the bell curve.  When I see low medium and high settings, this probably is more like the variance of like 20%, 50% and 70%.   You can modify it a bit with that bar.

What does this really mean for analysis?

It pretty much it means they are “Alerting” you to unusual activity in your account.  For my corporate accounts I am seeing things out of the ordinary showing up, like traffic from a specific geographic location has increased by greater than 40%  or time on site has increased from a specific location by over 88%.  You will have to become a hack like me to understand the fact that this stuff is relative.  This means that it really can mean very little depending on the situation.  You have to dig deeper to see why or how it is a variance.

Time Frame

Trying to get a handle on what time frame means, it appears that if you choose a small time frame at the top of the page, you create a statistically insignificant amount of data to analyze, and you will notice that there is no results.  There are two types of time frames involved in this analysis.  There is the time of the report you are looking at, at the top of the page, and there is the “Time Frame”, one day, that is in the middle of the page.  You can shift this “Time Frame” back and forward (if you are not looking at today).  This way you can see if the trend is common each day or just for this day.  Does not look like right now that you can switch this from a day to a week or month, but that would be a nice feature. Maybe I am missing something on the screen to switch this.  Also, seems like this is going to cost us money at some point, so maybe that is the paid feature?  But then, I have not yet gotten to custom intelligence.

Alert Sensitivity Bar & Significance Bar

There is a significance bar on the right side of each metric and dimension.  This significance bar is related to the variance discussed earlier.  It can be used as a measure of how important, or in my old stat class we called this the Sigma.  I actually turned down the Alert Sensitivity bar in the middle right and noticed that a few of the alerts dropped off.  This sounds right, if it is straightforward.  Obviously you will have less alerts if the Alert Sensitivity bar is lowered.

Graphing Anomalies (Alert Data)

Another part of this is the cool way you can quickly access graphing.  Let’s say you are looking at a specific alert.  There is a little graph icon on the left of each alert.  Click on it, and you can that specific target data across the page time line, and BAM a pretty powerful way to use this thing…  Try it and see what I mean.

Alerts Candlestick Bar Graph

Notice this secondary graph beneath the normal graph. This one is tracking how many alerts per day.  If you are just starting out this is set for the automatic alerts only.  If you set up custom alerts, it will track them on this chart as well.  I am going to set up a few and then report back through this blog and my twitter account @dgudema on how they work.

Group By Metric | Dimension Function

There are 2 ways to Group the Information, by Metric & Dimension.  Metric refers to the basic metrics listed on the right side.  Dimension appears to be the more deeper secondary level metrics like region, new vs repeat visitor and other more detailed items.   Interestingly enough the higher level metric may not show the alert that is in the lower level metric.

I am sure that there is more here that I have missed.  I will cover it in my second article coming up soon.

First off, Event Tracking is a new feature in Google Analytics.  If you are already a user, it is a different thing you have to implement with tagging.  So if you are not an existing implementer of Google Analytics, you will have to start from the beginning.

I have been working with Web Analytics for over a decade, and there isn’t an analytics program on the market in the top 3 or 4 that I have not worked with, from Webtrends, Coremetrics, Omniture (Now Adobe), to Google Analytics. 

So recently when I was asked to write a short spec on the Event Tracking tag and how it works in Google Analytics, I had to look carefully at this tag.  It could be confusing, because Google Analytics typically does not use a lot of tagging.  You just pop it in there and it works!   Good news though, is that 99% of Google Analytics users won’t need another tag, unless they want to dig deeper (or have to).

So, as a reminder, the basic chunk of Google Analytics code looks like this:

<script type=”text/javascript”>
var gaJsHost = ((”https:” == document.location.protocol) ? “https://ssl.” : “http://www.”);
document.write(unescape(”%3Cscript src=’” + gaJsHost + “google-analytics.com/ga.js’ type=’text/javascript’%3E%3C/script%3E”));
</script>
<script type=”text/javascript”>
try {
var pageTracker = _gat._getTracker(”UA-XXXXXXX-X”);
pageTracker._trackPageview();
} catch(err) {}</script>

You get this when you login to Google Analytics, go to edit your account and click on “check status”.  If you need more info on how to do this, I am going to write an in depth overview of google analytics and will cover the basics in that article.  

Basically replace the UA-XXXXXXX-X with your code, or grab the full javascript code from www.google.com/analytics/ , pop this code into your html, and you are ready to go with Google Analytics.

Extra Tags & Event Tracking

There appears to be about 5 or 6 extra tags for Google Analytics (depending on what article you view on the web), and finally 2 extra tags for the checkout process.  By tag, I am referring to lines in the analytic javascript code block.  The big one that seems to be getting lots of attention recently is Event Tracking by Google Analytics. This is because it is new in their program and answers a lot of problem areas that Google Analytics did not cover in the past. 

The basics of Event Tracking are you add a line of code per event.  This would mean adding the following line:

pageTracker._trackEvent(category, action, optional_label, optional_value)

to the google analytics javascript block.  I would place it right above
pageTracker._trackPageview();

You would replace the variables in _trackEvent in order to produce specific tagging of Events. 

Category

The Category is used as the highest level tracking.  This means it ties together all the other variables and should not be too unique per line.  It is required.  This means if you are tracking a series of events,  you would keep this the same per event page or action.  If you are tracking a WebForm for instance, then keep all of these the same wording like “Web Form” so they are grouped together.

Action

The Action is used to track the type of so called ‘actions’ the web surfer is taking.  So for this example, you would use “Viewed Form” or “Completed Form” or “Received Error” or something similar.

Labels

The label is used to show additional info about the event you are tracking.  It is kind of like giving a page a title.  In this case this could be “Seeing Form On Page” or “Filled Form Out” or “Got Error On Form”, or it could just be “Form Page”.  If it is unique that is fine as well.

Values

The values field is an Integer or Number field, so it often won’t apply.  A great article about about this subject, the Event Tracking Guide, says that it is great for counting or summing things up like download time.

That’s it for now.  There will be a future article on my success or failure in implementing the event tracking in Google Analytics.

To solve this problem, there are several places to look, including Google and doing a site:YOURWEBSITENAME.com search and clicking on every link to using Google Analytics.  The Google Analytics method is the best, since it automatically tells you what needs to be fixed.  This does not mean you should ignore the google search engine and looking at what are big incoming links in advance.  Remember once you find a broken link, and you are using the WordPress Redirection Plugin, I highly recommended in a previous blog entry (click here to learn more info), you need to enter the entry into Redirection and fix it.  If you are just learning that this did not require an .htaccess entry in Linux or htapi entry in windows, you are learning about a whole new world of improvement…

So back to Google Analytics.  First of all you need to have a 404 page, which is where all the pages end up when they find no entry in WordPress, including the redirection plugin.  This typically will be a 404.php page in the theme directory you are using.  If you are sending people to your site map page, that is fine as well. 

There are several ways to get this info in google analytics.  I recommend using page title.  For instance, on our site, the page title of our 404 page is called “Page Not Found”.  Go into Google Analytics and click on “Content” and then on “Content by Title”.  If you have just migrated, and the first day has passed, just set the date to the current day, to remove anything that is older than today, and find the “Page Not Found” (or other) title.  If you click on this title name itself, another screen will open and reveal all the URLs that arrive at that page.  This is your list of incoming broken links, sorted by the most important links at the top of the list.

Now is the detailed nitty gritty work.  You need to re-enter each of these links into a url to test them and make sure they are broken.  Once you are have found one that is being clicked on from a website out yonger, take it, copy it into the redirection plugin.  You do this by Adding a new Redirection, then copying the link into the Source URL.  Remember to not use your full path, just the /filedirectoryetc/ path, and then go over to another browser and find the end URL you want to send the visitor to and then copy and paste this path into Target URL and click Add Redirection.  The default is a Permanent 301 redirect. 

There are a lot of other options in the redirection program like using Regular Expressions.  These work fine as well, but make sure you read the instructions carefully, since some redirections are impacted by other redirections…sequentially.

Let me know if you have any comments are questions about this process, because almost all sites come to point of migration, and this blog entry is really about the aftermath of a big migration.  Our last WordPress migration did go through, but it had lots of small items that we had to fix along the way.  I am going to add another blog entry covering these additional problems we ran into, creating a load balanced WordPress Enterprise solution!