Build a better dashboard...

I was speaking with a client yesterday, and he was asking about how they could start to use their BusinessObjects Edge deployment to deliver dashboards. This is generally best done using Xcelsius (though not exclusively). Here are a few tips:

• Xcelsius is a great tool for developing visually appealing and easily understandable data presentations, and it's pretty easy to use. If you have a 5 CAL license for Edge, you already own one Xcelsius designer license for use in that environment. (And if you didn't see my recent LinkedIn post SAP was giving away licenses for Xcelsius Present - the free version only works with Excel data sources, but it's a great place to begin learning Xcelsius)
• Xcelsius can use a few different data sources in an Edge deployment - Live Office, Query as a Web Service, and a custom Web Service. Each have their own pros & cons:
• Live Office  - Create a report using Crystal or Webi that gives you the data you want to visualize; then put Xcelsius in front of it. It's great when you need to present summarized data at a level that doesn't already exist in the data source. Be careful about performance of the report though - you may want to consider scheduling it.
• Query as a Web Service - QaaWS allows you to connect directly to a Universe to draw data into a report. It's fast and pretty simple to use, but it doesn't allow much flexibility - no ability to create summaries, so sometimes work needs to be done in Excel.
• Custom Web Service - A custom web service allows you to create the exact data source that you need, which is a big plus. On the other hand, you also need the technical skill to create the web service, so that's it's downfall. Can be great if you've got the skills, but most of our clients don't bother.
• Want to do some fancy things with Xcelsius? Her are a few tips
• G-Maps Plug-In from Centigon - Ryan Goodman's company has created a really neat tool to integrate Google Maps with Xcelsius visuals. It's not free, but if mapping is key to your requirements, it's worth a look.
• Xcomponents - Donald MacCormack originally made these Xcelsius Add-ins available (for free), adding a few new Xcelsius visuals to the mix. Some neat stuff here.
• XWIS from Antivia - Antivia has been doing great add-ons for BOBJ products for years, and XWIS is another great addition. Hard to explain well, but take a look at their video demo.
• Remember that you can now put Xcelsius content into Crystal Reports - we'll be testing it soon to verify that this now allows scheduled deliver of dashboards (by cheating - hide the Crystal Report detail and just show the Xcelsius widget)
• Keep in mind that a dashboard is NOT JUST about gauges and charts - the "Flashy" stuff! Sometimes a well designed report is a great way to present information; sometimes a simple chart works too. Pick the best tool for your requirement, not just the one closest at hand.
• And keep in mind that in an Edge dashboard, you can now link various pieces together for great interactivity. For example, clicking on an Xcelsius map could cause a Crystal Report, a crosstab, and a line chart all to update to reflect the customers on the map. Really cool stuff - requires some planning, but can deliver great benefit.

Anyway, hope that give you some ideas. If you have questions, feel free to contact me by email or phone.

jpayton@dynamicintelligence.ca
(519)745-0722 x111

Colour me Purple!

I was just doing a bit of reading - catching up on some blogs written by friends and experts in the BI space. And I came across an interesting post by Wayne Eckerson called 'Purple People'. In simple terms, Wayne uses the term 'purple people' to refer to those exceptional individuals who can harness all the required resources (Business, technical, etc) for a BI project and bridge the gaps between them to achieve amazing results.

What makes this interesting? Well one of my other favourite reads is the blog and books of Seth Godin, and one of his books is called Purple Cow. Seth talks about a purple cow as something exceptional and memorable... something you'll never forget. See where I'm going with this?

Let me just say that it should be the goal of every BI consultant (and for that matter, consultant of any type) to be PURPLE!!! Be exceptional!!! Go the proverbial 'Extra Mile!' Deliver beyond your clients expectations and give them an experience they'll never forget!

As for me - I have a new favourite colour. Hope to share it with you soon!!!

:)

Are you really different?

Business Intelligence vendors create their products as 'toolsets' - platforms that let you deal with the unique reporting requirements of your business. Instead of providing a collection of fixed reports and metrics, they provide an opportunity for you to create your own reporting and metrics environment from scratch. That way, all of the reports are specific to your business.

So what's the problem with this? Well for starters, not every leader of an SMB has a clear understanding of what Business Intelligence can do for them. Though they really know their business operationally, they're not always sure what to measure and how it can be used to improve things. I don't know how many times I've met with prospective clients who have asked very early on for some samples of what other businesses are doing. They're really looking for a way to jumpstart the process.

Here's the question then - When considering a BI approach for your business, how important is it to have something that is ready out-of-the-box? Is it better to buy a toolset that let's you build from the ground up, or would it be preferable to have some content that is ready to go right away?

My belief is that, though all companies have their own unique aspects, there are also a lot of similarities. And with these similarities, there are a lot of standard metrics to monitor and analyses to be performed. Regardless of the business model, there will be some form of 'revenue', some 'customers' and obviously some 'vendors' (and lots of other similarities). It's probably safe to say that most businesses share about an 80% commonality. Wouldn't it be good if the 80% was ready out of the box? That way you could use the time saved to focus on addressing the other 20% more effectively.

Pervasive BI

I recently read an article (I do a lot of that) that talked about "Pervasive BI", and frankly I'm getting really tired of the phrase. Based on recent research, Pervasive BI is a myth. BI adoption rates in most businesses are around 10-12%. And from my perspective, that's just fine.

Please don't misunderstand - I think it is essential for knowledge workers to have access to good information to be effective. But shouldn't the focus of a BI deployment (at least initially) be on providing high quality strategic information that isn't readily available from other applications? Shouldn't the operational BI requirements be met by the operational applications?

Is the industry bias towards pervasive BI anything other than a push to get more licenses sold?

Professional Courtesy

A friend of mine recently decided to give up working in his professional field (he holds a CA designation), and get involved with financial planning. Through the particular organization he joined, he was required to do a lot of 'cold calls', knocking on people's doors. He's shared a number of stories about his experience with me, but one of the things about knocking on doors is that, if they are home, almost everyone will answer. And when they do, they will either say 'Yes, I'm interested', or 'No!, I'm not interested!', (or something a bit more colourful). But the thing is, he always gets a definitive response.

Let's put that into a Business-to-Business context. Imagine you download a product information sheet from a website, and in doing so provided some contact info. The vendor responsible for the whitepaper contacts you via email to ask about your interest. You ignore it. They follow-up with a telephone call, and you avoid the call. They place another three or four calls to you over the following week, and you ignore every one of them (Call Display at work!). Eventually the vendor gives up and moves on to the next prospect.

So what was accomplished? Well, let's see - you wasted the vendor's time and resources when they tried to contact you. You wasted your time by consciously deciding to ignore the phone calls and sitting there listening to the phone ring. You probably created some unnecessary anxiety for yourself, because you feel you're being harassed by the vendor. And you both probably have a lesser opinion of the other after the interaction (or lack of).

Here's option 'B' - you answer the phone. Give the vendor thirty seconds to tell their story. At the end of that, make a definitive decision to either continue the discussion or thank the vendor and say you're not interested. I know you're probably thinking you have no obligation to the vendor, and in simplistic terms that's true. But if you think about your own company's outbound marketing and sales activities, imagine how much time they would save if they weren't chasing ghosts. How much time collectively would we all save if people were just honest with each other at the front-end of discussions...

Keeping it brief...

Here's my thought for today, for you to ponder...

There’s no guarantee that good information will lead to good decisions, but there’s no doubt bad information leads to bad decisions.

Jim

Consensus

Wow! Three weeks into another year, and this is my first blog post! I feel great shame!!! :)

The topic for today is Consensus, and how it is critical for a successful BI deployment.

Many organizations that start down the path to enabling their users with Business Intelligence capabilities leap into the fray, boldly planning to put answers into the hands and heads of everyone. The problem is that sometimes the users can't ask the question in the way they want. Not because the tool isn't capable. Rather it is because the tool doesn't speak their language.

When you decide to implement a Business Intelligence solution, start by ensuring it speaks the same language as your users. Use their terminology and definitions. And if different groups have different meanings for terms like Product or Vendor or Customer, make sure you get them to either reconcile their definitions into one universal one, or define a different name for their respective elements. Whatever avenue you take, you must eventually reach a consensus about the definitions. Inconsistent semantics result in inconclusive or untrusted BI, and that's never a good thing.