Omniture Summit – Day 3

I called this day “The Maroon 5 Effect,” because everybody going to breakfast showed the effects of going to the concert and staying up late. I started noticing the effects of it right on the bus, where it was awfully quiet, even though the bus was full of people. Over at the Salt Palace, the scene looked of a zombie movie. There were only a portion of the attendees eating breakfast, but most were eating their meals. Not much was discussed during the morning session, other than key speakers selling their products to us.

The best breakout session in the event had Allan Dick, CMO of Vintage Tub & Bath. Not only was very passionate about his business and very entertaining (he gave us the story of the tub and toilet in 10 minutes). He explained how he used both SearchCenter, Data Sources & SiteCatalyst to bridge online and offline data to get a better view of what keywords are converting and generating sales. He explained that a visitor is given a unique customer ID, which can then be used when a visitor calls their call center to make the purchase. This data is then imported to SiteCatalyst using Data Sources and tied to the keyword data.

The closing session was a product road map discussion, which is precede by Brett M. Error, EVP and Chief Technology Officer. This is an interesting forum because here’s where us as clients and product users can provide comments and recommendations for improvements. Not only Mr. Error but his team of developers asses the situation and make notes, which are then reviewed and implemented. What’s different about other forums is that Mr. Error mentions which of the past year’s recommendations where implemented into production and provides with release dates. This year, the majority of the issues had to do with improvements of the Discover platform. I left the event with the same opinion that I have regarding the products: they are going in the right direction but there’s room for improvement.

That night, there was a reception, where we were given another opportunity to network with clients, as well as have some fun. The ambiance was of a winter wonderland, with a dancer dressed like a tree, moving around the bar. There was also a photographer, taking pictures behind a snowy backdrop. The party continued to a club, but I needed to pack my bags and get some rest before my return trip. So that’s Omniture Summit 09…at least from my point of view.

Omniture Summit – Day 2

Day 2 of Omniture Summit was the official kick-off of the event, and proved to be the longest day of the whole week. The event began with Josh James giving his “State of Omniture” address, where he announced the Omniture Marketing Suite, which encompassed some new products like Omniture Recommendations, which provides product recommendations based on the purchasing behavior of the visitors; Omniture DigitalPulse, which is used to verify the health of your analytics implementation by pointing out any discrepancy that might impede the tracking code from fully tracking each visitor behavior and CMO Dashboards. Mr. James also mentioned the integration of Test&Target with SearchCenter, which is the first step in integrating all Omniture products. Mr. James interviewed the CEO of WPP, Sir Martin Sorrell, discussing the advances in mobile tracking and the need to “get on the mobile tracking bandwagon.” Interesting enough, not enough emphasis is being given to the Social Media trend that is currently dominating how we interact with the Internet and each other. One case study that was impressive was given by Fred Kleisner, CEO of The Morgan Hotel Group, who presented a comparison between site road maps when their first one was launched (2001) and their new site (2009). It was interesting to see the attention to detail in all aspects of optimizing the site, mimicking the experience that a visitor would have when visiting one of their properties on the web.

The rest of the day was going to different breakout session, where a client would give a case study of how they used Omniture products to increase their revenue and better track visitor behavior. I attended the Financial vertical, in which the Omniture consultant mentioned the need to use internal search and mobile media. The next session that I attended talked about APIs. Omniture announced new APIs like SAINT and Discover. They also announced that our entry for Omniture’s Developer Challenge came in second place and our own John Nguyen presented the project and answered many questions. There was also a presentation by Martin Lindstrom, the author of Buy*ology. In his session, he talked about brand awareness and what make us purchase the things that we have. The best thing about the session is that everybody got a free copy of Buy*ology.

That’s the end of the first part of the day…the second part started as soon as the sessions where over. After a not-so-quick meal, where Kim, John and I discussed everything from analytics, to sports, to why our dinner was taking so long (seriously, I think they went to the market to get the products,) John and I went back to the Grand America Hotel for the Maroon 5 concert. There where a number of themed rooms, where you could get a bite to eat, as well as a drink. John and I decided to go to a room where a piano man was taking recommendations. He sang everything from Abba’s “Dancing Queen,” to Garth Brooks’ “Friends in Low Places,” to Flo Rida’s “Low (yeah, you heard it right and it was funny too.)” I must say that the Maroon 5 performance was better than expected. They played for an hour and a half (because they have only 2 CDs.) As far as the experience, they sounded phenomenal, better than the CD. Surprisingly enough, the Boston Celtics stayed in the hotel because they played the Utah Jazz on Thursday and some of them where seen among the analytics geeks in the crowd, clearly showing that Omniture knows how to throw a party. For an after-party, we got invited to a suite on the 23^th floor, where Omniture hosted a bar and food for us attending the party. John and I got the opportunity to network with some of the attendees (although I doubt that some will remember.) Personally, it was super fun, but going to bed at 2am was not as easy, especially since we needed to be at the Salt Palace at 8am.

Omniture Development Challenge

The Omniture Development Challenge results are in and our Page Ranking entry came in second place. Our John Nguyen made a demo of the entry and although a little nervous, he demonstrated the qualities of what our Geary family. Good job to everybody involved and let’s shoot for #1 next year.

Omniture Summit 09- Day 1.5

Hello from snowy Salt Lake City. As part of my attendance at Omniture Summit, I’ll be bringing you some daily updates on what I say, ate and learned in my week-long adventure. You can also see my “up-to-the-minute” comments (or rants) in Twitter (http://www.twitter.com/rmanzanet).

For this blog entry, I used “1.5” because I’ve been in Salt Lake City since Monday afternoon. With every story there’s a beginning and this one started with a snow storm, San Diego style (lot of rain), which delayed my flight to Phoenix by an hour and I forced me to do a 50 yard dash to the gate, barely catching the connecting flight and was able to sit in the tail of the plane. After checking in, I got lost in the trying to find my room and fought with the door trying to get it open. Although one can almost predict the rest of a stay by the beginning of a trip, everything has been good…so far. I attended a reception at the Grand America Hotel, where I got my credentials and needed materials for the week. I was able to meet with some old friends from Omniture, as well as made some connections with some analytics enthusiast. We discussed about how the field of analytics had changed in the last couple of years, as well as our expectations for the week. There was a band made up of Omniture executives, and although I knew of them, never heard them live. I must say that they where quite entertaining, even hitting some high notes. I departed the event around 10:30pm, only because it was over and the place was being cleaned.

This morning, I woke up expecting snow on the grown, or at least in the air. I figured, if I spent the money on a nice coat, I wanted to break it in. I watched the news with the expectation of snow, but the weatherman said that it was going to be warm. So I got ready for the day, walked out to building and what do you know, snow was falling like crazy. In a way I was relieved because I got to use my coat, but this told me two things: that weather people need to do their homework better and that when snow falls, it gets cold…really cold. After getting the snow off me, I got into the bus and off to the official beginning of Omniture Summit 09.

I was fortunate enough to take the Landing Page Optimization Certification by MarketingExperiments. This course is originally given in a day and a half but because of Summit, it was compressed to a one day event. I went in to the class expecting to hear things that I already knew, but what Dr. Flint McGlaughlin and his team showed us was incredible. They approach landing page optimization from a quasi-scientific model, explaining that the probability of conversion is just a way of remembering those factors that affect a conversion of a site and how they relate to each other. They discussed how the placement of content, the design of a site, as well as your calls to action can positively or negatively affect your conversions. Another interesting item that they discussed was the idea of how a connection is done from a search engine to the landing page and how that flow can be broken if the information does not match on both the search engine and the landing page. At the end of the session, we had to take a test that had 71 questions. Although, it was an open book test, and some of the questions where discussed, it was still nerve-wrecking to finish it in an hour, but I did and think I passed it too. We’ll see, so stay tuned.

The evening was spent in the official reception, where I met with Kim and found that John was still on his way (at the time of this entry, he’s in his room finishing the presentation after drinking a pot of coffee). The theme for the reception was Mardi Gras, with beads, a Cajun-style band and a food court with themed food, like fried catfish. I introduced to Kim some of my ex-coworkers from Omniture, and networked with some of the attendees. One thing that never fails is at the degree that some people will go for a something free. For this year, in order to qualify for the drawing of a 50 inch plasma TV, you had to meet with every vendor there, listen to their pitch and get a stamp. It almost look like a carousel of people trying to get their passport tagged by the vendor. Wonder if we could dump that in a prop?

Alright, it’s almost 1am, so before I fall asleep I’ll finish for the day. Tomorrow (or today) will be another full day of sessions…and the Maroon 5 concert. So with that, stay safe and adios.

The perception behind Web Analytics

A belief that I’m trying to erase when talking to clients or anybody in the web space industry is that there is more behind the practice of web analytics besides page views, visits and unique visitors.  Although the practice of monitoring visitor behavior, as well as the tools to track them has come a long way since the early days of Hitbox Pro, there is still a belief that web analytics tools are seen as “add-on” product.  I’m here to shed some light on what web analytics is and how the data that it collects from your visitors can help you better position your product/service in the web.

Web analytics is more than just knowing how to code a page and looking at how much traffic it attracts.  It forces you to not only stretch the limits of technology but allows you to use emotions and your own imagination to concoct a strategy that creates a following and ultimately fulfill your goals.  Web Analyst (or whatever they call us passionate about web analytics, I call myself a Web Analytics Geek) have to break the perennial box and look not only at what’s “hip,” but also at the technologies behind what’s cool.  It forces you not to look at the specific details, but at a greater picture, because in essence, we look at trends in data to create answers.  “And what do you do with this data?”  You may ask.  We look at past data to get a sense of how healthy a site is, and to see what people like/don’t like about the site.  This give us the necessary information to plan a campaign, focus on an SEO strategy or overhaul the site because it’s elements where not working properly to “hook” visitors to what you’re selling. 

The practice of web analysis is more of an “art” because it allows you to use different disciplines and strategies to come to the same result, as well as dissecting the data that is collected to create a dynamic picture of your site’s health and how it grows (or shrinks).  In order to keep up to create strategies, I learn the business of my client, as well as what’s happening in that specific industry.  This is done so you can create a solution from the perspective of the client and the visitor to the site.  By getting intimate with what your audience likes or dislike, will produce measurable traffic that can be used to fulfill the client’s KPIs.As you can see, web analytics can do wonders for your business and should be used as an intricate part of the creative process.  We are more than just number crunchers or dashboard monkeys; we look at the flow of a site and how it marries with the web.  So next time you’re planning on expanding your web presence, use web analytics in conjunction to your creative, media or SEO efforts.  Otherwise, you’ll be driving through the internet half-blind. 

So what do you think?  Do you think that you can run a business with or without an analytics tool?  Let’s get a discussion going.

Analytics and technology don’t move at the same speed.

I was reading a blog post by Eric Peterson (http://blog.webanalyticsdemystified.com/weblog/2009/01/thoughts-on-the-proposed-iab-guidelines.html), in which he mentions how the IAB is proposing to further define “Unique Visitors/Users,” and the stand that vendors must take to keep accuracy.  To add to what he’s saying, technology moves at such a fast pace that it’s hard for analytics vendors to keep up with the trend.  This creates a false sense of accuracy in when determine what a “unique visitor” is.

Per the IAB, “Unique User/Visitor” is defined as an identified and unduplicated individual Internet user who accesses Internet content or advertising during a measurement period.  In order to qualify, you must access the site with a cookie, browser and device that is identifiable by the analytics tool.  The issue that I see here is that at the rate that new technology is released to the public, the harder it is for vendors to identify, test and release solutions that adheres to the above definition.  Take for example our perception on what can access a site.  A couple of years ago, you needed a computer to do this.  As time progressed, this ability was then adopted by cell phones and now video game consoles can access the Internet.  The problem here is that it puts an added pressure to vendors to maintain their collection practices to follow a standard definition that is quite difficult to maintain.  In order to follow a strict metric definition, a vendor would in advance have access to new technology and ample time to modify and test its solutions before making the solutions available for public consumption.

Although I understand and applaud what the IAB is doing, I see a couple of reasons why vendors won’t follow them:

1. Analytics tools will never be precise and these definitions expect each behavior to be “linear.”  If it was, then it would cease to be an analytics tool and become a statistics tool.  
2. All vendors claim that their solution displays the most accurate data.  Although they follow similar paths, each uses unique methods to provide a representation of said behavior.
3. New technologies are not readily collected by analytics tools or might be tracked as a different device.

One idea to keep is that analytics is supposed to provide you with a representation of a site’s health, which can be recorded and displayed differently depending on the tool, the visitor’s behavior, the technology behind the site and the devices used to view the site.  Rather than putting too much pressure on vendors to tighten their practices and release proprietary information (which they won’t), the effort should be made to gap the bridge between what is track and any new technology, thus maintaining a higher degree of accuracy. 

Anyways, that’s my 2 cents.  What do you think?  Do you think that technology and analytics will one day advance at the same speed or will technology always have the upper hand?     

Sports + Analytics = Higher Returns

Organized sports have moved from being a form of entertainment to becoming a useful marketing tool.  If used well, a sporting event or team can boost the popularity of a product, as well as a company’s revenue.  But you might be asking,”How can analytics mix in this equation?”  Simple, by using the team’s health, analytics will show you the areas where you need to improve your marketing strategies to drive traffic to your site and ultimately convert your visitors into clients.

Take the Tampa Bay Rays or the San Diego Chargers for example.  Both teams have either disappointed or were believed to make it far because of their past or current experiences, but the Rays made it as far as the World Series and the Chargers are putting every sport analyst to shame by making it to the Divisional Series.  What this means to your business is that behind these successes are fans that support the team, and you can use those emotions to drive more traffic to your site. 

One strategy is to promote your product/service on sites that cater to a team’s fan base.  As the season progress, not only monitor the team’s health but also your traffic.  My experience with clients that use this strategy is that their products trend in conjunction with the team’s health.  Look at your referring traffic to see what domains bring the most visitors to your site and focus your marketing strategies on those domains that are popular, while improving your ad placement or using other tactics to drive visitor traffic from those referring sites that bring the least traffic.

Something else to keep in mind is that the use of a team to promote a product /service needs to be done regardless of the team’s health to maintain support from the fans, while avoiding the “bandwagon effect.”  The bandwagon effect is a behavior where people adopt something that others do without consideration, which is closely related to opportunism.  This is clearly seen in sports.  As a team succeeds, it creates a greater fan base.  In order to find success in a team’s downfall is to use the hope of a better season, to drive the core fan base to your product/service.  This not only evokes loyalty from that group of fans, but the word of mouth will spread the popularity of the product/service as the team’s popularity increase.

So, what do think and who do you think will make it big in 2009 (besides the Chargers of course)?  Happy analytics!     

Page views, clicks, hits, they’re all the same…right?

One misconception that I’ve seen when talking about analytics is how people misuse terms when trying to define an action. This comes from adopting a tool’s definition for a behavioral action which has been used for an extended period of time. When you get used to that term, you might think that everything is the same. Imagine if you look at an orange. You might think that all oranges are the same. Not exactly. Even among the same fruit family, there are unique factors that differentiate one kind from the next. The same can be said when looking at tracking tools. In my last blog entry, I mentioned that each tracking solution tracks in accordance to it’s own architecture and metric definition.With so many terms floating around, it’s understandable that you might confuse one term from the other. A clear example that I’ve seen clients, and even internally do is confusing page views with clicks. Note that in general page views is when you land on a page, while click is the action of clicking on a link, which in turn generates a page view.

Another misconceptions is thinking that visits are unique visitors. Note that unique visitors are reported in different flavors (depending on the analytics tool that you’re using):

So next time you seeing data from different tools that are worded differently, rather than generalizing, take the time and look at their definitions and do a little bit of testing. You’ll be quite surprise that just because it looks like an orange and taste like an orange doesn’t have to be an orange.

Comparing apples to oranges…same juice, different taste

As the web keeps turning from being an extension of a business to the driving force behind it, more and more tools are made available to track visitor behavior.  With this comes the “apples to oranges syndrome,” which occurs when you start comparing data from two different systems and expect them to match.  The one occurrence that I keep seeing is the comparison between an analytics tool to the client’s ecommerce system.  Before I continue, I’ll just give you the punch line:  THE DATA WON’T MATCH!!!!!!      

Now that we got that out of the way, I’ll start explaining why the data won’t match.  When looking at a tracking tool, whether is an analytics tool, or a campaign tool, or your company’s backend system, each have its own set of moving parts, with some parts set in such a way that even though they might look to be tracking the same data when comparing to another tool, in reality they’re not.  Here are a couple of reasons that explain this phenomenon:

·         Unique architecture – as tracking tools go, each is created with its own set of algorithms and definitions, thus rendering them, as well as the data that is presented unique.  This can be seen when comparing data from multiple analytics tools.  Since each tool collects, process and presents data in its own way, what one tool considers quantifiable, another might not. 

·         Metric definition – as well as architecture, each tool defines certain behaviors in its own fashion.  What one tool calls a page view, another might call it a hit, and another might call it impression.  Each metric definition is placed because it fits best with the tool’s objectives and is more relevant to what it reports. 

·         Cookies – cookies play a big part in tracking visitor behavior because not only it track specifics about where the visitor was and where it’s going but it recognized a visitor if it leaves and decides to return (provided that the visitor accepts the cookie and doesn’t delete it after leaving).  Although the trend is to use 1^st party cookies because the data that it collects is more accurate and is less likely to be blocked by visitors, there are tools that by default uses 3^rd party cookies.  The problem here is that if you use a tool that uses 3^rd party cookies and another that uses 1^st party cookies and a visitor is blocking 3^rd party cookies (which is the default security setting in most newer browsers), the tool that uses 3^rd party cookies won’t track that visitor. 

·         JavaScript – most tools require you to place some JavaScript code on the site in order to collect traffic data.  Depending on what tool you’re using, it might provide you with a noscript tag that is placed on the page as well.  This ensures that if a visitor is blocking JavaScript, that its behavior is tracked.  If you have multiple systems but only one uses noscript tags, then that tool will be the only one that collects data for visitors who block JavaScript.

·         Security settings – security settings affect how different tools accept incoming/outgoing traffic and collects data.  One example of this is IP blocking.  If your company doesn’t want to track traffic that comes from certain IP addresses (like internal traffic, for example), some tools allow you to block them.  Another item to consider is that some firewalls are set to block traffic to certain domains, which affects how a tool tracks traffic.  For example, say that an employee in your company decides to buy one of your widgets from his work computer (yes, we all are guilty of that) and your analytics tool blocks all internal IP addresses.  When you look at your sales report and compare it to the analytics report, that visit and everything that pertains to it was not recorded. 

Now that you’ve read this, you might be thinking that they’re all useless.  “Not really”, I must say.  Understand that each tool provides a unique representation of your site’s health and how visitors react to it.  Rather than trying to figure out why the numbers don’t match and dismissing them altogether, take a step back and look at the bigger picture, look at the trends in them.  If you see that the trending behaves in similar fashion, you know that each tool is working as it should.  Remember that it’s very difficult to capture every single page view, and multiple tools won’t give you an exact representation, but they’re pretty close to the truth. 

The Art of Page Naming

Although it’s true that “it’s the little things that make life fun”, when it comes to analytics, the little things can make web traffic reporting more like a nightmare. 

This is true when dealing with page naming a site.  If done correctly, they present a detailed picture of your site’s health and your visitor’s behavior.  On the other hand, if care is not taken when creating a page name, the reports look sloppy and in some instances, you could be inflating page views by grouping traffic from multiple pages into one report.        

It’s true that there are ways to automatically create a page name on your site’s page, but it may not produce the right information or the information that it’s presenting won’t make sense.  When I explain to clients the importance of creating a unique page name, I pose this argument:  What if you’re a new Marketing Manager and your boss asks you to create a traffic report of a site that you’ve never seen.  “Easy” you might think.  “I just log into my favorite analytics tool and download the page na……hmmm.”  I forgot to mention that what you find as page name are things like index.htm, q=123, http://www.site.com among other names that really makes no sense.  As a bonus, there’s no documentation detailing what each code, or page is.  You could either go through each URL to try to figure out what each page is about, or try to “wing it.”  In the end, the report won’t look professional and might create more questions than provide answers. 

In order to create a report that is both clear and concise and avoid any confusion, follow these simple steps:

1.       Before doing any modifications, look at your current reports and identify pages where the information presented is not the best.  This could be because of length, using query variables to identify a page, displaying a URL, among others. 

2.       Once you have a layout of all pages and domains, create a naming convention that can be used across all sites.  This is very important because if you’re looking at pages from different domain, you want to make sure that you understand the path that a visitor is taking.

3.       Remember that the page name is the lowest form of content.  Page Names should contain a brief description of where you are with relations to the site and what information is on the page.  If you’re looking at SEO information, for example, the best practice for implementation is [site name]:SEO.  This is very important if you have multiple domains that contain similar information.  You can have page names such as Domain1:SEO, Domain2:SEO, etc.  Not only this tells you the location but also provides you with feedback if the visitor goes from one domain to another in your site.

4.       Have someone else look at your work.  Even though a strategy makes sense to you, it doesn’t mean that it make sense for everybody in the organization.  By having a coworker review your page naming strategy, not only you can verify that the information that each page will produce is the right one but that is easily understood.

5.       Test, test, and when you’re done…TEST AGAIN!!!  There is nothing worse than a good plan that is badly implemented.  The best way to test data is by having a separate account for this purpose.  You can send traffic there for a determined period of time (I tend to test an implementation from 2 weeks to a month).  The reason behind this is to keep test data from being mixed with “live” data.  Plus, in the event that something is done wrong, it can be changed without any repercussions.  Once you’re happy with the reports that are being generated, switch the test account to the live account and SUCCESS!!

Although creating a proper page naming convention might look like a time consuming task and be more of an art form than a chore, at the end of the day, hard work is rewarded in quality.  And that my friends, is time well spent.