Web Analytics Guy's Blog

SXSW from a Web Analyst Perspective

webanalyticsguy — Sun, 07 Aug 2011 18:41:52 +0000

With the recent announcement from the SXSW Interactive Director to raise prices by 27% for early registration, I thought I would branch out a little bit from my usual posts and talk about what I thought about SXSW Interactive from a web analyst perspective. The reason that the Interactive Director gave for the price increase was the new price is competitive with other pure tech conferences. While SXSW also includes a music and film portions, the majority of the attendance is for the interactive portion and the price increases for music and film were not quite a steep. My personal thought for the price increase is to control the number of people that will attend the conference. There were rumors that SXSW was going to limit registration and while they did not limit registration, by raising the rates it has the same affect and if the same number of people attend as last year then SXSW will just make money. Now the question is, with the price increase or without the price increase, is it worth going to SXSW Interactive.

I was fortunate enough that my company paid for me to go last year and with the price increase it might price out people paying to go themselves or small companies that might not have the increased budget. Before I attended SXSW Interactive last year, I had heard a lot of things about SXSW and was really excited that I was finally getting the opportunity to go. I had pretty high expectations for the conference, rightfully so or not, about some of the new tech companies and new technologies that I would be learning about. SXSW was definitely a very social conference. They had an app where you could plan, scheduled, and see which presentations that you wanted to go see. For almost every presentation there was a twitter hastag that you could follow during or after the presentation. One of things that I noticed about the conference is that people are not shy. People were not afraid to walk out of a presentation and tweet while they were walking out sharing their not so positive feelings about the presentation and the speakers. But for the most part, attendees did a great job recap the session so if you could not attend you would get an idea of how the session went.

When I first looked at the schedule to see which sessions I wanted to attend, I noticed that the schedule was from Thursday evening to the end of the day on Tuesday. It is five days straight days of sessions and walking around downtown Austin trying to get from one session to the next one. Being a web analyst and web analytics implementation program manager and just leaving the Omniture Summit, I had a web analytics mind set. I had not really taken the opportunity to look at the schedule to see which sessions that I wanted to attend. But what I was excited for was to learn about some new up and coming technology. I was looking at the SXSW schedule on the plane right to Austin from the Omniture Summit. I started picking some of the sessions that I wanted to attend, which had nothing directly to do with web analytics but some of the topics are related to web analytics. The first session that I attended was about privacy and it was one of the best sessions that I attended. The session was about who should control your data and some thoughts on how to opt out of tracking for ads. I attended several sessions about privacy and these were some of the best sessions that I attended during the conference. When I looked at the schedule I saw sessions that were called workshops and there was a workshop that included web analytics and survey data. When I saw this session, I was very excited. I thought there was finally a session that I could really engage. When I saw the word “workshop” I was thinking that it would be a discussion group and doing some actual work in web analytics. My definition and thought of a workshop is different than SXSW’s. There definition of a workshop is a two-hour session with three to four presenters and no breaks. While I was not expecting experts, I was expecting more than I got and I walked away from the session extremely disappointed.

The sessions that I was really looking forward to were the keynote speakers hoping to learn about some new technology or at least something that was interesting. There are five keynote speakers, one each day, at the same time each day. Getting in to hear the keynote speakers is very difficult and usually a long wait so I would just go to one of the other rooms where they would have a simulcast of the keynote. Last year most of the keynotes where not very memorable and most were not very good. There were only two that I really remember and only one that I thought did an outstanding job. The one that I thought did an OK job during the keynote was Seth Priebatsch from SCVNGR. He talked about how to use game theory to solve problems. The presentation was pretty good and Seth had a lot of energy, but never having been exposed to game theory before he lost me about half way through. The other keynote that I liked and I thought did an awesome job was Blake Mycoskie, the founder of TOMS shoes. The main idea behind TOMS shoes is that Blake will give one pair of shoes to children in need of shoes in a third world country for every shoe that is purchased. He talked about how he started the company and how we came up with this idea “One for One.” While his presentation was not about a new technology or a technology, it was still a very interesting story and a great presentation.

The last part of the conference that I attended was the trade show, which had a combination of music, film, and interactive companies with different booths. The interactive companies had the largest presence at the trade show from big companies (Microsoft and Monster) to smaller companies. Almost all of them giving away some sort of swag and most of which you do not want. But there were a few good items. I got a couple of signed books and a couple of signed footballs from current UT football players. Again, this was another place where I hoping to see or learn about a new technology or at least see something new from a company. Again I walked away from the trade show disappointed, except for the books that I got.

While overall I enjoyed my time at SXSW and attended some great sessions and some sessions that were not so great. The trade show was OK, but not that good. I think SXSW is over valuing their conference a little bit. While I can understand a price increase 27% is too big of an increase. If my company would chose to send me again to SXSW again, I would be happy to give up another weekend to go, even to just a few sessions.

Overall I would give SXSW 2011 a B.

Filed under: Social Media, SXSW, Web Analytics Tool Tagged: all top, linkedin, Online Marketing, Rich Internet Applications, technology, web analytics

What was the percentage of the page that was viewed?

webanalyticsguy — Sat, 14 May 2011 15:49:47 +0000

In the years that I have spent in web analytics, one of the questions that I get is we are thinking about redesigning a certain page or creating a marketing promotional page on the site and we how far down the page did the users view of the page. My first answer is to recommend running a test on the page to see which design drives the most success. I usually then get a response back that they do not have a testing tool, the time, or the resources to run a test. After I stop crying from that response, I then say let me see what I can do. After doing a little digging to see what solution that I can find. I came up with a solution that will allow the teams to see how much of the page initially viewed and how much of the page was viewed in total. This solution will take the screen resolution of the users out of the equation to get the most accurate information possible. I then can create a classification file to create different groupings for the report, which include total % of page viewed and scrolling %. After I created the report, I showed the original report to the teams and explained each of the different views. For this report, you are actually going to need two variables (s.props) to take full advantage of the reporting: one to calculate how much of the page was viewed and one to capture the previous page the user viewed.

Percent of Page Viewed
The original percent of page viewed report is very detailed, but is the base report where the classification file is created original percent of page viewed report When looking at this report, the number to the left of the pipe is the total percentage that was viewed and the number to the right of the pipe is the initial percentage that was viewed. The zero means that the percentages could not calculated on that page. The report is very detailed and can be hard to consume, but it is the based for different classifications for the report.

Initial Percentage of Page Viewed Grouping
One of the first classification reports that can be created is the initial percentage of page viewed grouping. This report is breaking apart the two numbers separated by the pipe. This report is looking at the number to the right of the pipe and creating different groups. This is just a grouping all of the percentages, which will allow you to see in a aggregate view the how much of the page was viewed when the page was first loaded in the browser. Again, because this is a percentage of the page it does not matter what the screen resolution is for the user.

Initial Percentage of Page Viewed
The initial percentage of page viewed report is another classification report that you can create that is taking the number to the right of the pipe and adding a percentage to it. This report is a detailed report and can be hard to read. When this report is combined with the initial percentage of page viewed grouping report, it makes the report a little easier to digest. The best way to consume this report is to look at the grouping report first and then drill down to this report. I will show an example of this a little later.

Total Percentage of Page Viewed Grouping
Another classification report that can be created is the total percentage of page viewed grouping report. This report is looking at the number to the left of the pipe in the original report and creating different groups. This report is another way to group the report so you can see in an aggregated view how much of the total page was viewed to see how much of the content was viewed the user. One of the questions that I have been asked is how many times did users view more than 75% of the page. This report will be helpful in answering that question.

Total Percentage of Page Viewed
The total percentage of page viewed report is similar to the initial percentage report, except that is looking at the number to the left of the pipe in the original report and adding a percentage to the number. Again, this report is a very detailed report and can be hard to consume when just looking at this report. The best way to consume this report is to look at the grouping report first and then drill down to this report.

Scrolling Percentage
The previous reports are the building blocks for the scrolling percentage report. As I am sure you noticed in the original report, the scrolling percentage is not listed or mentioned. The way to calculate the scrolling percentage is to use the classification file to subtract the initial percentage of page viewed from the total percentage of page viewed. This report gives you the ability to quickly see how much of the page did the user scroll down to see additional content and information. One way to use this report is look run the total percentage of page viewed grouping report and then drill down to the scrolling percentage report, which will allow you to see how much users are scrolling who viewed over 75% of the pages in total.

Total Percentage of Page Viewed Gropuing by Page
The next few items are examples of some of the reports that you can run by drilling down from a classification report to a classification report or from a classification report to the previous page report. In this example I have drilled down to see which pages users have completely viewed over 75%.
What I did here was go to the total percentage of page viewed grouping report and then I drilled down to the previous page report. Using this breakdown will allow you to answer to the original question that was asked of which pages were viewed more than 75% of the total page.

Page by Total Percentage of Page Viewed Grouping
Now let’s say that you want to know how much scrolling or what was the total percentage that was viewed for a specific page. So now you are going to go to the previous page report and then drill down to the total percentage of page viewed report. This report will give you a good idea of how much of a specific page was viewed. If you put a specific piece of content or a button a lower on the page, this is a great way to check if users even say the content or button on the page.

Page by Total Percentage of Page Viewed Grouping by Total Percentage of Page Viewed
While the previous report is a great view of the data, you are viewing a aggregated report. After you drilled down to the grouping report, now you want to see more detail. You want to see the specific percentages within the grouping report to get more information. Here is an example of the report, where I picked a specific page and drilling down to the total percentage of page viewed grouping report and then I chose a specific grouping and drilled down to the total percentage of page viewed report so I can see which percentage in the grouping was viewed the most. The groupings have a fairly big range and by adding this view it will allow you to see if it is higher in the range or lower in the range.

Finally, if you have Discover or Insight you can create a segment or filter on these reports to be applied on your other reports. While I think these reports are great and give a lot of good information, they do have their limitations. The percentage of page viewed reports have the same limitation as calculating time spent, the percentage cannot be calculated on the first or only page in a visit or on the exit page in the visit as the percentage cannot be calculated for the page until the another page is viewed in the visit.

Filed under: Percent of Page Viewed Tagged: all top, linkedin, omniture, optimizing, web analytics

People, Process, and Technology?

webanalyticsguy — Mon, 08 Nov 2010 01:55:15 +0000

It has been a little while since I wrote my last blog post and I thought it this was an appropriate topic to blog about. Some of you might recall Eric Peterson has done several presentations on this topic. This topic has also recently been mentioned on Twiiter. As I have mentioned from the beginning I am going to write about I am going to write about my personal experiences and this topic was a good topic to write about as I have started my new position. I am going to write about my experience with each of these with my various companies that I have worked. I am going to start with technology, then people, and finally process.

Technology
This is the one area that I have been very lucky to have the companies that I worked for buy into, for the most part. All of the companies that I have worked for have selected top tier tools, which could be for web analytics, surveys, testing, email marketing, etc. Some of the tools that my companies have selected over the years and that I have used are: CoreMetrics, Omniture (SiteCatalyst, Discover, Test and Target, Survey, Genesis, Data Warehouse), Google Analytics, comScore, and Nielson), and a couple of business intelligence tools (Oracle, Business Objects, and Teradata). Though I have been fortunate to use a variety of tools throughout the years, not all of my previous companies have wanted to invest in some of these tools. The reasons why they did not want to invest in some of these tools came down to the companies did not think they could get a good ROI from the tools and they were not ready to invest in the tools (either not in the culture or not enough people). When I say invest I am even referring to the free tools and the companies did not have the people in place to take on the new tools. Though I started thinking about these tools and as an analyst has our relaying on some of these tools too much actually make our analysis more difficult?

People
Of the three, this one has been the most challenging with some of the companies that I have worked. I have worked in departments that range anywhere from 2 people to 50 people where I have worked on multiple tools to just one tool. The departments that I have worked in that were 2 to 3 people where I was able to work on multiple tools was a great learning experience, especially earlier in my career, though with all of the tools that companies can use today I do not know how realistic this is today. When I have mentioned adding some new tools, besides not understanding the new tool, it also came down to not having enough time for someone to work with the new tool. In other words, there were not enough people in the department to spend time with the tool. Besides not thinking they could get a ROI from the tool, the companies also did not think they could get a good ROI from adding new people to the department, even if I suggested using a free tool to show the company how the tool can help the company make better decisions to increase the conversion. With all of the different tools that companies can use, should they hire at least one person for each of the tools that they company uses?

Process
Of the three, the process is the one that I have continuously worked on throughout my career and it is the hardest one to achieve. With so many different tools, a good process needs to be set to either implement the different tools, add a new item to be tracked, or new tests to be run. A process needs to be set up for all of the tools that you have, but these are just some examples of the different processes that need to be set up. Though it does not require any monetary investment it does require a time investment and buy off from your internal clients to follow the process that you have set up. This can be challenging as people have different ideas about which process might work and the different departments might have different requirements, so getting buy off from the different departments can be challenging. Another challenge for creating a process is when two requests that come in have the same priority level, how do you determine which project to work on first? There are a few different ways you can determine which project to work on first. You can look at the potential revenue for each of the projects and whichever project has the most potential revenue is the project that gets the top priority. You can also look at the time frame for each of the projects and whichever project has the earlier time frame could get the top priority. But what if both or more requests that came in have the same revenue potential and both have the same time frame then what do you do? This is where setting up a group from different departments can be useful to talk through the different requests to help set the prioritization for the different requests. Even though the top requests have the same revenue potential and due date having a group of people from different departments discuss the different requests can be very beneficial in deciding the prioritization of the projects. When I have been in a department of two, the process was easier as there were fewer people that I had to get approval from for prioritization of the projects than when I was in larger departments. For larger departments a good process is a must to make sure everyone is on the same page with the prioritization.

Conclusion
Though I think that all three of very important, I think developing a good process is the most important of the three. Convincing the companies to invest in new tools and to invest in the people to manage and use those tools for analysis can be difficult, but if you have a good process for analysis and prioritization of the different projects and requests you can show the company what kind of ROI that can be achieved from the new tool and people to manage the new tool.

Filed under: people, process, technology Tagged: coremetrics, omniture, omniture implementation, optimizing, people, process, technology, web analytics

Merging Data from Different Sources

webanalyticsguy — Mon, 03 May 2010 03:19:07 +0000

In my Year in Review post I mentioned that I was not going to focus on certain tools and focus more on my experiences that could be applied to different tools. In this post, I am going to cover a little bit of each. I am going to focus more on a specific web analytics tool and how to merge data from different sources to get a more holistic view of the visitor. My experience of merging data from different sources comes mainly from a media site perspective and that is what I am going to focus on. As some of you may know, Omniture allows you to use Data Sources to import information from another source into Omniture. That is one way to get an overall view of the customer. However, I am going to focus on option from Omniture that is not as widely known called ‘Data Feed.’ The data feed is a great option if you want to export your Omniture data into another tool to do more analysis.

Data Feed Information
The short description of the data feed is the raw data for one report suite. Another way to think about the data feed is the data that Omnitue sees before any scrubbing of the data has occurred, which includes before any of the of the IP Address blocking has taken place or any of the VISTA (Visitor Identification, Segmentation and Transformation Architecture) rules have run. Each compressed data feed file; one per report suite contains several tab-delimited text files. One file contains traffic data and the other files contain reference data. Some columns in the traffic data file contain a numeric ID number that references a value in one of the reference data files. The data feed can be sent to a FTP account as frequently (hourly) or less frequently (yearly) as you want, though in my experience I have found a daily feed to be the most effective. The FTP account that the data feed file is sent to can be one either one that Omniture sets up or one that your company sets up. As visitors will be viewing your site at all hours of the day, some of the visits might not have finished when the data feed starts to run. For those visits that were not completed when the data started running, they will be included in the next data feed file. The size of the file can vary greatly depending on how much traffic the one report suite is collecting. Though on average, the data feed will be 500 bytes for each page view or 500MB for 1M page views per day. To get more information about setting up a data feed for your company, I would contact your account manager.

Business Case for Using Data Feed
Working for a media company presents some challenges that companies in other verticals do not necessarily face. A little while ago, the core site team and the online marketing came to me and they wanted to know how much revenue we were making per page view and per visit. To a lot of companies, this is a pretty easy calculation but for a media site this can be a little challenging especially with the systems that we are using. The other reason this can be a little challenging is the revenue can change on a page by page basis, depending rate for the ad impression. The first hurdle that I had to overcome is the system that we use to display ads on our site. The system that we use to display the ads on our site is DART for Enterprise (DE). DE is exactly the same thing as DART for Publishers (DFP) except that DE is housed on our servers and we do not get the updates to the software as often. But we still have the same flexibility to target the ads that will meet the client’s needs. Since the DART data sits on our servers, it is a little more difficult to export the data out of DART and into Omniture. So after doing a little digging and making a few phone calls, I soon learned about the Data Feed that Omniture offers. After thinking about how to best use the data feed, I went and talked to our Business Intelligence team to see we can work out. After running through some of the different options, we ultimately decided to send the data feed to our FTP account and to upload the data into Business Objects. Once we loaded the data into Business Objects, we then could link the Omniture data to the DART data since one column in both files are the same. Remember when I mentioned above that the data feed is the raw data before any of the scrubbing has been done yet. One of the first things that we had to do was apply the same VISTA and/or ip address exclude rules to the data feed that we do to the Omniture data to get the data feed numbers as closely as possible to the numbers in Site Catalyst. Once we accomplished getting the both sets of data as close as possible then we can now merge both the DART data and the Omniture in Business Objects. Once the data is merged together in one table, we can then calculate the revenue per page view and per visit. Also, since we are using the data feed, which has all of the data, we can then also calculate revenue per page view per traffic source. This information will allow us to see which page or pages are our highest revenue generating pages and which traffic source is driving the most revenue. Even though a particular traffic source is driving a lot of page views, but are those pages high value pages or lower value pages. This information will then let us know which traffic source we need to focus on to drive more traffic to higher value pages.

Filed under: Merging Data, Traffic Sources, Web Analytics Tool Tagged: Data Sources, linkedin, omniture, Online Marketing, web analytics

Traffic Sources Reporting

webanalyticsguy — Mon, 15 Mar 2010 03:16:13 +0000

I have been back for a little over a week from another great Omniture Summit and my fourth summit. I went over my notes from each of the sessions that I attended and some of the slides of the presentation that I could not attend. I am not going to write a full recap of the summit this year, as there are already a few of them out there that do a great job recapping the summit and some of the sessions. Some of the sessions during the summit covered different ways to track the traffic sources. For those who attended last year’s summit, you might remember the announcement of the Channel Manager plug-in. The Channel Manager plug-in is very similar to the Unified Sources VISTA Rule (USVR). The biggest difference between the two is the Channel Manager plug-in needs help from the consulting group and the Unified Sources VISTA Rule needs help from the Engineering Services group. I am going to write about the Unified Sources VISTA Rule, some of the reports that we had created from the USVR, and some of the decisions that can be made from these reports. When we decided to go with the USVR, the next decision was decided which reports we would like to have set up. We talked to several of our internal clients and then thought about the business and which reports would help the business make better decisions. I worked very closely with the engineering services group to make sure the logic would be exactly what we wanted.

After talking to our internal clients and thinking about the business needs, we decided on three reports that we wanted created for the USVR. The first report that we decided on was an easy decision. We wanted a channels report to get an overall view of all of the channels in one report. After we talked for a little bit, the second report was also fairly easy to decide on. The second report that we wanted as a part of the USVR is a referring domains report. I am going to explain a little later why we decided to have this report added as a part of the USVR. The third report that we wanted to have for the USVR was a little more challenging to decide on. After doing some thinking, we decided we wanted to have a natural search report created. I will also covered this report in a little more detail a little later.

Traffic Sources – Channel Report
We knew very quickly that we wanted to have the Channel Report. For our weekly and monthly KPI reports, we report on these channels each week, but we were using several reports to pull this information together. From the out of the box Traffic Sources reports to our tracking code reports. We wanted everything in one report so we did not have to use multiple reports and we also wanted to be able Visits for every channel. The only drawback to the channels report is the direct load traffic shows up in this report as “None.” The “None” traffic source can make it difficult to trend the direct load traffic source. But before I get into how to fix this, here is a look at what the Traffic Source report looks like. This report has all of our traffic sources in one report, which gives us the ability to add Visits to the report as well any of our success events, such as, page views, video ad begin, video content begin, and sweepstakes entries. This report allows us to see which channel is bringing in more visits and which channel is driving more page views, video plays, or sweepstakes entries. In the time that we have had this report, we discovered very quickly which channels were bringing in more visits, but those channels were not necessarily driving more page views or video plays. It told us very quickly which traffic sources that we need to focus on to drive more page views or video plays on the site and which traffic sources are working. One of the best things about this report, is that you can trend the report and see the traffic sources next to each other. This report can also be used for trouble shooting. To give you one example, one of our sites has a huge spike in traffic one week. We were able to trend the channels report and see which channel had the spike. It turned out to be a partnership campaign that we were running. We then went to the partnership report to find the exact campaign that drove the increase in traffic. When we found that information, we used that information to also drive a home page promotion. As I mentioned in the original report, direct load shows up as “None” which cannot be trended. To fix this issue, you can classify the original channel report and classify “None” as “Direct”. After “None” is classified, “Direct” can then be trended and compared to the rest of the traffic sources.

Traffic Sources – Referring Domains Report
This report was an easy decision to include in the USVR for two reasons. First, we wanted to be able to see Visits for each of the referring domains instead “Instances” that comes in the out of the box report. Second, we enabled full sub-relations for the Channel Report, which then allowed us to drill down from our Sister Site channel and Other Site channel to see which specific referring domain have the most visits and drove the most success. This report might not seem like an obvious choice to pick for the USVR. We picked this report for a couple of reasons. One of them was to be able to see visits for each referring domain, but the biggest reason was to be able to see exactly which “Other” sites are driving traffic to us. In the out of the box reports some of the your campaigns (twitter, facebook, etc) will show up in Other sites, but in the USVR channel and referring domains report those campaigns will put in a bucket and the true “Other” sites will show up.

Traffic Source – Natural Search Report
The third and last report that we decided we needed was a Natural Search report. This report is also the most complicated, but can drive a lot of decisions. Once of the driving factors for this report was we have been focusing on optimizing our sites for natural search and this report will show us if those efforts have been working. After we decided that we wanted to have a natural search report, the next decision was what did we want the report to look like. Knowing that we only had one variable to use for this report, so we had to be creative. We decided early on, that we were going to use classifications for this report. We also wanted to go one step further than just passing the natural search keywords. What we decided to do was pass the natural search keyword and the landing page concatenated in one field (keyword:landing page URL). This report will let us see which natural search keyword and landing page combination has the most Visits and which ones drove the most page views and video plays. One of the things that we noticed that certain combinations driving more video plays and certain combinations that were driving fewer video plays. When we looked at the landing pages driving more video plays, we noticed there was one thing that was different about those pages than the other pages, which gave us some information that we can use to add to the pages driving fewer video plays and see if that change will drive more video plays. As I mentioned earlier, I knew that we wanted to classify this report. The first classification we wanted to do was break out branded vs. non-branded keywords. By classifying the keywords into these groups, it allows us to quickly see how our branded vs. non-branded keywords are performing. Though we decided to breakout the keywords into these two groups, classifications are flexible enough to allow you to bucket the keywords in whatever keyword groups would be best for your company. One of the next classifications that we created was classifying the landing page into the different page types. Besides just looking at which page types were getting the most page views or which pages were users entering the site from, we wanted to see which page types were getting visits from natural search and which page type was driving the most traffic. Your home page will probably get a lot of traffic from natural search, but is their another page type that is driving more page views and video plays. Two more classifications that can be created is breaking out the keywords and the landing page URL into separate reports This is very useful since the original report is a concatenation of the natural search keyword and landing page URL. This will create several keyword and landing page combinations and breaking these out in classifications then aggregates all of those combinations into one line item. Another capability that classifying this report gives you is to be able to sub-relate any of the classified reports by one another. One of the reports that we run all the time is breaking the keyword group (branded or non-branded) by the landing page URL. This report allows us to see which landing page URL brings in the most traffic from branded vs. non-branded keywords and which one drives the most traffic. One of the last things that can be created for all of these reports are some calculated metrics (ex. Visits/Page Views, Visits/Video Content Begin, etc). These metrics can then be added to all of these reports to see not just which channel, referring domain, or natural search keyword/landing page URL has the most visits or driving the most page views, but also which ones have the most visits/page view or visits/content video begin.

Filed under: Traffic Sources Tagged: KPI, linkedin, omniture, search engine optimization, Success Events, Videos, web analytics

Adoption of Web Analytics in Your Company

webanalyticsguy — Sun, 31 Jan 2010 20:29:21 +0000

In the last seven years that I have spent in web analytics one of the biggest challenges that I have had is getting adoption of web analytics in the company. This can be very challenging for a variety of reasons. Some of the people in the company might not know about, know how to use, or understand web analytics. Some people in the company might not think web analytics can help them. These are just a couple of the reasons why adoption of web analytics in a company might be challenging. This was not any different for the first company that I worked for when I first entered into web analytics. They are a multi-channel retailer and were primarily a catalog company. They are a small company and became very well known through the catalogue. They launched the website as an after-thought. In the beginning, they really did not have any type of tracking on the site. Of course this was way before Google Analytics came around. The only tracking they had were which products were purchased on the web site. I think they might have had something that told them how many visits the site received and the entry and exit pages. Though no one looked at it and all of the decisions about the site were made based on how many products were purchased through the catalogue. When I came on board, I was the second full time employee dedicated to the web site. My main responsibilities were marketing of the site and web analytics. After we picked a web analytics solution, one of the next things that I worked on was to get adoption of web analytics in our company.

In the last seven years, I have had varying degrees of success and challenges in getting adoption of web analytics. The biggest challenge that I had at the first company was getting buy-in from the merchandisers. Being a small company, we only one set of merchandisers for the whole company and they decided which products would be sold in the catalogue and web site. They were catalogue merchandisers and did not think much about the web site, much less about web analytics.

In my years of trying to gain adoption, I have tried seven different ways to get adoption of web analytics in my companies. I would start slow with one or two and then try one or two more.

1. Automate some basic reporting – This one might seem obvious, but is often overlooked. When I say automate some basic reporting, I am not talking just about a KPI report. I would create and automate reports for different departments. This is a good way for each of the departments to know what metrics are available. This step is not to convince users to use web analytics, but just to let them know that the data is available and to give the users and idea of some of the information that they can receive.

2. Run a more detailed report with analysis – After a little while of receiving some basic information and letting the users know what data is available, I would then create a more detailed report (a few more metrics) and provide some insight for each of the reports. At this point, other departments will start to see what other information is available and how they can use the information to make decisions.

3. Monthly meetings with departments and recommendations – The next thing I would do is set up a monthly meeting with different departments at one time. This meeting will cover the basic reports and the more detailed reports, as well as recommendations based on the information. In this meeting, I would also include some comparisons against the different departments which could create some friendly competition for the departments to see which one can improve their numbers the most.

4. Partner with one department – One of the other things that you can do is partner with one of the other departments in your company. I would partner with one of the departments that have taken more of an interest in the information that you have been providing. I would work with this department on providing a little more analysis; including segmentation from traffic sources and which segment drive more conversions. Once one department starts using more of the data, the other departments will want to know more.

5. Training a handful of people in other departments – One way to get more adoption of web analytics can be referred to as “train the trainer”. In this case, train one person from each department, who shows the most interest in the data and who you consider a “power user”. You can train these people through the whole tool, how to quickly run different reports and how to make decisions from the information. This person will then train other people in their department on how to use the tool. Though there might be some questions that the person you trained might not be able to answer, this is a great way to get more people to use the tool.

6. Hold a monthly training class – As the person who you trained in the different departments might not be able to answer all of the questions, a monthly training class is a good way to help fill in the gaps and to show a few more tips on how to use the data. It is also a good refresher for the users in the company. As much as I like to think that everyone logs into the tool as much as I do, that unfortunately is not the case. This is a good way to keep the tool and the data in front of a lot of people and to get a lot of people to log into the tool at least once a month.

7. Create training videos specific to your company – Creating training videos is a great supplement to the monthly training class. It is also a good resource for new employees who need to use the tool. Though Google Analytics, Yahoo, Omniture, Web Trends, and other tools have training videos, they are great for basic information. Also, their training videos are not specific to your company or your company’s needs and the training videos that you create can be a little more specific. I have tried to use two different tools to create the training videos. I am sure there are more than two, but both of these are free and offer some basic video features. The two tools are: BB Flashback Express and Jing by TechSmith. The flashback express tool offers a few more features with its free version, but I would recommend trying both and see which one works the best for you.

There are more than these seven ways to get more adoption of web analytics in your company, these are just the seven that I have tried and have found some success in using.

Filed under: KPI, Web Analytics Tool Tagged: all top, linkedin, omniture, Online Marketing, Videos, web analytics

Year in Review

webanalyticsguy — Wed, 23 Dec 2009 09:00:25 +0000

It was a year ago this week that I was traveling for the holidays and I made the decision to start this blog while sitting in a hotel room. I knew there were quite a few blogs about web analytics already with various opinions on different topics. When I decided to start this blog I wanted to focus on one thing, sharing my experience in web analytics both good and bad. The experiences that I have shared have been about recent topics and some as far back to when I first started my career in web analytics. I have had a great time writing this blog this past year sharing my experiences. I hope that I was to give you more information about a decision or give you an idea about some things to try. I look forward for the coming year and sharing more experiences with you. Though in the past year, I have focused on certain tools, in the coming year I am planning on sharing experiences about various tools and analysis. In the great time that I have had sharing my experiences, you may have come to my blog and not find the answer you were looking for. I compiled a list of the top search terms that brought users to my blog. I am going to try to give more information and/or links for each of the different search terms.

Omniture Implementation Manual
Getting your hands on the implementation manual can be tricky. In order to get a hold of the manual, you have to either be an Omniture customer or a consultant working with Omniture. If you either one of those, the implementation manual can be located under supporting docs > manuals. There is a column of all of the implementation manuals including the Site Catalyst implementation manual.

Implementing Omniture
An omniture implementation can be a little tricky and you need to plan out what you want to track to help you figure out what users are doing on your site. I wrote two blog posts about this that I hope you will find helpful.
First post about Implementing Omniture
Second post about implementing Omniture

Web Analytics Solutions Profiler
This is a great tool to help you de-bug various web analytics tools. This tool was originally developed and built by Stephane Hamel and recently purchased by iPerceptions. I use this tool all the time to make sure my pages are tagged as well which analytics tool are the competitors using and what information are the capturing. It is a very easy to install plug-in for FireFox. I also take advantage of the crawl feature of this tool.
Web Analytics Solutions Profiler site

Omniture Blog
This is a great resource for some quick tips or to see what is happening in Omniture on different topics from SEO, SEM, Web 2.0, Web Analytics, and mobile analytics and more.
Omniture blog main page

Charles Debugger
This is also another great de-bugging tool. I use this tool to most of my QA when a new tracking feature is launched, we click tracking are added to links, and for flash applications.
Charles Proxy

Omniture Custom Link Tracking
Custom link tracking is great way to track clicks on a link, especially for promotions or links that change out on a consistent basis. It is a great way to aggregate links or modules that reside on multiple pages on the site. It also gives you the flexibility to pass any value into a prop or eVar or pass events. You are also able to trend the click data. The one drawback about custom link tracking is it a call to Omniture, so make sure that your contract can handle the additional calls. To find information about implementing custom link tracking, go to knowledge base and search for link tracking. In the knowledge base article there will be pdf that will give you all of the information.

Tracking Marketing Campaigns
This is another topic that I have written a blog post about. Tracking marketing campaigns is something that is very easy to do, but is something that not everyone does. By tracking your different campaigns, you can see which campaign drove the most success and which ones did not work as well. Depending on which tool you use, will depend on how the tracking will be added.
Traking Marketing Campaigns post

Omniture iPhone Tracking
Tracking of the iPhone applications is one of the great features of Omniture. Though very recently Apple has changed their Terms of Service and the original Omniture iPhone code does not meet that change. Omniture has updated their iPhone tracking code to meet the new Terms of Service. If your iPhone app was rejected in part due to the Omniture tracking, just log into the admin console and download the latest version and replace the old code in the app. That will bring the tracking into compliance with Apple.

Omniture plugins
The plug-ins is one of the greatest codes to implement for an Omniture implementation. You may not need to implement all of the plug-ins, but there is certain plug-ins that is almost mandatory regardless of the type of site. To find a list of some of the plug-ins, you can go to Web Analytics Land to get a list and the code.
SiteCatalyst Plug-ins

Omniture Flash Tracking
Tracking a flash application or flash site with Omniture is fairly easy to accomplish with Omniture’s ActionSource. ActionSource uses the native language to flash, ActionScript. This makes implementation of ActionSource easier or so my developer tells me. There are two ways to implement ActionSource, either through a component that can be placed inside of the flash application, where you can assign the variables inside of the component or you can write the ActionSource directly. To find the implementation manual can be located under supporting docs > manuals. You will want the Media Tracking and ActionSource document. The ActionSource piece is found in chapter two. Some of the items that I have tracked with ActionSource are: flash videos, hotspot clicks, clicks on links that leave the site, etc.

I hope this information was helpful. I look forward to sharing more information with you in the coming year.

Posted in De-bugging, Rich Internet Applications, Social Media, Success Events, Videos, Web Analytics Tool Tagged: De-bugging, linkedin, omniture, omniture implementation, Online Marketing, Rich Internet Applications

Web Analytics and Voice of Customer

webanalyticsguy — Sun, 06 Dec 2009 19:53:22 +0000

And the survey says……(this reminded me of watching Family Feud when I was younger). A couple of weeks ago I read a blog post from iPerceptions about how they are going to integrate web analytics with behavioral and attitudinal data. As I kept reading the article, I kept thinking “WOW that is awesome.” Then I got to the Features and Benefits and I was really impressed with the crosstab capabilities of the survey and clickstream data and to look at your analytics and survey KPI’s in one dashboard. After reading the article, I started to think about my other co-presented breakout session at the Omniture 2009 Summit about surveys and VOC strategies. If you can combine clickstream and survey data, create segments based on those on a both sets of data would give you great insights about your customer. When you are thinking about putting a survey on your site you need to think about what type of questions you need to ask. Do you want to ask Idea questions (comments, feedback), Opinion questions (ratings, polls), or Effects questions (site changes, product variations). Before you create surveys, you need to understand your business goals: establish a voice within the community, build trust with your customer base, generate positive brand awareness, encourage loyalty, improve customer perception. There are many different types of surveys that you can put on your site that will help you get into your customers mindset.

Some examples of the different surveys that are on the sites: product reviews, news story reviews, conversion funnel monitoring, time on site assistance, customer service feedback, remarking surveys, quick polls, or exit surveys. Most of these surveys are “traditional” surveys, but they are a good way to get into your customers mindset. If you can combine the answer to these surveys with the clickstream data, gives you a better understanding of your customer.

I am going to go over some of the different survey types and some of the survey companies. Different types of surveys will give you a different view of your customer. Quick Poll’s and Product Reviews are are not the traditional surveys, but can give you some great information.

Quick Poll: This is a short survey that can be put on different pages on the site (section front pages,

home page, or registration page) or in an email. A quick poll will allow you to learn more about your customer a little at a time and will help you to either target certain content and/or send the customer an email based on their preferences.

Product Reviews: A product review is another good way to learn more about your customers, but it is also good information for the buyers to see if they should continue buying more of that product or buy a similar product. It is also good information to go to the vendor with any feedback about the product.

For remarketing surveys and exit surveys, there are several different companies that can help you with your business goals. There are surveys that will fit different budgets and that have different functionality.

4Q from iPerceptions: 4Q survey is a free and easy way to reach out to your visitors and capture their feedback about their user experience. This survey will help you answer four questions; How satisfied are my visitors? What are my visitors on site to do? Are they completing what they are set out to do? What are they saying in an open-ended text box? These are great questions to get to know your visitor better and can help you answer you answer your business goals. If you do not have any surveys on your site, this might be a good to get started.

Survey Monkey: Survey Monkey is a survey tool that that has a free and low cost option that you can completely customize with whatever questions will help you answer your business goals. Survey Monkey will allow you to collect responses either through the site or through an email. There are several types of questions to choose from, as well as, survey templates and custom themes. You can also add a logo to the survey.
Once you start to collect data, the results will start to show up immediately. You can share the results of the survey without having anyone log into an account. You can also download the results from the survey and combine the data with your clickstream data for more analysis.

Opinion Lab: Opinion Lab is an opt-in, voice of customer solution by inviting customers to share insights in their own words from every page on the site. Customers submit feedback through an opti-in comment card. They structure the visitors insights and deliver the data through their reporting tools. Opinion Lab’s approach is based on opening an ongoing pipeline of communication with the visitors to gain valuable insights that include information about your site, products and services, brand image, and competition. Opinion Lab is a passive survey. There is a link on every page on the site (usually in the footer), giving all of the visitors an opportunity to give feedback about the page or their experience on the site.

Fore See Results: Fore See Results is a survey that will appear randomly while the visitor is browsing the site. Their surveys include a broad range that drive user satisfaction, including navigation, site performance, and look and feel. Fore See uses the American Customer Satisfaction Index (ASCI) methodology to analyze the user satisfaction which allows you to identify what is driving customer satisfaction and what will have the greatest impact on site behavior and revenue. ACSI also allows you to benchmark against your competitors. The surveys are on a 10-point system to allow for greater differentiation of responses. The surveys take approximately 2-3 minutes to complete. If you are an Omniture customer, Fore See is a Genesis partner which will allow you to upload the survey results into SiteCatalyst.

Omniture Survey: Omniture Survey is a survey tool that you help you measure visitor sentiment on your site. Surveys can be distributed to your customers in various ways, either through email, when a visitors lands on a certain page, exits a page, or when a particular event occurs. Omniture Survey allows you to measure opinions, attitudes, and motivation in real-time, and take immediate action based on the responses. If you are an Omniture customer, Survey will not use any additional variables and the reporting will work very similar to the conversion reports. Once you start to collect some survey results, you can see how much success was had after the visitor answered a certain question. You can also correlate survey answers to other conversion reports to see which products where viewed or videos were played by visitors who answered the different questions. Omniture Survey is customizable to your company’s needs. You can create as many questions as you think are necessary to help you answer your business goals. The survey results are not only passed into SiteCatalyst, they are also passed into Omniture Discover. This will allow you create a segment based on both clickstream data and survey answers and apply that segment to any report to get a deeper insight into your visitors.

As you can see, there are several different survey options. No matter what your budget is or what web analytics tool that you use, there are options out there to learn more about your customer and merge clickstream data with survey data to gain more insights.

Posted in Survey, Web Analytics Tool Tagged: KPI, linkedin, omniture, surveys, web analytics

Advanced Analytics for Search Engine Optimization (SEO)

webanalyticsguy — Sat, 17 Oct 2009 18:22:24 +0000

As I mentioned in my Omniture Summit 2009 recap, I was fortunate enough to be a co-presenter for a breakout session – ‘Makin’ Gold Records: Turning your SEO team into Rock stars.’ One of the most difficult things to do is track SEO and to determine how successful the SEO efforts have been. I am not going to talk about tips about how to optimize your site for search engines. There are already quite a few blogs and forums that talk about how to do this. What I am going to talk about are reports and analysis that can be set up to help determine how successful the SEO efforts have been. I am going to borrow from one of the slides in the presentation that really helps with this point, ‘43% of marketers don’t or can’t accurately measure ROI from SEO. To be successful with SEO, you must accurately measure success to drive value and action. In the end, it is not about driving traffic to the site, it is about making money.’ The question now becomes where to begin? To borrow another quote from the presentation, ‘With so much data available…the most important metric is that which helps improve your site.’

This is really good advice, but before I get into some of the reports that can set up to help analyze SEO, you need to make sure to get some of the basics down first. No matter which tool you use, you need to make sure you use tracking codes for your paid search marketing campaigns. This is a small, but very important thing that needs to be done. Depending on which web analytics tool you use, will depend on how this get accomplished. To accomplish this for Omniture specifically, you will need to go to the admin console to the Paid Search Detection area and enter which query string tracking code(s) that is going to be used for your paid search campaigns. For example, you will enter ‘cid’ into the text box next to Query String field. For the Search Engines, I would leave the setting to ‘-Any-‘, unless you want to add the tracking code to a specific Search Engine.

The last thing that needs to be decided, before you start to set up reports for tracking SEO, is determining which KPI’s are the most important that will help you improve your site. The KPI’s that will be the most important to you will depend on the vertical of your site. Here are some examples of some SEO KPI’s:
• Keyword Conversion Rate – Revenue Per Search
• Entry Page Conversion Rate – Leads Per Search
• Conversions – Leads, Orders, Subscriptions
• Bounce Rate – Keyword or Entry Page
• Average Time on Site – Keyword or Entry Page
• Keyword Conversion Rate – Page Views / Search
• Conversions – Page Views
• Conversions- Video Plays
• Keyword Conversion Rate – Video Plays / Search
• Keyword Conversion Rate – Sweepstakes Entries / Search

SiteCatalyst has a couple of Natural Search reports out of the box that do a pretty good job of analyzing most of the KPI’s above, but you can go one step further. There are two ways that will allow you to create additional natural search reports. The first way is through a Unified Sources Vista Rule and the second way is through the Channel Manager Plug-in. These are two different ways to help you crate marketing reports. Each one has its good points and bad points. The biggest difference between the two is the Unified Sources Vista rule is created through Engineering Services, while the Channel Manager Plug-in is created through Consulting Services. The Channel Manager Plug-in does allow you a little more flexibility, but with some good planning you can accomplish the same reports with either option.

Here are a few of the reports that you can have create with either option.

Natural Search Keywords – This report might seem a little strange to put on this list since it is an out of the box report. Even though there is an out of the box report for Natural Search Keywords, there is one metric that I would like to this report and the only way to get it is to pass the keywords into an eVar. The metric that I would like to add to this report is Visits. By adding visits to this report, will give you a slightly different view of the keywords report and allow you to create calculated metrics that are visit based. You can use this report to quickly see which keywords are performing the best for different metrics or KPIs. You have the ability to easily sort by each column to see which keyword drove the most engagement or conversions (ex. Page Views/Visit or Video Plays/Visit)

Natural Search Entry Page – This report can be very valuable by itself or in combination with the Natural Search Keywords report, which is another reason to pass Natural Search Keywords into an eVar. You can enable full sub-relations for both of these reports and you can drill down from the keywords report to the entry page report to see the top entry pages for different natural search keywords. The report by itself can help you see what the pages visitors are entering your site on, which can help you discover gaps in your keyword strategy by identifying entry pages that drive the most engagement or conversion from natural search. You can see how deep into the site visitors are landing and do those pages drive more engagement or conversion than the home page or a section front page. You can create a calculated metric for bounce rate and see which entry page has a higher bounce rate. You can then go to your SEO team and let them know which entry page is not performing as well as you would like it to, you then can do some testing on a new design to see if a new design might drive more engagement or conversion.

Classifiy Natural Search Keywords – Classifications is another reason why it is a good idea to pass the Natural Search Keywords into an eVar. It will allow you to classify the keywords, which helps you to logically group your keywords into different buckets that make analysis more effective and actionable. One of the questions that I have been constantly for a SEO team is how well branded and non-branded keywords perform. When you classify the keywords, there is not a need to download the Natural Search Keywords report each time and then try to go through each keyword to find branded and non-brand keywords. The classification will allow you to easily see this information in a single report. Classifications also an easy way to breakout Natural Search keywords by any dimension that is needed, besides just branded or non-branded. Classifications are set up in the admin console through the SAINT upload.

Backlinks Report – This report is another valuable report to give you information on which site that is linking your site is driving more engagement or conversions. The backlinks report is based on filtered referrals and can give you insight into your existing link popularity strategy. If you have a linking strategy that involves blogs, facebook, twitter, or some other site this report will allow you to quickly see which linking strategy is working the best in driving engagement or conversion.

Another good report that you can utilize to help you with your SEO strategy is an internal site search report, which can help to identify additional SEO keywords to optimize your site. Though this report is not specifically tied to the Unified Sources Vista Rule or the Channel Manager Plug-in, it can be very helpful in finding SEO keywords. You might not want to just look at internal keywords that drives success, but also at keywords where users did not find what they were looking for. You can go to the search team and the SEO team to talk about how you can leverage that information by building content on the site that is SEO friendly. Some additional tools you can use to help you research which Natural Search keywords to focus on are: Hitwise, Google Trends, and Webposition. These tools will help you look at natural search keywords that are driving traffic to your competitors.

SEO Forum and Blog
Highrankings Forum

Jordan LeBaron – Omniture SEO Consultant

Posted in Online Marketing Tagged: linkedin, omniture, Online Marketing, search engine optimization, web analytics

Sizing It Up: Measuring and Optimizing Rich Internet Applications

webanalyticsguy — Fri, 25 Sep 2009 12:00:22 +0000

When I started thinking about what I was going to write for my next blog post, I had a couple of topics in mind trying to keep with my original plan for my blog of sharing my experiences in web analytics. Then the news broke that Adobe has agreed to purchase Omniture. Whether this acquasition is going to be good or bad, only time will tell. There has already been enough written on this topic, so I am going to focus this post on another topic. When the news broke of the acquisition, an idea for a post came to me. About a year ago or so, I co-presented a webinar with Omniture for an Adobe online seminar. Little did I know what would transpire a year later. As you can see from the title of the post, the webinar was about measuring and optimizing Rich Internet Applications. I have been very lucky in the last couple of years to have the opportunity to measure and optimize RIA’s and to work for the company that worked with Omniture to develop their flash tracking (ActionSource). Though I am not a developer and have never actually added the ActionSource tracking I work very closely with our flash developers to make sure the tracking that I would like is added to the RIA’s. Though I have tracked many different applications, most of the RIA’s that I going to cover in this post involve flash video, but I think this will give you a good idea of some of the tracking capabilities.

Before I get started going over part of the presentation, here is a link to the Adobe Webinar. In order to play the webinar, you will need to have an account with Adobe. If you do not have one, you can create an account. Creating an account is free and after you register will take you directly to the webinar.

Optimizing Rich Media
As I mentioned in one of my other posts, Video Tracking, video has been one of the bigger priorities for us and we are incorporating video in a more integrated fashion through the site than we have before. When we first developed this player, we started with the video list, but questions come up, would a Full Video View be a better experience, which would give users have the option of viewing the video list by clicking on More Videos link in the upper right hand corner. Instead of just letting our gut instinct take over and fall into the typical decision making trap, we decided to let our visitors tell us which video player they preferred. We ran an A/B test on both of these versions with a 50/50 split, using the Video List as the control and the Full Video View as the test. Take a second and think about which player you think generated more video plays? We have asked this question to people in the past and majority of the people thought the Full Video View preformed the best. After we looked at the results we learned that the Video List had 35% more video plays than the Full Video View.

Hotspots or Overlays
When we added hotspots to the videos, users had the ability to click on certain links within in the video as the video played to learn more about certain products. But this was not driving user engagement with the hotspots because we believed user’s were not aware that there were links with in the video that they were able to click on. Knowing this information, we then wanted to test this against having an overlay in the bottom portion of the video where users can click on the overlay and bring up a window over the video. The video behind the overlay will pause while the user views the information in the overlay, where they can play a video or click on links that go to the advertiser’s site. It is not a big surprise that the overlay drove more click-throughs, but the question is by how much. We quite surprised to see that the overlay had a 56% better click-thru rate than the hotspotting. After seeing this information, we are now in the process of adding overlays to all of our videos and this also gives us an additional revenue source with our advertisers.

Other Platforms (Widgets)
Now that we have tried the traditional modules and traditional applications, Scripps wanted to try something new. Go outside of the box and saw an opportunity to try the next new thing. Since the iPhone is one of the hottest phones and the market, we decided we wanted to build an iphone application for Food Network. Was this going to work? We were not really sure, but we knew we wanted to tracked how our users interacted with the application, including which tab users clicked on click, which videos users viewed, and if they were using the search box which would drive users to the Main Food Network site. The iPhone application was considered a success and we are now in the process of developing iPhone applications for some of our other sites. After the iPhone application we wanted to try another, more fun application within Facebook, so we built a cupcake application, where users can create their own cupcake, chose the wrapper they want, the type of frosting, and place a message on top of the cupcake and send it to their friends. On the right hand side of the application, we listed some cupcake recipes and we tracked if users clicked on any of the recipes that would then take the user to the Food Network site. This application was not quite as successful as the iPhone application but was pretty successful. After seeing this, we wanted to really go outside of the box and different and we created a Food Network widget for the Chumby, where we would upload a recipe of the day for the Chumby, where users can log onto the Chumby and recipe the current recipe or view any past recipes. We tracked which recipe users viewed and how they moved through the widgets, either through hitting the arrows or scroll through the slides. After looking at the data for the Chumby, we saw that the Chumby was a failure and we have discontinued creating any new widgets for the Chumby, which goes to show you why it is important to track and test these initiatives so you know where to focus the resources.

So what have we learned from all of this? By using the ActionSource tracking (except for the Facebook Cupcake application, where we used tracking codes), it gave us the ability to track the different interactions that our visitors used and helped us better optimize the experience. It also helped us make decisions on which video player technology would drive the most user engagement which and be the most effective for our advertisers and it let us determine which non-traditional applications would be worth investing more time.

Posted in Rich Internet Applications Tagged: linkedin, omniture, optimizing, Rich Internet Applications, Videos, web analytics