Archive for November, 2011
by Bill Ives
November 28, 2011 at 4:36 am · Filed under
Web 2.0
This is the second part of my interview with IBM’s Mike Rhodin, SVP, Software Solutions, about their moves in the software space. We continued our discussion by covering IBM’s smarter commerce initiative. This is an emerging market where the idea of “the educated consumer” has reached a whole new level with explosion of user generated content within Web 2.0.
Mike said that while they had sound offerings in the order capture area with WebSphere Commerce and clients like Macy’s and Target, they wanted to expand to cover the entire B2C lifecycle. He mentioned their acquisition of Sterling Commerce for order capture, order management, shipping and logistics. To provide supply chain optimization they acquired iLog. They also acquired Coremetrics for Web analytics of order tracking and Unica for automated marketing. Unica looks at what the next best action should be taken to drive sales as people explore a commercial web site. It can also be used in physical stores at point of sale terminals. On the analytics side of this process lifecycle, Netezza was acquired to enable better visualization of data.
We went on to discuss what’s next. A big move that IBM is taking involves a transformational look at the use of computers with Watson. The traditional use of computers was to standardize things and allow the machine to consistently provide the same answer every time for the same transaction. Watson is a system that you teach rather than program so its answers continue to evolve as it learns. It will get better but it can also make mistakes. You teach it through a Q&A process. When Watson takes in new data there is a seven fold expansion in its memory as it is making connections between the new data and the old data.
This approach is much like more modern views of human cognition. The old behaviorist view modeled, in part after machines, was a simple stimulus response. You filled the brain with new data and it responded. Piaget introduced the concept of assimilation and accommodation. You assimilate new information based on your existing cognitive structures. You also accommodate these structures based on the new information. There is a give and take process going on as you learn.
IBM is partnering with WellPoint for the first commercial application of Watson to help with medical diagnosis. They system wisely does not tell doctors what to do but offers them the top five probabilities for a diagnosis based on what is known about the patient. It does not take over the doctor’s job but acts as a doctor’s assistant. This use is designed to both increase proper patient healthcare and reduce the cost of care by reducing unnecessary tests and procedures. This makes a lot of sense to me.
IBM’s software business has become one of the fastest-growing contributors to IBM’s profit. In the last decade, IBM Software profit has tripled, driving more than $9 billion in profits in 2010. With coordinated moves like those Mike described, I can see why this success is occurring. As Mike noted in his blog post, Beyond Collaboration: Becoming a Social Business, “it’s one thing to create networks of customers, employees, or partners — it’s quite another to change the processes that run your business, to make them truly social.” It is this change that is required to put social software to work and move it from a useful utility to the core of a new way of doing business.
by Bill Ives
November 21, 2011 at 4:18 am · Filed under
Web 2.0
Over the last decade IBM has driven a significant transformation of its business model as they shifted to higher value areas, improved efficiencies, and invested in a number of firms to fill out their portfolio of software offerings. They have acquired over 80 software firms since 2000. More recently, they are focusing a significant portion of their efforts on social business as I wrote about in my coverage of Lotusphere 2011. IBM’s most recent earning reports indicate a strong growth in demand for their Social Business offerings.
Recently, I spoke with Mike Rhodin, SVP, Software Solutions about their moves in the software space. We also discussed the convergence of social business and analytics. We had a wide ranging discussion that I will cover in two posts, with the next one appearing tomorrow.
Mike began by diagramming the interactions of their efforts in analytics, social software, and smarter commerce. At the core is a three-way overlap where web analytics serves all three areas. At the intersection of analytics and social is sentiment analysis and content analytics such as that performed by Watson. At the intersection of social and smarter commerce is the creation of brand-aligned communities. Mike gave Moose Jaw outdoor products as an example where people are passionate about the brand and this passion is the foundation for an online community. Analytics will intersect here to offer a useful view of what is happening within the community.
We next focused on analytics in more detail. IBM projects $16 billion in business analytics revenue by 2015. Business analytics software and services is a key contributor to IBM’s growth with revenue up 20 percent in the first half of the year. In the last 5 years, IBM has invested more than $14 billion in 24 analytics-related acquisitions. They have also dedicated nearly 8,000 business consultants worldwide to this space.
The Cognos and SPSS acquisitions helped extend their base for predictive analytics. In the predictive analytics space they are initially looking into two areas: financial performance management and risk. They augmented the capabilities within Cognos last year with the acquisition of Clarity to support automated regulatory filing. In the risk space they added OpenPages and integrated it with IBM’s Tivoli brand. They now have eDiscovery capabilities through FileNet and support for legal hold requirements through PSS Systems. Another acquisition, Algorithgmics, also provides deep analytics in the area of financial risk.
We next stepped back a bit to look at what Web 1.0 and Web 2.0 brought to business. Web 1.0 demonstrated that the Web was ready for business. It showed that open standards could work and provided a language for the Web with Java. SOA was created and this supported the modernization of back office functions.
Now Web 2.0 took this further by bringing modernization into the front office. You now have Web apps that look at workforce automation and human capital management. For example, one internal IBM tool allows people to bid on work within their internal market allowing for greater utilization and a better fit of skills with job requirements.
Social tools are being adapted to created the connected enterprise and provide a useful and work related knowledge Web unlike the old style disconnected versions of knowledge management. I certainly agree with this need for connecting knowledge with work. The only knowledge management systems that I saw that were successful from the early 90s on were those integrated into work process. This was harder to do with Web 1.0 tools and so much easier with Web 2.0
To properly leverage the capabilities to handle the front office that Web 2.0 brings, IBM has built a multi-functional social business platform with Connections. They have gone beyond social networking to social business. Here is a blog post that Mike wrote on the topic, Beyond Collaboration: Becoming a Social Business. As I saw at Lotusphere, the capabilities of Connections are being integrated into other IBM software offerings to make them more social.
They are also developing situational workflow offerings, another good move. Two acquisitions in this space are PureEdge for secure XML forms and Lombardi for department level business process management software. IBM is working on rounding out its software capabilities to cover many business capabilities. They are also building an app store provisioning model to provide small business applets for the mobile business market.
We next went on to discuss their efforts in smarter commerce, as well as their next generation of computing, Watson. I will pick up these discussions in part two of this interview.
by Bill Ives
November 14, 2011 at 4:28 am · Filed under
Web 2.0
There is a great rush to move applications to the cloud and enable them for delivery on mobile devices. One of the critical tasks to make this move work is the provision of efficient performance testing. Mu Dynamics has stepped into this essential market with Blitz. I recently spoke with Ajit Sancheti, Co-Founder at Mu, about their move.
Mu began six years ago with a focus on security for network infrastructure. They allowed firms to look at the impact of apps on networks, including the rush of iPhone apps and the extensive use of apps like Twitter. They then realized that their core solution would fit well for performance testing in the cloud and mobile apps where there was a lack of this capability. So they built an extension of their core platform to cover this market.
Blitz is a performance testing solution specifically designed for mobile, cloud, and web application developers. They offer usage-based, tiered, low-cost pricing that makes it accessible for both large and small development organizations and platform providers. You can quickly do load testing and you only pay what you use. Developers can instantly load test their APIs and cloud apps without creating complex test scripts. With the Google Chrome extension and integration with Google Analytics, Blitz brings load and performance testing into the development cycle, enabling developers to simply launch a load test of their application wherever they want. It can also be integrated into the application development workflow.
They have integrated with Acquia to provide support for Drupal application development. This allows Druplal developers to run load tests in a matter of minutes, ensuring their sites can handle traffic spikes before they occur. For my most recent review of Acquia see – Acquia Provides Drupal Commons to Support Open Source Enterprise Collaboration). Mu has also integrated Blitz with CloudFlare, a web site development platform. This allows CloudFlare developers to measure the baseline performance of their web applications with quantifiable metrics.
Ajit said that the simplicity of Blitz allows developers to do their own performance testing without having to rely on testing experts. Most of the process is automated and you can access Blitz in a self-service mode through the cloud. While they are constantly adding new features, the goal is to keep it simple at the same time. I think this is a wise move as too many applications add complexity with new capabilities and lose their original attraction.
Ajit said they looking to extend load testing to performance monitoring. They also plan to offer and enterprise version, as well as an offering for private cloud use. Testing is generally not something that makes the front pages but it is critical for application success and simplifying it can allow developers to focus on enhancing their apps in an efficient way. Mu is providing a great service to the market with this offering.
by Bill Ives
November 3, 2011 at 3:44 am · Filed under
Web 2.0
Traditional BI software can be difficult to use, is often expensive to maintain, and can force users through pre-determined thought paths. QlikTech is taking an alternative approach. I recently spoke with Erica Driver, Senior Director of Global Product Marketing at QlikTech. She said that QlikTech turns this traditional BI approach upside down, putting data where users can explore it, share it, and combine it at will to find new opportunities and make better decisions. Instead of pushing insights from the top down, QlikTech’s software enables users to discover insights on their own, anywhere in the organization, as well as anywhere outside it.
Erica talked about business discovery as a new BI approach that puts business users in total control. It enables them to create tailored insights that meet their individual needs and timelines so they can solve specific business problems. It provides instant answers, and also delivers live data and analysis on any mobile device, so users know what’s happening in real-time.
Erica noted that we make decisions based on multiple sources of insight: data, people, and places. QlikTech has made investments in each of these three areas. For data they offer an enterprise platform with rapid analytics and comparative analysis. For the people component that have now added social business features and they have developed mobile capabilities to address the location aspect.
Before we got into discussing the five themes Erica walked me through some QlikView basics focusing on the associative experience. QlikView maintains all of your relevant data and the associations within in it so you can drill down any way you want to depending on the questions you are asking. For, example you can bring down sales data by regions, products, customers, sales reps and other factors, as well as combinations of these. For example you can see sales data by customers and sales reps on a global basis as shown below.
Then you can drill down all the way to an individual customer in Europe as shown below. The drill down and aggregations are done on the fly. You cold also explore individual sales reps, for example.
QuikVew has also added social capabilities so you can add a chat session directly into the report. Below you see a chat session added to the data in the first image. This allows the conversation to take place where the data is located. As I have written many times, social capabilities need to be integrated into work processes to have a lasting impact (see Putting Social Media to Work). This is a great example.
You can then always go back to see the data as shown when the conversation took place. This view is not a screen shot so you can also see the same data filters over time. The social capabilities also allow you to do real time sharing for increased collaboration. Erica showed a demo where she and I both had real time access to the same data and took turns filtering it. This was quite powerful as it increased the speed of our exchanges. We did not have to toss control back and forth as we both had it at the same time.
In addition to building collaboration capabilities into QlikView, the ability to integrate with other enterprise collaboration tools is also offered so you take QlikView to where people are working such as within SharePoint or Salesforce Chatter.
QlikView also provides comparative analysis. You can create your own groupings and then make comparisons as shown below. In this case two food product groups are being compared to see patterns and relationships.
QlikView looks at structured company data like the examples we have just seen but it can also look at unstructured content like the Twitter example below.
There is also support for many mobile devices and a version that is optimized for smart phones. Their rapid analytics platform allows you to build custom reports with selected subsets of the your data. Their enterprise platform capabilities give business users control over what they do while also giving IT the ability to maintain overall control for such issues as security, scalability and reliability.
I was very impressive with the comprehensive set of features and their accessibility. This could be a breakthrough offering in the business discovery space. Gartner has placed QlikView in its upper right quadrant for business intelligence as an industry leader with the ability to execute. Gartner also noted that “ease of use” now surpasses “functionality” for the first time as the dominant BI platform buying criterion. They also note a new group of data discovery tools led by small vendors. This is all to QlikView’s advantage.
by Bill Ives
November 3, 2011 at 3:11 am · Filed under
Web 2.0
I have written about Traackr before on this blog (see Traackr Introduces αlpha‑Lists). As I wrote then, I have been using Traackr for several months. For those who want to better understand the influencers in a field, Traackr provides a comprehensive, yet very easy to use, tool set that gives you a list of the top 25 in your topic of interest. As a disclosure I have no connection with Traackr but they have given me limited free use of their tool so I can better understand it.
I recently spoke again with Courtney Vaught from Traackr to see their excellent new analytic tools. They provide some useful visualizations with this new capability, so there will be more than the usual number of screen shots in this post. For many of the analytics examples we used the Traackr campaign I set up for Darwin Ecosystem on influencers related to keywords relevant to our Awareness Engine™. We began by looking at the location visualization (see below). This shows the states within the US where your top 25 influencers are located. You can drill down within each state to see their location within the state.
Next we looked at the distribution of key words used by all of the influencers to see what is prevalent in their conversations (see below). Creating the original list helps you discover who these influencers are. This keyword analytics feature and some of the other following tools helps you listen more closely to what they are writing. The original top 25 lists, along with the analytics pieces are updated weekly. These keywords not only give you a good sense on the topics of interest, but can also help you adjust the keywords you use to derive the top 25 list. We can see which ones are being used and which ones are not used. The results on our list made sense to me and were useful.
Next we look at the scores table. Traackr uses three measures to determine its top 25. Reach is the ability to generate views based on audience size. Resonance is the ability to spark conversations based on such measures as retweets, comments, and links. These first two are actually independent of your keywords. Then relevance is the ability to cover specific topics based on relevance to your selected keywords. As you can see below you find reach and resonance on one axis and relevance on the other. This grouping makes sense and I find the results helpful. In our case the scores are stronger for reach and resonance and less for relevance. This seems appropriate as we are trying to create a new market with our Awareness Engine.
Lastly we looked at sentiment. Here we moved away from Darwin examples as it takes some elapsed time to set up and run a sentiment analysis because of the factors involved and we had not started one. The actual user time is short, it just takes time for the process to run. Sentiment analysis can be marked in the monitor tool and further broken down and reviewed in the analytics suite.
Sentiment is determined by users as they manually rate each piece of content by the influencers as: positive, negative or neutral. I actually agree with this manual approach for two reasons. The data set is small and the automated sentiment tools are still quiet primitive and often return “unknown” as a result. Below is a screen shot of sentiment within analytics where you see the percentage of positive (red), negative (black), and neutral (grey) sentiment.
Sentiment can also be found in the Traackr monitoring tool. Here you can see the absolute number of positive, negative, and neutral content instances. See the example below for the same data as the analytics visualization above.
Monitoring gives you much more functionality than this sentiment reporting. For example, you can put in up to five keywords and monitor their use by your top influencers. These new keywords do not have to be, nor should they be the ones chosen from your original list used to determine your influencers. For example, if your list is for hybrid cars you can monitor specific cars as shown in the example below. You can also select a subset of your influencers as also shown below.
I really like the analytics features and look forward to using them. I will also make greater use of the monitor feature. Traackr provides a nice balance of ease of use and useful features. It is great tool if you are trying to discover and understand the key influencers on your topics of interest.
Thought...
