Data visualization and the dynamic dashboard

 I’m a big fan of data visualization because it really helps people understand information.  You can certainly derive a lot of insight from a report, but sometimes it helps to look at data in a different way.  Dashboards are one way to get information to people in an easily digestible format.  However, in order for dashboards to be effective, they need to be:

 

  • Engaging -  meaning you have to want to look at them 
  • Useful – meaning that they provide valuable information that is easily understandable

 

Anyone can say they provide a dashboard if they have a few gauges and charts, but if they are flat representations of information they don’t go a long way to help to display complex information. On the other hand, there can’t be so much happening in the dashboard that the user becomes overwhelmed and confused.

 

Now, I have to admit that I wasn’t a big fan of certain dashboard visuals – such as the gauge.  This may be because when I was developing executive information systems (way back when), our gauges were fairly static, so they didn’t convey much information.  However, I recently had a conversation with Shadan Malik, CEO of iDashboards, and I found the interactive dashboards he showed me quite engaging and useful.

 

Here’s an example of what I mean.  This is a static screenshot of a dashboard, from iDashboards, (which actually illustrates my point about some static displays of information).  But, click here to see the actual dynamic dashboard built using Flex® and Flashâ technology.

 

 

callcenterscreenshot1

 

This dashboard presents information from two different bank call centers – one called Auburn Bank and the other Regel Bank.  This is actually a real dashboard, but the data and names have been changed to protect the innocent, which explains the fact that some axes aren’t labeled and why some of the data is a bit suspect – but you can get the idea of how useful a dashboard can be. 

 

There’s a lot of information on the dashboard, but it isn’t overwhelming.  The dashboard actually tells a story.  In this case, it is the story of two different banks and how well each bank’s call center is performing.  Right away, you can see that (for Auburn Bank) over the past six months, there have been some issues with cost and budget, that the percent of calls coming through the call center (vs. the web) has increased and abandonment rates are decreasing.  The interactive mode helps to drive some of the information, home however.  For example, Auburn Bank has been having some issues with answering calls in a timely manner.  I found this out in just a few seconds of looking at the dashboard by hovering over the cost/budget chart by month and looking at what was happening in gauges and when it trended into the red zone.  You can also toggle between the two banks to see how each is performing.  You can also drill down into any piece of data to get more information.

 

The company recently announced that it has incorporated Flex technology into its product.  This increases the speed of processing data and enables even more interactivity.  Of course, it is a fine line between visualizing just the right amount of information and being overwhelmed by too much information.   My understanding is that iDashboards works with many of its customers to help deal with data issues, determine important metrics and walk through the storyboarding of the dashboard.  This is probably a good thing, given the power behind the product.

Practical Operational BI

I recently had an interesting conversation about business intelligence and operational BI with Steve Mauss, President and COO of Knowledge Relay.  Knowledge Relay was founded in 1984 and has its roots in project management software for the aerospace/defense and government verticals.  Over the years, it has expanded its reach to support other industries such as utilities and construction as well as adding new data access and analysis capabilities.  The company leverages open source software.

 

Knowledge Relay

 

The goal of the company is to help organizations “improve operational efficiencies by reducing the time between collecting data and acting on it”.  In fact, Knowledge Relay promises that it will get your large or mid-sized company up and running on its platform in 30 days.  The company has several products that help it to do this.  One is called Data Integrator.  The other is Information Visualizer. 

 

  • Data Integrator.  Data Integrator comes with over 250 connectors out of the box to help companies connect to a myriad of data sources including CRM, ERP, databases, financial applications, and project management software. .  Data integrator also performs the Extract, Transform, and Load (ETL) part of the integration puzzle.  It uses a drag and drop interface to map source to target destinations. The company deals with data quality ty in a couple of ways.  Its ETL solution has built-in quality control as the data is migrated and transformed.  It also provides an applet that performs quality checks in the background (pre- and post-migration) and, if/when it finds a problem, it notifies (i.e. via email to computer or mobile devices) the responsible parties and asks them to fix it.  
  • Information Visualizer.  Information Visualizer combines multiple data sources into a single view.  It includes Gantt, PERT, X-Y line, pie, and text reports as well as some dashboards (which it OEMs from iDashboards).  The company also supports sharing of information via a job server that distributes PDF reports automatically via email, a web portal, or a number of print options.  I’ve included some of the visualizations, below.

knowledge-relay2

 

Knowledge Relay supports some pretty practical operational applications.  For example, an energy company might use the software to analyze regular maintenance tasks that involve the integration of project management, ERP, and MRO information.   The system will help management understand how it is performing on these tasks.  It can also be used to generate a list of items that the guy on the floor needs in order to deal with something like outage management. 

 

Operational BI

 

There has been a lot of hype over the past few years about operational BI and particularly real time operational BI.   Think embedding and automating analytics in a process so a person (or another process) can act on that information in real time. A good example of this is a call center.  In this use case, a call center representative can use the information about a customer he or she is speaking with to up-sell, cross-sell, and so on during a phone call.  How does this work?  When the call is received, the system retrieves the information on a customer and it is analyzed and passed to the call center agent who can then act on it. This sounds exciting, and it can be, although industry studies indicate that this is not yet mainstream technology. 

 

Operational BI, at its core, is really what Knowledge Relay is talking about – providing information to people to help them make decisions and take action in the context of a business process.  At Knowledge Relay, operational BI is about utilizing BI to analyze the operational aspects of a business and provide this insight, not only to executives, but also to the people on the front lines.   This analysis is embedded (and automated) into the process.  As Steve put it, “It’s not always enough to find out about an operational problem, it’s also important to identify (data quality and other) issues as soon as possible in order to avoid cascading problems throughout the system; problems that will later have to be undone just to get back to a baseline. “  There is a lot of value in this.

 

 

Sales metrics that matter

Sales are a critical component of what makes a company tick and metrics that measure sales performance are used by many organizations. Some companies track sales by category and examine pipeline metrics, sales metrics and win loss ratios. However, while companies may have metrics in place, this doesn’t mean that they are tracking the right metrics or tracking them properly. For example, a company might track days to close a deal, but not think to classify deals by deal size or by vertical, which may be where the real insight lies. Or, a company might track this metric but can’t track it over time. Some companies don’t necessarily even know the right questions to ask, or how to interpret the results.

I recently spoke to LucidEra, a company that specializes in sales analysis. The company targets mid size organizations and provides business intelligence capabilities and best practices to help organizations improve sales effectiveness. Rob Reid, CEO and Ken Rudin, VP of Marketing at LucidEra, explained that the company has compiled numerous metrics and analytics to help provide insight to sales organizations. The company delivers this functionality via a SaaS model. It works like this: LucidEra brings in your company data from a number of sources including salesforce.com (it can actually be installed as a tab in salesforce), Microsoft CRM, Excel, or a custom data source. It then analyzes the data and provides reports, graphs, and other visuals to the end user. Here’s a screen shot of one kind of analysis that examines new versus repeat business:

new-vs-repeat1

LucidEra offers a number of prebuilt metrics and dashboards and also allows companies to customize their own analysis. In this example, you can see that sales associated with new business is essentially flat and the revenue associated with this kind of sale is lower than that associated with repeat business. Repeat business, on the other hand, is higher and has a repeated pattern that shows an increase in the 4th quarter. This then begs the questions – what is driving the spike? Can it be used in other quarter? Why is first time business low? You get the idea. Other kinds of analysis include reporting on and analyzing what’s changed in your pipeline, leads analysis, and sales representative performance.

To help companies get started, LucidEra provides a Pipeline Healthcheck service which focuses on identifying trends in key sales metrics, interpreting what the trends mean, and quantifying how changes in those metrics can affect your business. LucidEra will take your sales data and give you an assessment within 48 hours. Some of the topics a Healthcheck includes are:

  • What changes have occurred in your pipeline in the last week, month, quarter, etc?
  • What are the trends across your critical pipeline metrics such as win rate, average deal size, and the average length of your sales cycle?
  • What are the key factors that influence these metrics? For example, how does the length of your sales cycle vary by industry, by lead source, by the size of the deal, etc?
  • What is the likelihood of a deal closing based on the characteristics of the deal, such as whether it’s a new customer or a repeat customer?
  • Which forecasted deals are at risk? For example, which deals are stuck in your pipeline?
  • Where are the characteristics of the deals that you’re most likely to win; that is, what’s your sales “sweet spot”?
  • How do sales reps and regions compare on key metrics such as win rate or deal size, and how is that trending over time? How do these metrics differ based on how long a sales rep has been with your company?
  • What is the quality of your sales data, and what key information is missing which would give you valuable insight into your pipeline, sales people, or sale process?

The upshot?

I’m a big believer in asking meaningful questions about data and interpreting it properly. Sometimes, companies have good intentions – like analyzing sales data – but don’t know the right questions to ask. They may have the data at their fingertips, but this doesn’t help if they can’t effectively analyze it. Don’t get me wrong, I’m not saying that making BI more accessible is a bad thing. My point is that it does take some talent, however, to ask the right questions and interpret the results. Companies like LucidEra that provide a best practice approach in a useable way can help organizations that are struggling to figure out what all of their data means.

Metrics that matter actually do matter.

Follow

Get every new post delivered to your Inbox.

Join 1,190 other followers