InetSoft Webinar: The Goals in Operational Application Development

This is the continuation of the transcript of a Webinar hosted by InetSoft on the topic of "What does Agile BI really mean?" The speaker is Mark Flaherty, CMO at InetSoft.

So if we just look at some of the goals in operational application development, they came up with the word agile. They just wouldn’t be more nimble in their work. And in a lot of ways agile is the wrong word. I think what we are all looking for is just to speed things up. And a lot of times, also with methodologies, lot of times people say, well you know the method is great, but I am a business man or woman. I don’t really care what the guys in the cubes do as long as they crank it out.

Well, actually you should care especially in business intelligence because what the men and women are doing in the cubes takes a long time. And I mean we have all heard the classic complaint: ‘I just want a new report,’ or even simpler, ‘I just want a revision of an existing report, and it took six weeks for me to even get a response and another six weeks for kind of a review, a third six weeks for approval; six months out I am still waiting for this report.’

Where did it come from? And it’s no joke. The pace of business just accelerates so it’s really a need for the business people to get the products of business intelligence sooner to them because with the pace of business, especially where the recession drove all kinds of changes, even the recovery is driving changes because companies are trying to realign to the new reality, the new economy.

Six months later the thing you ask for is not really all that relevant anymore. So really instead of agile, we are just trying to deliver a product, a useful product sooner. And this is true, whether you are in operational applications, or you are just trying to create that report, or you are the data warehouse architect or data warehouse modeler, and you are trying to get new tables, new data structures ready in the data warehouse.

If you are the data integration specialist, trying to get more interfaces in place, trying to get more ETL jobs established and ready to go. Sometimes with business intelligence and data warehousing, we talk about it like it’s one big monolith but really it’s different layers of disciplines. So there are different men or women who do different things there. In lot of ways it’s hard to talk about Agile BI because you are going to have to do different things to speed up development if you are at say the data warehouse and modeling level as opposed to the data integration level.

Agile BI started in report development, and we started calling it agile BI. A lot of teams had gone to this highly iterative process where someone would propose a new report. Typically it would be some kind of business user. And then you would get a fairly quick response to them. You would go through the catalog, and say well in our body of reports, we have something that seems really similar to what you are asking for.

View a 2-minute demonstration of InetSoft's easy, agile, and robust BI software.

And for a lot of teams, they would just start altering existing reports, and say is this getting closer to the look and feel you want? And you try to tell the business person forget the data for the moment, we are just talking about presentation and delivery, does this look like the kind of delivery you are looking for? And that will be highly iterative, and in fact you would have the report developer and the business person in the room together, and some people were doing this 12 years ago. And so that’s actually been a really great precedence for agile BI, and I think of that kind of report development as being the thing that’s easily made agile.

This whole agile conversation has brought us to to get business and IT to talk. So, now as a cornerstone of this methodology, it's an open collaborative environment between business and IT. You have to have that to make it work. And the other thing that I think has really accelerated agile business intelligence is we don’t have to come up with a monolithic requirements document before we can start.

Because of the close collaboration and the true interaction between business and IT, that requirements process is much more interactive, more iterative, and it can change and be more flexible on the fly. We used to have to wait six weeks to get an approval on this. Everything is much more condensed, and the collaboration effort is greater, and your business requirements gathering effort doesn’t become a case of, ‘I am going to go do that then I am going to put it over here and let it collect dust while I go do something else for six months.’

View live interactive examples in InetSoft's dashboard and visualization gallery.

The self management of BI teams, that’s very, very different than how we have run Traditional BI. For it really to become more agile, more speedy in delivering BI products, there have to some changes made in process but also in mindset. The mindset I run across, especially when dealing with backend data issues as well as data integration is this.