For the first public project, we’ve loaded historical Major League Baseball stats compiled by player and team going back to 1871. There’s a ton of info there and baseball fans should have a good time compiling lists of players from different eras or computing formulae for their fantasy baseball team.

Here is a report I created showing the top 50 “Wild Things” in the history of baseball. Wild Things are pitchers who have hit a lot of batters in their career. I couldn’t believe that there was a major league pitcher who hit 14% of batters faced! (If the report is cut off, try viewing this blog post on its own page.)

So if baseball’s your thing, click through and check it out…. There are a bunch of sample reports, but I’d really love to see the kinds of things you can come up with. There’s a ton of data in there it would be great to see what the community can generate. After all, this site is called Collaborative Analytics, right?

… it is important to document the current measures, as well as those that have been discarded. Some examples of information about the measurements that could be tracked might include: description of the measure, how the measure is calculated, role responsible for measuring, information source, type of presentation, frequency, usage, etc.

On some level, this seems so basic as to not need explaining. But communication problems arise all time, even with people you see on a daily basis. I’ve had this happen to me recently, and it’s not a good feeling to realize that something’s messed up that could have been avoided with more communication and transparency.

Everyone knows this intuitively on some level, yet why are communication tools an afterthought to most BI tools on the market? To my mind, this is something that should be standard with every solution. But it’s not, and even more amazing, it’s not something that anyone is really asking for!

It’s great and even essential to be able to crunch a lot of numbers and integrate tons of data into a BI solution. And it’s essential to be able to derive some knowledge from this data to plan a course of action. But what comes next, what causes alignment? Communication! Getting the word out to the members of the company, so that each person can align their priorities with the overall goals.

What this means to me is that there is a large opportunity out there for collaborative analytics, and I’m excited about the upcoming full beta release of Good Data. The upcoming release has a lot of new features that incorporate collaboration and communication across metric and attribute creation in addition to finished reports and charts. Just a couple weeks away…

You can have the best software in the world, with the most sophisticated features, analytics and integration, blah blah blah - but if people don’t use it, it isn’t going to add value.

This quote appeared in an article on ReadWriteWeb by Jason Rothbart. The point of the article is that the most critical factor for enterprise software success and return-on-investment is effective user adoption. In a recent study, 70% of respondents said the most important factor for realizing value was effective user adoption, while only 1% said that the most important factor was software functionality.

How, then, do you create effective user adoption?

Certainly, software features do play a part in this. In order for someone to even consider using something, it must address a particular need. But chances are there is already a “standard” tool out there. Features certainly help to differentiate, but how often is there something truly new and innovative, versus marketing spin?

Ease of use, I think, is another thing that helps user adoption. If new people can get on board and begin using the software intuitively, they will likely stick with this solution. Complexity out of the box is a sure fire way to limit adoption.

Price could be another factor. It may not necessarily help user adoption, but if the price is low enough, and the software does something particularly well, then ROI could be achieved.

But in my opinion, the key factor to getting user adoption is collaboration.

In this day and age, I believe that everyone creating software applications should incorporate some kind of collaboration. People can create things, share things, make their voice heard, correct wrong assumptions and enter into meaningful discussion with their colleagues. Collaborative features are empowering. Collaboration is viral. If done properly, people not participating can miss out, or feel left out. And then you have a very human reason for adopting.

What do you think? What are the forces that really drive user adoption of software applications?

I’d like to continue from here and talk a little about why collaboration is a good thing.

Good Data aims to provide two main things: visibility and alignment. Visibility is the act of analyzing data and extracting answers to relevant business questions. These answers can then determine a course of action based on facts. By continuing to monitor the data, a company can benchmark its progress and (hopefully) improve performance. By and large, this is the aim and the promise of all BI software vendors worth considering, and the manner of doing it (MOLAP, ROLAP, HOLAP, blah, blah, blah), while important, is really just a way to answer “how” to provide visibility. Alignment, on the other hand, deals with having all stakeholders working effectively as a team. This means that all stakeholders should be able to see the past performance, agree about the present, and create and work toward common goals for the future.

Collaboration is a special ingredient that can greatly enhance the effectiveness of both visibility and alignment.

Visibility
Similar to Wikipedia entries, Good Data users can collaborate on the report definition process. Any perceived errors can be corrected by an appropriate person, and the history is available so that others can see the progress and even contribute to the final report. This adds a layer of validation to the answers. Unlike individual data silos, everything is centrally located and available for all to see. Further, other users can leave comments and annotations on the reports for additional clarification or alternate interpretation.

Alignment
The result is that stakeholders are empowered to be active participants, whether by creating reports or by offering interpretation. If an instinct seems contradictory, they have the means to participate and verify the report. In this way, results are disseminated throughout the organization, consensus is ultimately achieved, benchmarks are set, and future goals can be created and effectively measured.

So… I invite everyone interested in visibility and alignment to check out the Good Data demo. In just a few short weeks, we will launch a newer beta version that will allow you to upload your own data project and collaborate with your colleagues. Stay with us… it’s coming!

To finish, I thought I would post this picture. To me, this picture shows that when you have visibility and alignment, it can be a beautiful thing.

Georgetown on the Potomac

Microsoft has an edge in that the most popular and widely used BI tool is Excel. Leveraging this, they are building plugins into Excel that lets users retrieve information from multiple sources and create pivot tables within Excel. When finished, you can then publish results to SharePoint. This is Microsoft’s version of collaboration. Doug Henschen wrote a good description of Project Gemini here.

On the surface, this seems good, and certainly plays into Microsoft’s strengths of owning the desktop. But is this really any closer to the elusive “one version of the truth” or just compounding the problem of “Excel hell”? The Galloping Data Architect, Michael Horne, writes a well-thought piece about how this solution is more the latter. He argues that giving everyone access to the source data and the ability to bring this data onto their desktops via Excel does not unify disparate data silos. In fact, it makes it easier to create them (and in so doing, creates a nightmare for IT).

Everyone having the ability to perform their own queries in a product they use regularly certainly solves accessibility. But without a collaborative, or shared, workspace, there are no checks and balances. This can also lead to lots of redundancy, as people build things that have already been created. And what if their results are different? This can easily happen as part of personal bias in the report building and selective disclosure of information. We’re back to multiple versions of the truth. Here’s a real world example of how working in isolation with Excel can have painful consequences.

To me, this all sounds like more of the same. I think that the real innovations are coming from startups who can think out the box and not be tied to the success of an existing product or platform, and therefore have nothing to lose by being creative.

And speaking of “one version,” do you find it funny that Microsoft’s code name for this is Gemini (twins)?

I view working for a startup as making a bet. You are betting that this startup has what it takes to be successful. At the core, you are betting on the company’s ability to innovate. In times like this, your innovations can be your only shield against the hostile environment.

Now is the time for focus. Not only in financial matters, but across the entire business. For all startups, it’s necessary now to focus on the components of your business that you believe in, that you are proud of, and that make you different. Now is not the time to be pitching another instant messenger client, web-based email service, or some other copycat service.

At Good Data, I believe we are bringing something new to BI. I believe in our SaaS model. And I believe that this environment may even be a boon to increase the rate of adoption for SaaS applications. Sure the odds are long, but I knew that already. I was sold on the vision before I started, and nothing has changed. I guess that for me, things just aren’t that bad.

So for what it’s worth, my advice is to not be overwhelmed, keep innovating, and stay focused on the prize.

How is the downturn affecting you?

The author points out that SaaS is not ASP 2.0:

The notion that SaaS is taking us back to the days of dumb terminals is simply ridiculous. Have they not heard about AJAX and all the other rich client stuff that actually uses PC cycles to enhance the user experience?

The problem I have with statements like this is that the tone seems to presume that SaaS is a means in and of itself. It is a “holier than thou” attitude that implies that by virtue of being SaaS, you are already a half step ahead of people that aren’t. This simply isn’t true.

SaaS is simply a delivery method. This delivery method has some inherent advantages, such as being scalable (running in the cloud), automatically updated (by the SaaS provider), cheaper upfront (no hardware costs), etc. There are many arguments in favor of SaaS, and for the most part, they are all true. But these reasons are not why companies buy software. 

For any software firm to be successful, the firm must sell solutions that directly address the buyer’s pain. I read a lot of articles about SaaS and this point always seems to be overlooked. Claims are made that Monolithic Company X will soon be usurped by nimble SaaS Company Y in some huge market. What seems implied is that Company X and Company Y are identical, except for the delivery model. This is simply not true.

While SaaS vendors have some advantages as mentioned above, there are also some huge obstacles. Creating an application that is up to par, or better, than the incumbent and being able to deliver it through the browser is no simple task. There are bound to be some tradeoffs. The trick is creating enough value (ie, relieving enough of the customer’s pain) to make the customer consider the solution in the first place. Only then does the different delivery model come into play. Without the steak, the nice platter doesn’t matter. 

The challenge for all SaaS firms is the same as with traditional vendors: create a compelling product that people will use. This is done by understanding the customer, understanding the problem, and non-intrusively introducing a component in the customer’s workflow to make the job easier. Here, incumbents have a clear advantage in that they already have users locked in to some extent. Then, after solving the basic need, you can introduce features that enhance the experience and increase the productivity of the users. It is here where most SaaS companies have an advantage through innovation and rapid adoption of the latest technologies.

To finish, I just wanted to end with by sharing a comment that was posted on the ReadWriteWeb article. It was left by Jan Erik Paulsen, who doesn’t seem worried about the SaaS threat (at least for now) and in my mind helped spell out the challenge I tried to describe in a humorous way:

Let’s be honest for a minute.

[Big, traditional software vendors] pay the bills while some Web 2.0 guy talks about his mad AJAX skills. Yeah, because those 79 lines of Javascript makes him stand out in the marketplace…

Scared? Why ? Oracle and similar companies are paying the bills. Basecamp is just crashing the party and stealing beer from the bathtub.

Basecamp is a hotdog stand. Great hotdogs though.

What do you think? Which SaaS companies do you see out there that have what it takes to unseat an incumbent?

1. 90% of BI will be delivered on mobile devices.

I think the iPhone is really forcing the mobile industry to create products that are actually usable for internet content. It won’t be long before mobile devices are mainstream methods for getting content, and BI should be no exception. 

How does this affect Good Data? Well, just a couple nights ago Jakub showed me Good Data working on his iPhone. Good stuff! Notification services (via RSS) are a no-brainer. I’d say we’re well positioned here

3. Information will be managed by business professionals and not technology professionals.

This is what it’s all about, isn’t it? This could have been a theme for BI for the past 20 years. This is something we are definitely addressing, and I like our prospects here.

4. Do not listen to BI best practices that say “18 months to deliver” and “high costs for licenses/services”. We need to think out of the box.

This is where SaaS excels. With Good Data, you will be able to upload CSV files into the system and begin creating reports immediately. 100 MB will be free, so there is no risk in creating a proof of concept. We’ve nailed this one, and you can try it in November.

6. Better adoption of BI will happen when BI is integrated within productivity suites, like Office. BI will be an extension of business productivity.

I wholeheartedly agree with this one. Why should anyone have to go to an unfamiliar application or be trained to use something that is not a part of their workflow? You can argue that it should be, but I’ll use the example of airline check-in desks again. There is no added value in having a trained team of people who know that when you check someone in, you should type *xQydMJKL*, press “Enter” five times, and then some other arcane nonsense. This is similar to BI tools today.

We are building Good Data in such a way that BI results can be taken away from the application and embedded into the places where users want it. Inside other applications, on corporate wikis or blogs, anywhere. This is very exciting and key to making BI accessible to wider groups.

9. SaaS BI’s success will be based on two serious roadblocks: Data Quality and Data Security.

This point is interesting because he directly addresses SaaS. In my mind, all the points apply, and all points prove the value of SaaS. But I don’t feel that these are really “roadblocks.” Certainly, these are issues to be dealt with. But Data Quality is not a problem specific to SaaS, but rather to BI in general. As for Data Security, this isn’t really my domain, but I feel this can largely be overcome.

For an SaaS company, security has to be the name of the game. Without proper security, there is no business. In many cases, my feeling is that SaaS companies worry more about security than many firms who keep data in-house, thinking that keeping it in-house is enough. A non-software example I like to think of is bank. This is probably one of the most sensitive issues for most people: their money. Yet not many hoard cash. And even in the face of today’s financial collapse, most still keep their money there! Now is the time to be questioning bank security, yet for many, the bank is still the most secure option. SaaS companies should be hiring these bankers for the marketing department!

I was disappointed to see that collaboration and social networking only received a passing glance. But I guess that just means that we have our work cut out for us, and it will be up to us to prove the value in collaboration. That’s ok with me.

What do you think about the future of BI?

In remember when I got my first Web job in 1995. 28.8 modems were a luxury. At that time, many of the biggest websites (read search engines) were hosted at universities and didn’t even have their own domain name. The Web was hardly “graphical,” and multimedia meant a bunch of JPGs. And when Real launched (nee RealAudio) people could embed MIDI on their pages. (And they did.)

Over time, we know what happened. More and more information became online and accessible, broadband became commonplace, and pretty soon, we have Web 2.0 everywhere, allowing people to interact with one another (Facebook), build content together (Wikipedia), influence purchasing decisions (Amazon), etc. In essence, people have a platform to be heard, and they are using it! And the open-ness of the platform is allowing them (us) to build amazingly useful things.

Is there anything we can learn from this evolution and apply to enterprise software? I believe there is, and that SaaS BI, with the Web as it’s base, is in a unique position to deliver.

The company I work for, Good Data, is building a collaborative SaaS BI solution. Since the application will be hosted and delivered directly through the browser, why not allow all the things that make the modern Web so great?

The application allows for Wikipedia-like report creation allowing people to extend analysis of other people’s reports with built-in versioning. People can comment on reports, create annotations on particular values, share reports, tag reports, subscribe to reports and so on. As with the Web, the more participation, the more valuable this becomes. Not only will the output be shared across an organization, but it’s interpretation can be validated by the multiple perspectives of the participants.

This area is really exciting for me, as I can think of a lot of applications of collaborative analytics. And I hope to write about this theme often in more well-thought out posts in the future. But I wanted to get the ball rolling and pose the question to the wider community:

Are you a believer? How can collaborative BI help you?

Ken points out that the core problem of BI is not about technology, complexity, level of commitment, and all the other typical problems blamed for lack of BI success. His argument is that the core problem of BI is that people don’t know what questions to ask. In other words, people can have the best tools, but if they don’t know what they are doing, the result will be worthless.

To this end, LucidEra offers a free pipeline health check, where they take your data and “apply generic best practices analyses” to produce insightful results. But while this is a powerful offering that I’m sure is helping people, I have a couple problems with his core premises.

First, while I am sure best practices exist, I don’t believe that they are the same for all companies. My feeling is that this application of best practices is a very limiting and narrow world view. If I am running a company, I don’t want to be told what to do, particularly if the company is successful. Do companies searching for BI solutions really believe that their BI application provider is the best entity to evaluate their business?

This leads to my second problem, which is that this approach seems to have a clear conflict of interest. It seems to me that LucidEra is offering two distinct products: business consulting (best practices analysis) and an SaaS BI tool for OLAP and ad hoc query. With both of these on the table, how is one affected by the other? Is the best practice analysis limited to what is possible with the tool?

If I am running a company, and required business consultants, I would want consultants who are laser-focused on increasing efficiency and promoting productivity. I would not want them to be tied down or have any limitations related to a particular software package. It seems to me that the consulting and the tools must be decoupled.

What do you think? Is business consulting a part of business intelligence?

The main goal of this blog is to discuss business intelligence issues, particularly around Software as a Service (SaaS) BI. I started this blog to post my ramblings about the subject, and hopefully generate some good discussion around key topics.

I believe this space is undergoing significant progress. The available technology today is allowing things that simply couldn’t be done as recently as three or four years ago, and the influx of bright startups in this space is bringing lots of new innovation to address current BI limitations.

My goal for this blog is to highlight things that I notice in the space. People who know me will say I’m opinionated, some will say I can be funny, others will think I’m an asshole. No doubt, some of all will be present here. I hope you will find it useful, and I hope you will be active in supporting the discussion with your thoughts. Even better, I’d love to hear your dissent. Together, I hope we can gain insight and move forward.

After all, isn’t that the point of collaborative analytics?