BETA
This is a BETA experience. You may opt-out by clicking here

More From Forbes

Edit Story

Bimodal Considered Harmful

Following
This article is more than 7 years old.

In my conversations with customers and vendors over the past 18 months, I've been trying to find someone who speaks highly of Gartner's bimodal approach to IT. This, by now famous, approach classifies IT into two modes: Mode 1 is traditional IT, and Mode 2 is new-style DevOps/Agile/buzzword laden IT. Mode 1 keeps the lights on, while Mode 2 goes off and does all the new, fun, 'innovative' stuff.

I haven't found any.

Usually, I am met with a polite "Well it's not completely useless" response before we discreetly move on, or something like "well, what they mean is..." explanation that re-frames bimodal to mean something completely different to what Gartner actually say.

However there are plenty of people who are happy to stick the boot in to what they see as a wrong-headed approach to running IT: Author of Lean Enterprise and Continuous Delivery Jez Humble, fellow Forbes contributor Jason Bloomberg, Tech Evangelist at Hewlett Packard Enterprise John Jeremiah, to name just three who have written about this publicly. Add to this dozens of executives with whom I have personally spoken.

I agree with them. I think bimodal is dangerously wrong and organizations should absolutely not use it.

Why Is It Wrong?

The core of the problem with bimodal is that it ignores change. What is termed Mode 1 IT didn't start that way. Once upon a time, mainframes were new and exciting. So was client-server. The fact that many people are currently falling all over themselves in a rush to embrace the new, shiny, DevOps/Agile/nimble approach to IT doesn't change history. What is now termed Mode 1 started out as Mode 2.

So how did Mode 2 become Mode 1?

Gartner's model is silent on this issue, and therein lies its fatal flaw. It is inherently a static model. Somewhat ironically, it codifies lack of change in an attempt to show organizations how to embrace innovation.

Huh?

"It creates a two-class system," says Meredith Whalen, Senior Vice President, U.S. Insights and Vertical Business Units, at IDC. "Ultimately it leads to morale problems. We also think it leaves part of the IT organization to fossilize and doesn't really drive innovation, as any type of organization should be evolving."

IDC is proposing its own framework, which it calls Leading in 3D (download a PDF of the executive brief on this framework here). Simon Wardley, a Researcher for the Leading Edge Forum, has also proposed his PST model that shares many of the same ideas (and which I think is more clearly defined, and more easily implemented).

Both of these approaches share two important differences from Gartner's bimodal approach.

Why These Alternate Models Are Better

Firstly, they introduce a middle mode: Settlers in Wardley's PST model, and Integrate in IDC's Leading in 3D model. This stage is the vital, missing piece from bimodal that connects and transitions Mode 2 ideas to Mode 1. It is what enables brand new ideas created in isolation off in one corner of the organization to be adopted by the organization as a whole, magnifying their impact. It's what enables core systems to provide value out into the brand new customer engagement systems, as the Settlers figure out how to combine the two. It's the connection, the network, between what Gartner would have you isolate from one another.

The second, and more subtle, difference is that these models both provide a way for new ideas to improve Mode 1. The Town Planners and Incorporate stages are both continuously fed new ideas that have been tested and validated by the Settlers/Integrators. Not every new idea will work -- that's part of the point of innovation, to try and fail until you find something that works -- but the ideas that do work are then improved upon, refined, standardized. Standardization is not a strength of people who are focused on newness and innovation, so where does this happen in Gartner's bimodal model?

What astounds me is that Gartner's bimodal model has established any form of traction whatever. Even this most cursory of investigations has shown it to be fatally flawed, so why on Earth does anyone attempt to use it?

Feel free to let me know in the comments.

You can hear my full conversation with Meredith Whalen on Episode 018 of The Eigencast here.

Follow me on LinkedInCheck out my website