When Procurement Met Siri: A Simple Future for Spend Analysis? (Part 1)

This post was co-authored by Jason Busch and Ryder Daniels.

Even as data and visualization geeks, we’ll be the first to admit that we’ve overcomplicated our professional and personal lives with unnecessary and confusing interfaces that to a data analyst from another world, planet Earth is quite an amusing place. Consider BMW’s iDrive interface (a “convenience” in our car) in finding out about traffic on a designated route. To an advanced alien anthropologist peering at the activities of earthlings, it would really look like a bad joke that has done nothing but complicate life – and how to best get from Point A to Point B, let alone turning the radio dial.

With iDrive, a basic query takes at least 3-4 clicks and spins of the device to input your destination and pull up a traffic map. Ideally, in the best of circumstances, this takes 30-45 seconds. Of course alternatives do exist. Consider Apple’s Siri (and similar technology). Ask Siri, for example, “how is the traffic on the road ahead” and she will respond to your question (she might send you to the wrong location, but that’s another problem entirely).

How simple and clever! But there’s limitation to this. Siri won’t reroute you based on traffic down the road (a good navigation system will). But if you want a basic answer fast, with a good probability of finding what you’re after in a matter of seconds, it can’t be beat.

Ironically, this approach harkens back to the old days of basic queries and text-based simple interfaces (i.e., non visual). The great complexity we’ve wrought with new types of spend- and supplier-driven BI mash-ups is overkill for much of the routine of the non-power analysis accessing procurement information.

A rich and engaging user experience needs to be balanced with practicability and speed. Simplicity in general doesn’t have the same “wow” factor in demonstrations as highly intricate and complex cascading dashboards, faceted navigation, and n-tier drillable data. But simple, text-based questions – either spelled out or asked to a voice recognition system are hands down the fastest way to answer questions.

Just as with the BMW navigation experience, think about the similarities to navigating many spend analytics packages (if not all). Consider the number of log-ins and the number of clicks to get at a particular piece of data or to open a report or to click through a dashboard. And think about the context in which you need to find information, e.g., in a meeting with a supplier, with a senior manager with business stakeholders, and how it can be distracting to pull up information that is not already at your fingertips (“oops … that didn’t make our briefing book”).

What’s needed is something new, something that harkens back to the future. Siri has no new content - it sits on top of content. It really is simple – a basic text query applied to the spoken word. And it can tell us “what is the traffic ahead” much faster than the over-engineered navigation systems that cost at least 2-3X the amount of an iPhone.

Below, we feature a number of screen shots of a mock-up we created that shows the power of a simple widget sitting on top of spend analysis to answer basic questions. As we continue our analysis tomorrow, we’ll turn our attention to more advanced examples of the types of queries such a text or voice-based system could answer – as we head back to the spend analytics future!

(Click images to enlarge)

4 3 2 1

In the meantime, think about how simple life could be if we could ask our system such queries in free text or voice!

Discuss this:

Your email address will not be published. Required fields are marked *