“It depends”. When talking about a development “style” then the answer is Yes, but when talking about a stiff framework the answer is No. The important parts of agile are a set of core values and guiding principles. In analytics we shouldn’t use agile as a rigid methodology rather as a method to deliver working features early and on a continuous basis within the development lifecycle. Do we need a Program or whatever owner? Do we need daily standups? In my opinion No. What we need is a very close collaboration with the business and all parties involved in the development lifecycle. Providing results in short time intervals, getting feedback from the customers, incorporating the feedback and providing additional or enhanced results again. That’s the whole magic and we can call it prototyping or “fast delivery” or whatever we want. All we need is a list of desired working features, agreement on delivery phases and regular meetings involving project members and the business stakeholders.
So let’s rephrase the questions from the beginning: “Do we need a process enabling us to deliver results in a good quality within a relative short timeframe”? To that question my answer is: “Yes, definitely”. That’s one of the huge benefits we have in the data analytic environment. With our methods, tools and knowledge we are able to fulfill all these topics. Discussing requirements with the stakeholder, sourcing and cleansing data, loading into Tableau or into another visualization tool, providing the result to the customer and discussing potential changes or enhancements. After some or many development cycles a new product or prototype is born. Especially in big firms where IT processes often have become dull and slow, these ways of providing fast results to the customers (potentially in a first instance only) are in my opinion “pure gold”. But please, don’t understand me wrong. I do not think that all IT data projects can or should be replaced by data analytics but it’s very efficient when first prototyping and then going operational including all IT processes when wanted and needed.
So please go (smart) agile in data analytics projects.