As we look to integrate analytics directly to line of business solutions we are forced to completely rethink our approach to User Experience. Unlike our traditional customer, the Analyst who lives and breathes analytics, this new class of user is not impressed with or frankly interested in all the cool analytics we can do. They don’t care that we used a regression model or network analysis; and they definitely don’t want to see ultra cool 3-D multidimensional visualizations that will blow their minds. They have a life ;-)
For the average business user analytics is only a means to an end, and frankly not the most interesting part of that end. All they want is for us to share some valuable insight about their business in the context of what they are doing; thereby enabling them to do it better. It might be advising a purchaser what products to order (because stock is decreasing and demand increasing), a sales guy which customer to call (because sentiment is decreasing and contract renewal getting closer), or a manager which employee to promote (because impact is increasing, external influence is increasing, and attrition risk is increasing).
This shift is very hard for analysts to stomach for two reasons:
Firstly… analytics is tough! If we sweat blood and tears to generate some amazing insights we want to show that off to the world. No-one wants to make their work look simple or trivial, especially when it was so difficult to do in the first place. What we really want is to show the customer that amazing visualization which will impress the pants off them. However, simplify we must. We have to take a back seat to the business application essentially “hiding our light under a bushel”.
Secondly… it’s damn hard to simplify! Condensing analytic insights down into simple contextual business-relevent recommendations is way more difficult than presenting up the analytics in our traditional analyst-style dashboards. If you simplify the analysis and insight too much you loose the ability to make meaningful business decisions from the analysis. However, if you don’t simplify it then you loose the business user, and they are the person best positioned to leverage the insight.
So folks … we have to make ourselves invisible. We have to align our analytics to business problems and integrate ourselves into the fabric of doing business. No-one should see that we are there until we are not. It’s like mothers, no-one appreciates what they do around the house until they take a vacation and then everything goes to rack and ruin :) The more invisible we are, the better the job we are doing. Eventually people will move from talking about “analytics-driven business decisions” to just “business decisions” because no-one will be able to imagine making a decision that isn’t validated by analytics.
To paraphrase the great physicist Richard Feynman; “if you can’t describe a theory simply, then you don’t really understand it”.