How Agile is Your Business Intelligence?

[This is something of a 'live-blog'. Yesterday I had the chance to watch a webcast on 'Nimble Business Intelligence' and I jotted down some of my favorite highlights from the presentation.]

You know 'BI', right? In a recent report Gartner described Business Intelligence (aka BI) as software that delivers 13 capabilities: Reporting, Dashboards, Ad hoc query, Microsoft Office integration, Search-based BI, BI infrastructure, Metadata management, Development, Workflow and collaboration, OLAP, Advanced visualization, Predictive modeling and data mining, Scorecards. This dirty dozen featureset certainly covers a lot of functionality and all of it is for 'decision-support'.

But considering the high failure rate of traditional BI projects, maybe BI alone isn't enough. In a webcast titled 'Nimble BI: Enterprise Mashups for Agile Intelligence', BI expert Seth Grimes addressed what he thinks is an under-served area of decision-support: agile or 'nimble' intelligence that supports the more dynamic decision-makers. And like the title of the event says, enterprise mashups are the way to get your agile intelligence.

Early in his presentation Seth showed a bunch of 'composed' dashboards from popular BI vendors, created by one group for consumption by another group. He emphasized the word 'composed'. And then he emphasized it again. There was very little about them that were 'nimble', he said.

'Flexibility, self-service, speed-to-insight'. These are the fundamental elements of ‘Nimble Business Intelligence’, according to Seth. Given the reasons for BI project failures, these qualities would certainly help avoid some of those issues. One of the audience members seemed to violently agree when she called self-service the 'Holy Grail', lamenting that 'adding a column to a ResultSet for a BI app requires an entire spin through the waterfall SDLC'.

Seth emphasized that self-service means 'dynamic data access', 'visual' tools’ and ‘no ETL’, but also pointed out that the sources for these self-service solutions should be the same ones that feed the BI solutions you have today. Of course, I wholeheartedly agree that 'information agility' goes hand-in-glove with a self-service solution, like my buddy John Crupi recently wrote. Although Seth didn't say it, I expect one of the reasons for this is the ubiquitous IT backlog. 'Backlog' and 'agility' are archenemies, it seems.

Of course, Nimble BI isn't a panacea for all that ails BI. One savvy audience member asked about data quality and how Enterprise Mashups might fix or exacerbate that issue. Truth be told, Seth somewhat agreed, pointing out that 'bad data' certainly gets more common when you use open sources from the Internet. But mashups, he said, could be used to 'explore your data quality issues' and ultimately improve your data quality.

Toward the end, Seth gave a recap of what he believes enterprise mashups deliver, value-wise. I'll quote him verbatim, since I don't think I could say it any better:

Seth then concluded with a very important point: Enterprise Mashups do not replace BI solutions. They sit side-by-side, complimenting each other. Certainly, Enterprise Mashups can consume BI services and make part of mashup solutions. And the opposite is also true.

In the end, I left feeling that Enterprise Mashups perhaps deserve a place in the BI practioner's recipe book. The recording and slides from Seth's entire presentation are available on our Event archive. Seth has also written a bit of this topic, with a whitepaper on 'Nimble Intelligence: Enterprise BI Mashup Best Practices' and a blog asking 'Is It Time For NoETL?'.

Take a look, start thinking nimble.

Read More...

Summary only...

Have a problem? There's a mashup app for that.

You know 'apps', right? Those easy-to-pickup, easy-to-use, easy-to-share mini-applications that you can get through your phone, your portal and even your desktop.

I am a big user and an even bigger fan of apps. So I was quite excited to meet my first app creator last week. By day, Jon Conway is a consultant and a member of our Mashup Developer Community. By night, Jon creates apps for the iPhone, like iPotato and iPoke the Hibernating Bear.

I knew how to consume apps, how they were distributed and socialized. But after I chucked a few iPotatos I realized I had finally a representative of the ‘other half’ of the lifecycle of an app. Now I appreciated where they came from.

That's when a post from Dion Hinchcliffe, master of all things '2.0', really came alive for me. Dion believes the app store is destined to move into the enterprise. He recently wrote about the opportunities and issues associated with the application of the app store model in the enterprise. Here's the summary, in his own words...

The premise of an app store model for enterprises is simple: By removing the middleman, the famous bottleneck between the business and IT demand can be reduced in many cases. Application backlogs can shrink, consumption of internal and external IT resources will increase, and fierce competition to provide the best solutions to niches can greatly improve overall quality (the long tail of IT argument), all while reducing costs.

In case you’ve missed the whole app phenomenon until today, the most important part of the app isn't the app itself. It's the place where app users (like me) and app makers (like Jon) exchange these magical nuggets of productivity, entertainment and utility. The folks who do the marketing call these places 'App Stores'.

They are simple online marketplaces that make it easy to find and select/download (and in some cases pay for) the app. Blackberry, Nokia, Microsoft and Palm all have app stores for their mobile devices. Google just launched their Apps Marketplace, presumably for browser apps. The Apple App Store is the most prolific of the bunch, with over 100,000+ apps and over3,000,000,000 app downloads. Yes, 3 billion. That certainly sounds like a model that works.

But the model is somewhat different in the enterprise. Where will the apps in the enterprise app store come from? The title of Dion’s blog gives the answer: ‘The Enterprise App Store And Self-Service IT: How SOA, Saas, And Mashups Will Thrive’. ‘Self-Service IT’ is shorthand for ‘they get created by the people who want and need them’. In other words, they won't necessarily come from the ‘software people’ or ‘IT’.

Unlike the consumer world, the apps makers and the app consumers will be the same people. They’ll get created in visual mashup-making tools by your techno-savvy business users (and their line-of-business developers, perhaps), for themselves, and then shared with others like them.

Let me give you a working example of this enterprise app store thing. JackBe has begun a ‘Mashup App Store’ experiment with the launch of our Presto Cloud (Community Edition). The premise is simple: Presto Cloud (Community Edition) is a shared version of our Presto Enterprise Mashup Platform (on Amazon’s EC2) that lets users create, share, and reuse mashup apps with each other.

And our experiment would be incomplete without a ‘Mashup App Store’ showroom for the folks who only want to browse/use the mashup apps, not make them. So we’ve create a ‘Mashup Wall’ in our Mashup Developer Community where you can browse, comment, rate (and, yes, download, if you want) the mashup apps created by the users of the Presto Community Edition.

We're going to treat this as a ‘perpetual beta' and all of this is simply ‘Mashup App Store 1.0’. There’s more to come. Try it out and share your feedback in our Community Forums.

Get your mashup app on!

Read More...

Summary only...

Does Your Enterprise Suffer from the Stalactite Effect?

A recent survey by Accenture of 1,009 Fortune 500 managers found some truly surprising results.

That’s not a great track record. But WHY? We spend $20 billion a year on databases, reporting tools, marts and warehouses, cubes and OLAP and much, much more. After 4+ decades of information technology advancements, how can we be so wrong so often?

Any honest 'expert' (if there can be such a thing on this topic) would probably admit that we are less-than-good for many reasons. I'd like to add an often-overlooked candidate to the list. It’s something I call the 'stalactite effect'. We all know what stalactites are, right? Those long pointy things that hang from the roofs of caves, formed as calcium carbonate-laden water drips over them. On any given day, you'd see no change. But given enough time they can become massive. (The biggest in the world is thought to be a 27-foot monster in the Jeita Grotto in Lebanon.)

In the enterprise, your water drops are your documents, like budget spreadsheets and project plans. Any one document/drop isn't likely to be a huge loss information-wise. But take what you create in a year, add what your nearest cube-mate creates in a year, maybe what the other 17 analysts in your department create in a year and you start to see wonderful stalactite of information. And it's all completely inaccessible to anyone but you. Enterprises like this literally exist in a data cave with lots of drip, drip, drip.

Is the stalactite problem something we just gotta learn to live with? Certainly not! Some very interesting approaches to these ‘micro-silos’ have emerged recently. For instance the Excel Services in the upcoming Microsoft SharePoint Server 2010 allow you to use and publish Microsoft Excel client workbooks on the SharePoint Server. And on top of this they’re adding features that provide multiple users with the ability to edit any workbook simultaneously; a ‘Slicer’ feature that is a new type of interactive GUI filter; and a REST API that is a client server software architecture/protocol that defines workbook and how to access them.

Of course, Enterprise Mashups can also help. JackBe has tied spreadsheets directly to our enterprise mashup server through our Excel Connector and Server API. A user can make a workbook into a mashable service right from the spreadsheet itself. The user can also refresh the mashable service at any time. And the reverse is also true, where a user can create a workbook that is directly based upon a live mashups. This is a great roundtrip solution. Want to see it in action? Check out the recording of our 'Mashing Microsoft' webcast. During the webcast we mashed Excel (as well as Project files and a few other things) in real-world scenarios like project management and financial analysis.

This approach to the stalactite problem can be perfect if an organization is losing valuable data on a drip-by-drip document-by-document basis. The document owner contributes the document into the ‘mashup cloud’, tags it, describes it and sets all the right access controls on it. After all, who knows the spreadsheet better than the person who created it?

It’s time enterprises got out of the data cave.

Read More...

Summary only...