Data infrastructure optimization, availability & security software
Data integration & quality software
The Next Wave of technology & innovation

Expert Interview with Donald MacCormick from on Big Data Dashboards

Photo: Donald MacCormick, Chief Product and Strategy Officer, Antivia

Donald MacCormick, Chief Product and Strategy Officer, Antivia

Donald MacCormick has been an active member of the business intelligence community for over 20 years, having served in a variety of senior product management and marketing positions for SAP, BusinessObjects, Crystal Decisions, Seagate Software and Holistic Systems prior to Antivia.

Antivia features the business intelligence dashboard concept on its home page. Does this mean the dashboard, once the calling card for business intelligence apps, has been mistakenly left for dead?
I’m not sure how dead the dashboard is. We still see a lot of demand in the market for people wanting to deliver dashboards across their organizations. However, the concept of a dashboard has definitely evolved in the last few years. Originally, a dashboard was a static overview of a process or area of the business, that visually highlighted performance above or below target. The problem is that as soon as someone sees summary level data, they immediately want to ask follow-up questions.

The new generation of dashboards addresses this with interactivity. So when someone sees a number, they can quickly and easily ask questions around it: how does it break down over regions? What does it look like for a different product? What it is predicted to be in the future?

These interactive dashboards with their application-like functionality have much in common with the mobile applications we are familiar with using in our everyday lives. A great example of this is a weather website which is something we all use regularly. It sits on a truly huge dataset and involves predictive analysis,; but nonetheless, the interface allows us to quickly get an overview of the weather in our area. Then with only a couple of clicks or taps, we get the specific information we want (the forecast for another day or another area, or if it is likely to rain this afternoon) all from amongst a huge set of predictive data – all without any training or the need for documentation. We just know how to use it.

Ironically, this simplicity of access to information that is so common in our personal lives – accessing bank accounts, weather, travel timetables, sports results – is pretty much absent in the business world. The solution to this is the style of interactive dashboard which we are championing with DecisionPoint™, which packages up business information to make it as easy to digest as a consumer mobile app.

How has the target community for the dashboard shifted now that the concept is considered “mature?”
I don’t think the target community for dashboards has changed. There has just been a dramatic change in the level of value dashboards can provide now that they are interactive. The target community for dashboards has always been business users at all levels across the organization – sales people, store managers, product managers, the executive team, call center team leaders and so on. However, static overview dashboards were of limited value to most of these people and so fell into a little bit of disrepute. With the addition of interactivity, the game changes. Suddenly, information which can drive the minute-by-minute activities of everyone across the organization is easily available on the desktop and on a mobile device.

Excel has long been seen not only as “Business Intelligence Lite,” but a sort of Common Core competency for business analysts. Since Antivia offers an Excel enabler, DecisionPoint For Excel, how has this changed over the past dozen years or so?
Excel is, has been and always will be a key component of any business user’s armory. As a colleague of mine once said: “You will only take Excel out of their cold, dead fingers!”

This is one of the driving reasons behind DecisionPoint™ for Excel. We know there is a massive amount of business data sitting in Excel spreadsheets, and there is a great need to free that information and deliver it to other people across the organization. Often, this is done by mailing the spreadsheets themselves around; however, this is far from ideal as it leads to complexity and confusion. DecisionPoint™ For Excel provides an alternative method for this information to be distributed by packaging it up in an easy-to-use, no-training-required, interactive dashboard which broadens its appeal, use and value.

What is your take on Microsoft PowerPivot? Has it enlarged the perspective of the Excel community of interest?
Microsoft PowerPivot is an interesting innovation. It is probably fair to say it is one of the first significant extensions to Excel’s functionality for a number of years. It appears to be targeted mainly at the analyst user in competition with Data Discovery tools such as Tableau and QlikView, providing easier ways to slice and dice relatively large volumes of data. Anyone who uses Excel extensively should certainly look at this and its companion PowerQuery. However, as with all these tools, it falls short in the wider distribution of information.

For many years, the business intelligence industry has been pushing the idea that empowering non-analyst business users to do everything for themselves without the need for intervention from IT or the BI team is a desired endpoint for business intelligence. Unfortunately, this has held BI adoption back more than any other single factor. The truth is that although this type of ad-hoc self-service is ideal (actually, vital) to analysts, it is something that has never taken off despite many attempts in the non-analyst user community.

Going back to the weather analogy, it’s like offering consumers a web service in the form of a huge database of raw weather observation data and providing an ad hoc query tool or Data Discovery tool on top of it. Clearly, this is never going to fly. While it may allow for enormous flexibility, it would almost never be used because people have neither the time nor the inclination to access their weather forecasts in this way.

Similarly, operational business users simply don’t have the time to sit down for extended periods to learn a tool, wrestle with data and do business intelligence for themselves. They need these easy-to-use, task focused, interactive, connected information apps to give them the data they need at their fingertips.

It may take a little more effort on the part of the technical teams to deliver this, but anything less is an abdication of responsibility.

Has Antivia had particular success in settings where SAP Business Objects is traditionally employed?
Antivia has a strong history with SAP BusinessObjects. Indeed, one of our early products was an extension to one of SAP’s products. This history is reflected in the fact that DecisionPoint™ Enterprise has some of the best connectivity to the BusinessObjects platform. As a result, many of our customers are active BusinessObjects users, and DecisionPoint allows them to leverage the significant investment they have made in the BusinessObjects platform but deliver content in this new interactive paradigm.

How was DecisionPoint engineered to address the challenge of Big Data Volume?
There are two main ways in which DecisionPoint™ is engineered to address the challenge of Big Data. Firstly, DecisionPoint™ operates a three-tier data architecture between the dashboards, (running either on the desktop or a mobile device), the server, and the underlying data store. Data is managed between the three layers using our advanced cooperative processing architecture, which means that all three layers perform some of the computation for maximum efficiency and performance while providing off-line access.

Obviously, it is not possible to process a multi-terabyte dataset with an iPad, but it is very possible in some of the new analytic database platforms. DecisionPoint™ sits between these two to make sure that data is handled at the most appropriate level; with high-level data coming down to the device so that it can be stored and accessed offline, while queries that involve huge volumes of data are pushed down to the server and processed on the Big Data database.

The second way in which DecisionPoint™ addresses Big Data goes back to the comparison with consumer websites. These often service large volumes of data in an easy-to-use, easy-to-manage interface. Similarly, both the flexibility and the application-style interface building capability of DecisionPoint™ allows our dashboards to package similar interfaces, thus helping the user navigate through appropriate parts of a Big Data set without ever overwhelming themselves or the technology.

How has HTML5 enabled products such as yours to deliver a better user experience?
The key factor in our choice of HTML5 as a delivery technology for our dashboards was its cross-platform delivery. Every platform provides a way, and often a number of ways, to render HTML5 content; so it is the closest you can get to “build once, deliver anywhere” for desktop and mobile. However, cross-platform delivery would make no difference if the technology were not capable of delivering the required functionality. Fortunately, here too HTML5 comes up big. We have been continually impressed by the capabilities HTML5 provides both in terms of user interface development and in data processing. These are at the heart of two of the key features of DecisionPoint™; namely, interactivity and advanced corporative processing.

What other cloud or on-premises competitors are you following with interest?
Obviously, we follow all the players in the BI market; although we believe we are different from most and possibly all of them in that we are focusing towards the “information application” end of the spectrum, whereas most others are focused on data discovery, standard reporting, or static and semi-static dashboards.

In producing dashboards that bring together disparate data sources that could potentially weaken privacy protections, do you worry that end point apps like dashboards could be blamed for outcomes that should be associated with other data pipeline stages?
There are very real privacy concerns in terms of what is possible when cross-correlation between seemingly anonymous datasets allows personally identifiable information to be deduced. As we collect more and more information, this risk is only going to grow and is definitely something we all need to guard against. However, it is unlikely that dashboards or dashboard technology will be blamed for this as it is a problem at a much deeper level in the data.

How will Big Data Velocity change the nature of dashboards over the next five years?
Even now, there is a lot of discussion about real-time data being used to drive business decisions; and as data capture and processing systems mature, this will only intensify. There is also a growing expectation of access to real-time information. Just today, I took some money out of an ATM and for some reason my first attempt failed, so as I walked away from the machine I logged onto the mobile banking app on my phone and saw that the second transaction was already registered in my system maybe only 15 seconds later.

As a result, there will inevitably be a push towards dashboards which have some level of real- or near-time information in them. However, the challenge here is not for the dashboards or the dashboard technology which cope with this problem well today, but more for the information supply chain underneath.

Load up a Big Data dashboard with Syncsort DMX-h.

Related Posts