Skip to Content

dashboard forecast.png

This was supposed to be a series of three blog posts exploring the lessons that the BBC weather site can teach us about BI in general and dashboards in particular (the first 3 posts were on the future of dashboards, Big Data and HTML5). However,  after finishing the original series I realized there was a fourth lesson staring me in the face, which should have been more obvious than the first three, because it is all about predictive analysis.

The reason I say this lesson should have been obvious is because a weather site is all about the future. People care about what the weather will be tomorrow and not about what it was like yesterday. As a result, almost everything you see on a weather site is a prediction of the future.

However, when you are using a weather site you don’t consciously think (or even care) about the methods and algorithms that have been used to generate the forecasts you are viewing. Of course not; you just trust the experts whose knowledge and understanding make the forecasts possible. All you have to do is navigate the information provided through an easy-to-use front-end interface (i.e. a dashboard).

This is exactly the same in the world of BI; as an end-user, you may need forecasts in your dashboard, but you should never have to think (even fleetingly) about how those forecasts are calculated. That is someone else’s job. Your job is to use the forecasts through the same simple, intuitive dashboard interface that you view all of your other data through.

Interestingly, SAP has just announced that it is merging its recently announced predictive analysis capability into its Visual Intelligence product. This makes perfect sense. Both are tools for analysts; people who specialize in (and understand) technical data manipulation.

If you don’t have “analyst” (or perhaps, more recently, “data scientist”) on your business card then you should not really be using such tools. Someone else should be providing you with interactive dashboards built on the work of people with specialist skills, knowledge and training who use these sophisticated analytic tools which then enable you to  get in to your data, get what you need, get out and get on with your job as quickly as possible. After all you have a business to run 🙂

To report this post you need to login first.

1 Comment

You must be Logged on to comment or reply to a post.

  1. BW BW

    Hi Donald,

    I like the blog series and the rest of your contributions, a lot – they are informative and stimulating.

    I would however like to make a comment.

    My experience of working in the BI arena for more years than I care to admit to is that the Business *always* wants to know what logic is behind the numbers (its what I call the “amateur Excel” syndrome – because almost all users of BI have also taught themselves how to use Excel and therefore have an inherent need to validate the logic). Its inescapable.

    For me this is the key challenge with the Dashboard format – both historical and predictive. When Dashboards are used by Subject-Matter-Experts they have to be supported by all of the qualifying logic before they will be accepted by SMEs.

    Unsupported Dashboards are a threat to SMEs and instill resistance. This is because part of being an SME is that I have to have confidence that the aggregations (abstractions) I use to make a decision can in fact be relied upon. Until I have that confidence I will always be liable to shoot the messenger …

    Dashboards used by non-SMEs (as in your BBC example) tend, in my experience, not to attrach a similar level of suspicion, but few of us encounter these in our professional lives.

    Just my 2c worth.

    (0) 

Leave a Reply