Emerging Technology Blog at Web Analytics Demystified

Tel Aviv (Israel) based beencounter is a service unlike any you have ever seen we suspect. By “hacking” your web visitor’s browser history files, beencounter provides interesting insights into your audience. An API makes this information actionable for personalization and targeting efforts.

What does beencounter do?

beencounter (the little “b” is their branding, not a typo) is, in their own words, a “Behavioral Targeting and Behavioral Tracking” service. Built around a single line Javascript tag, beencounter basically gives you access to your site visitor’s browser history files. And while they cannot tell you all the places your visitors have been, by providing the service with a list of URLs, beencounter can tell you whether your site visitors have been to those URLs sometime in the past.

Here you can see the results for Web Analytics Demystified over a few weeks:

What this report tells me is that during the timeframe I selected about 18% of my visitors had also been to the main Google Analytics page, about 6% had been to Omniture’s home page, and about 4% had passed through blogger Avinash Kaushik’s blog main page.

Pretty nifty, huh?

How does beencounter work?

beencounter is doing something surprisingly simple: they are taking advantage of a known hack of the browser Document Object Model (DOM) that identifies URLs that have been previously visited. The way this manifests normally is showing visited links on a page as a different color than unvisited links. You can also see this information if you open your browser history.

All you have to do is provide beencounter a list of URLs that you want to watch out for and they do the rest.  What’s more is they provide an easy-to-use API so that you can actually query the visitor’s browser in real-time and see if they have been to a site or group of sites:

Once you’ve added your sites, beencounter will provide a Site ID, and then just a little bit of Javascript is all you need (and the “is good but we are better” example text comes from beencounter, not us):

Sweet, huh?

What makes beencounter different?

I don’t follow behavioral targeting solutions very closely, but I’ve never seen a solution that provides this much programatic access to insights into visitor behavior across the Internet. When I saw the application I immediately thought of a dozen cool things I could do with only a tiny bit of programming.

Don’t get me wrong: beencounter has some particular challenges … specifically the fact that many people who learn about the solution are likely to think beencounter is incredibly invasive and crosses the line regarding consumer privacy. For example, I am watching for people going to ZAAZ and the WAA, but I could just as easily try and determine who in my audience spends time at Playboy.com or one of the many gambling sites out there.

Fortunately Nir Ben Levy and his team at beencounter had the same thought, and they have an active program to block risque, inappropriate, or otherwise potentially invasive sites. Mr. Ben Levy said they have over 2,000 “blacklisted” sites in a variety of domains, perhaps most importantly including financial services domains where someone could use this service for phishing or other nefarious activities. They manually maintain the blacklist which I’m sure makes for interesting conversation at beencounter HQ.

Who will benefit from beencounter?

In terms of benefits to the web analytics community a few immediately go to the top of the list:

1. If you have a powerful segmentation tool you could use the beencounter API to dynamically populate custom variables with the list of sites and/or API IDs for post-hoc segmentation. For example, I could create a segment in Omniture Insights of “people who love Avinash Kaushik but not the Web Analytics Association”
2. If you’re interested in measuring “loyalty” to a site, you could mine beencounter via the APIs to determine what percentage of visitors are also visiting your competitor’s sites (e.g., “disloyal”.) This is particularly cool because loyalty, as used in web analytics today, is a pretty bastardized term to say the least.
3. From a targeting and testing perspective, it would be pretty simple to write a dynamic rendering engine that swapped out key site messages based on the results of calls to the beencounter API. For example, we might have a different homepage message for people we know have been to Gartner or Forrester, and we could then test differential messages using web analytics. While not a proper A/B test (unless you really randomized the audiences) this would likely yield some interesting data.

On point #3, given that the pricing plan that includes API access starts at $49.95 per month, for motivated individuals beencounter is likely the cheapest behavioral targeting solution in the world.

Things we like about beencounter:

• Provides amazing access to visitor information and offsite audience behavior
• Very simple to implement and use for folks with even basic Javascript skills
• UI is easy to understand (nothing fancy) and includes export to Excel

Things we’d like to see from beencounter:

• More information about their “blacklist”, including access to the list, and more information about their efforts to prevent abuse
• Faster servers and an uptime SLA (we have seen some slow response from their Javascript) OR a server-side option so that we could run locally and thusly manage our own uptime/response
• An option to have their script “walk” the DOM and look for pattern matches, not exact matches exclusively

Learn more about beencounter’s service at http://www.beencounter.com

Posted by Eric on Monday, February 22nd, 2010 | 28 responses | Add a Comment | Share, Save or Email

Washington based Scout Analytics offers unique and actionable analytics for subscription-based services and recurring revenue models. Their patent-pending technology identifies usage behavior and generates a comparable index of customer demand that can be used to increase customer lifetime value.

What does Scout Analytics do?

Scout Analytics is a SaaS-based predictive analytics tool that offers insight into your existing customer base. It provides behavioral analytics that can be used to recognize opportunities and risks based on subscriber usage patterns. For example, the solution can predict the effectiveness of cross-sells and up-sell opportunities by associating individual customer behavior with similar customers in a segment. It can also identify patterns within user behavior that are indicative of risk, which can be used to take action to prevent churn for individuals or groups of customers (see screenshot below). Additionally, it can help recover lost revenue from unlicensed subscription use. It’s a tool designed to ensure revenue retention, expansion and assurance.

How does Scout Analytics work?

Unlike most analytics tools that rely on cookies to track individual user behavior, Scout Analytics uses javascript to collect a “signature” of customers. Signature inputs include: biometrics (via typing patterns), device (via device parameters), and network (via IP address). The device and biometric signatures increase the accuracy of unique user counts, while the network signature aggregates context called “firmographics” to organize customer segments by industry, size, geography and rate plan. These metrics incorporate individual user sessions, downloads and transactions that are cataloged to formulate a customer demand profile. This profile is then translated into a numeric ranking, called “Demand Rating” that offers a relative comparison of subscription cost vs. usage for each customers within a segment. Customers with ratings are generally identified as at risk, while high rated customers are targets for cross-sell or up-sell opportunities. Additionally, the ratings of customers within a segment can be averaged to create a score for the segment enabling comparisons of different customer segments (e.g., is demand stronger in SMB or enterprise?)

What makes Scout Analytics different?

Scout Analytics belongs in a new category of tools that uses data within a very specific context to drive action. It provides the ability to effectively reach out to customers based on their actual behavior and revenue potential to deliver relevant, targeted services, which builds and strengthens the relationships. The Demand Rating metric developed by Scout Analytics offers relative comparisons of customers and prospects allowing businesses to prioritize sales, marketing and product development activities.

Who will benefit from Scout Analytics?

This tool is not likely be used by data modelers or analytics power users. In fact, that’s one of the things we like about it. It’s designed for use by everyday marketing and sales staff to identify opportunities and take action. Anyone that can read a stock ticker will benefit by understanding where revenue growth opportunities exist and when churn is a potential threat. Any business with a recurring revenue model or one that delivers subscription-based revenue services will recognize immense value in this tool.

What you need to know about Scout Analytics?

Scout Analytics is a new kind of data tool that will empower your sales and marketing staff by pinpointing opportunities and threats. This information can be extracted and synchronized with SFA applications like Salesforce.com; augmented with existing web analytics data; or integrated with CRM and Marketing Automation tools. Scout Analytics also has a track record of delivering between 10% and 15% increase in revenue for existing clients.

Things we like about Scout Analytics:

• No need to be an analytics guru to operate this tool.
• Demand Rating is a single, informative and actionable KPI that can be tracked over time, across segments or product lines.
• With average revenue increases ranging from 10%-15%, Scout Analytics pays for itself.

Things we’d like to see from Scout Analytics:

• Greater expansion into multi-channel customer activity.
• Comparable metrics to Demand Rating that extend the toolset capabilities to other business models.
• Just how far Scout Analytics can take their concept of cookie-less tracking with device and biometric signatures.

Learn more about Scout Analytics by visiting their site at: http://www.ScoutAnalytics.com

Posted by John on Monday, February 8th, 2010 | No responses | Add a Comment | Share, Save or Email

Dutch based Nedstat recently integrated the possibility of measuring Engagement within their standard Sitestat technology. While Engagement means different things to different users, through their integration with Live Segmentation, Nedstat’s Visitor Engagement offering is flexible enough to adapt to audiences’ needs.

What does Visitor Engagement on Sitestat do?

Well basically, at first glance, it’s a scoring methodology as it allows to attribute points to certain events undergone by visitors on a website. Nothing new here for those who have already done some scoring exercises in the past. However, it goes further as with their Live Segmentation feature, you can zoom in on specific visitor behaviors, according to traffic acquisition methods for example (email campaigns, SEM keywords, social media initiatives, etc.) and add rules in order to come to data that would reflect your site’s definition of high, medium or low Engagement.

How does Visitor Engagement on Sitestat work?

Sitestat makes use of labels upon which you can then segment your data. A new label ns_ilevel has been introduced that allows you to attribute scores to important events such as specific pages and/or actions (contact us, form registrations, member sign-ups, downloads, etc.).
You then use Live Segmentation in order to focus in on visitor behaviors, whether from a purely scoring point of view or together with other metrics you deem important in your Engagement calculations such as the more classical metrics of number of page views for example as it all depends upon how YOU define Engagement for YOUR website(s).
You can then retarget those visitors with high Engagement scores: those who have seen a specific product but did not convert by exporting their data (if you have an email address for example) and pushing this towards your emailing systems (in this case) with a specific promotion.

Why is Visitor Engagement on Sitestat different?

It’s not the new label that’s interesting to be totally honest as we’ve seen that before and can be rendered by a lot of web analytics tools but also content management systems. It’s Nedstat’s endless pursuit for easy integration with existing and available technology that makes it attractive, something they had already shown with their Microsoft Suite integrations, making it easy to extract data and get it into Excel, or their Attribution models which go far beyond the usual “first click/last click/something in between” solutions proposed. Clients can choose from ten attribution models including the visitor engagement option whereby the engagement scores of the visitors determine the attribution per channel.
Additionally, their data is apparently not aggregated and can always be reprocessed. You attributed a too high score to a specific event? Not a problem! The event can be rewritten and another score can be attributed to it.

The example shows the marketing touchpoints a visitor came across before placing an order. The attribution  to each touch point is shown based on various models. In the final column the engagement model is used to determine the attribution value.

Who will benefit from Visitor Engagement on Sitestat?

Ok, so let’s be honest here, while Engagement has become quite of a buzzword, when we talk about Engagement, true Engagement, it’s not something that you can just figure out for your website while having a coffee!
Those who will benefit from Sitestat’s addition of scoring are those who are savvy enough to know what they are looking for and how they define Engagement in order to get the right data out of the tool and use it to the benefit of their digital strategy.

What do you need to know about Visitor Engagement on Sitestat?

It’s not just a module; it’s an additional scoring method in order to be able to compute Engagement according to your needs. It should therefore not be seen of some kind of standalone; it’s an integral part of the Sitestat technology and Nedstat’s offering as it goes together with Live Segmentation at the very least.

Things we like

• The flexibility of using Live Segmentation together with the scoring methodology
• The Attribution models
• The Microsoft suite integrations
• The possibility to reprocess the data

Things we’d like to see

• Being able to attribute scores within the module as opposed to having it hard coded
• Proposals for Engagement measurements per sector
• Engagement distributions: not only a number but a more context based view where the distribution of Engagement would be shown

Posted by Aurelie on Tuesday, February 2nd, 2010 | No responses | Add a Comment | Share, Save or Email

Delaware based Shufflepoint has taken the most user friendly web analytics application in the world today and managed to make it even more useful. Their ability to move data easily between Google Analytics and Excel makes GA an even more viable option for businesses of all types.

What does Shufflepoint do?

In a nutshell Shufflepoint connects Google Analytics indirectly to Microsoft Excel and Powerpoint, arguably two of the most widely used analyst tools in the world today. While somewhat time-consuming to use the first time with a little practice Shufflepoint has the ability to dramatically reduce the amount of time analysts and GA-focused consultants spend gathering data (you know, so you can spend more time actually analyzing the stuff!)

As you can see in the following screenshot, Shufflepoint allows you to drag Metrics, Dimensions, and Profiles into a series of “targets” that govern how the report will look. If you’ve used Yahoo Web Analytics to create custom reports this interface will feel extremely familiar to you.

How does Shufflepoint work?

Shufflepoint simply makes excellent use of GA’s free APIs, essentially serving as a bridge between GA and Excel (or Powerpoint) to allow you to build moderately complex custom reports. Leveraging both a drag-and-drop interface and a query language that is in many ways similar to SQL (see image below), Shufflepoint makes Google Analytics native export functionality more or less obsolete. More importantly, while the team at Google has done a pretty good job with the new “custom reports” interface in GA, Shufflepoint extends GA’s custom reporting paradigm to the logical conclusion: relatively easy to generate documents that can be shared the same way more or less all other data is shared in a business environment (Excel!)

Why is Shufflepoint different?

Shufflepoint differs from the only other automate-into-Excel solution that I am aware of (Excellent Analytics, http://www.excellentanalytics.com/) in a few ways:

1. Shufflepoint is not free, although the $29 to $199 monthly pricing is not particularly expensive;
2. Shufflepoint does work for those of us using non-PC machines because it is browser-based SAAS at the core, not a true Excel plug-in;
3. Despite being a service, Shufflepoint is available as an installed instance, although a consulting arrangement with the company is required for this.

Additionally, Shufflepoint extends Google Analytics in interesting ways, including the ability to create queries with multiple profiles at the same time, good support for creating Google Gadgets which essentially allows anyone to create totally customized views of GA-collected data in iGoogle, and support for augmenting and moving GA-collected data in a variety of unique ways (for example, into Google Earth!)

Who will benefit from Shufflepoint?

Two main groups will benefit from spending a little money with Shufflepoint every month: any Google Analytics Authorized Consultant (GAAC) and any analyst who has standardized on Google Analytics and has ongoing report distribution needs. The benefit to the former is pretty obvious — if you can spend the time to build slick-looking dashboards and then couple them with Shufflepoint to quickly import different client’s data then the quality of your ongoing deliverables should go up along with your margins! The benefit to analysts is similar — the value of a standardized report template for distribution throughout the organization is not one that anyone should overlook — and if Shufflepoint can reduce a day’s worth of generating reports to less than an hour … well you can either spend your newly freed time reading some awesome free books on web analytics, contributing to the web analytics community, or, you know, doing analysis.

What you need to know about Shufflepoint?

I really like what these guys are doing, and not just because their very existence makes me look good having predicted their existence back in October, 2008 when Google first dropped the APIs on us. The emergence of “value-added” applications in Google’s analytics ecosystem, especially coming from companies who aren’t afraid to charge money on top of an otherwise free service because they are delivering incremental value, bodes well for the future.  In my consulting practice more and more very large companies seem to be considering moving away from incumbent paid solutions designed for widespread use and adoption in favor of the ease-of-use and simplicity they associate with Google Analytics. While this may be a case of the “grass being greener”, certainly the existence of Shufflepoint as a replacement for Omniture ReportBuilder and Webtrends REST make the consideration process more realistic (if not any more likely.)

Things we like about Shufflepoint:

• The fact that the app is platform agnostic.
• The fact that the app is really pretty easy to learn and use.
• The pricing is very easy to digest, even if you’re a cheap bastard like me

Things we’d like to see from Shufflepoint:

• Automation, automation, automation, automation … did I mention automation?
• Some type of Excel integration that would make launching the SAAS app even more straightforward.
• An expansion of their SQL-esque functionality, perhaps as integration into Microsoft Access

Learn more about Shufflepoint at their web site at http://www.shufflepoint.com.

Posted by Eric on Tuesday, January 26th, 2010 | 5 responses | Add a Comment | Share, Save or Email

Boston based Localytics is, in a nutshell, web analytics for mobile applications. Based on the projected growth of mobile applications and the likely surge of tablet computing this year, Web Analytics Demystified thinks this company is really onto something.

What does Localytics do?

Localytics is a digital measurement solution designed to provide metrics specific to mobile applications. The interface allows clients to view details about their apps such as usage patterns over time, user type (new vs. returning), user software (platform and OS), as well as user hardware and device versions. It can also track events within applications such as start, stop, play, etc. which are customizable to enable capture of the most important app functionality. The tracking also includes geo location capabilities to determine where apps are dispersed across the globe. All of these metrics are provided in real-time empowering clients with instant cognition of the use and utilization of their mobile apps.

The screenshot below shows the interface and you can view the short video introduction Localytics provides about its services. They also have an online product demo that is worth checking out.

How does Localytics work?

Localytics developed client libraries to integrate with the Localytics API which requires placing just three lines of javascript code into your mobile applications for basic monitoring (custom events require just one more line each). Client libraries (available for Android, BlackBerry and iPhone) were created using open source SDK for transparency, consistency and easy integration. The client libraries also provide flexibility for deeper integrations and customization. Once integrated with your app, the client library sends function calls (Required: open, upload, close. Optional: tagEvent, setOptIn / isOptIn.) to control your session. These anonymous and secure calls to the server can be uploaded upon start of the application (recommended) to place one single server call with batched information or can be configured to fire more or less frequently. Localytics data is written to persistent storage immediately after it’s recorded and is available within the user interface in real-time.

Why is Localytics different?

Localytics considers its primary competitors to be: Pinch Media / Flurry, Mobilytics and Medialets. It claims to differentiate from these solutions by offering segmentation capabilities that enable drill down on specific devices, geographies and other attributes beyond the aggregate level. It also offers real-time reporting that allows clients to identify changes in user behavior in a matter of hours rather than days. I particularly like that Localytics is fully transparent and has even developed a client wiki to offer information, FAQs, documentation and community aspects to their solution. This company was a member of theTechStars program in 2009 and was awarded seed funding to launch its operations, which are headquartered in Boston. Localytics was also recognized in 2009 by MITX (Massachusetts Innovation and Technology Exchange) as a finalist in the category of Analytics and Business Intelligence.

Who will benefit from Localytics?

The primary beneficiaries of this technology are the marketers who will be able to keep tabs on the use and utilization of their mobile apps. Not only can they watch what’s happening in real-time, but the segmentation and analysis capabilities (if wielded properly) can empower them to take action on this data. For example, marketers could test price sensitivity by watching app download activity when prices are adjusted – or – identify problematic areas of apps if users tend to drop off at specific areas. While, this technology may initially appeal to the app developer looking to understand their project adoption and utilization, a great deal of value is delivered to decision-making marketers.

What you need to know about Localytics?

Localytics is onto something pretty cool here. One source cites an estimated 2.3 billion mobile apps downloaded in 2009, which is expected to grow by 23% in the next five years. This burgeoning playground for developers needs a measurement solution and our initial view of Localytics offers an eye-catching and friendly user interface. We expect the major measurement vendors to come barreling down on mobile app measurement in the coming year – and for a burst of new entrants in 2009 – but Localytics has done a fair job of gaining a head start. Their launch price point was free, but this will undoubtedly change in the coming months. In any case, if you’re actively developing mobile apps or using them for marketing purposes, we recommend that you give Localytics a look.

Things we like about Localytics:

• The capacity for this solution to deliver metrics in real-time is awesome!
• The transparency of their client libraries and ability to modify the API rocks!
• The fact that Localytics built a client wiki is just plain thoughtful.

Things we’d like to see from Localytics:

• Users of this product would really benefit from alerting functionality.
• Ongoing expansion to other mobile platforms (Palm, Symbian).
• Thought leadership from the smart team at Localytics around this very hot topic of mobile measurement.

Learn more about Localytics by visiting their web site, http://www.localytics.com

Posted by John on Sunday, January 24th, 2010 | One responses | Add a Comment | Share, Save or Email

If you’re reading this hopefully you’ve already noticed we made some changes recently at Web Analytics Demystified. In addition to adding two new partners (Aurelie Pols and John Lovett) we over-hauled  the web site and have added a new blog (this one) called The Emerging Technology Blog!

In the next few weeks we are going to start reaching out to technology vendors — both those who are our consulting clients and those that are not — and asking them for 30 minute briefings on their most recent software or service releases. The output from these calls will be what we’re thinking of internally as “TechCrunch for Web Measurement Technology.”

We hope to use this new blog to surface great, new, and innovative technology for the entire community. We promise not to regurgitate press releases — trust me, if we start doing that I will just kill the blog — but given our global presence and experience as analysts and technology evaluators in a variety of situations we think  we can do a good job.

If you have a technology that you think is cool we’d love to hear from you! It can be any kind of technology that measures connected channels — web analytics, voice of customer, customer experience management, social media measurement, mobile analytics, virtual space measurement — you name it!

Finally, on behalf of all three Senior Partners at Web Analytics Demystified we want to thank Judah Phillips, Paul Holstein, Daniel, Joseph, and everyone else who has contributed to a blog here at Web Analytics Demystified. Especially Paul and Judah who wrote such great and compelling stuff — and Judah’s stuff is still here, at least for the time being!

Posted by Eric on Monday, January 18th, 2010 | 3 responses | Add a Comment | Share, Save or Email