Business Agility: How Does Cloud Computing and Big Data Help?

How do cloud computing, big data, and other “agile” technologies improve business agility? Why, “Automation!” of course. That’s part of it, but you’re missing the biggest gains.

This is really part 2 of a series of posts on business agility. Part 1 is titled “Business agility, fighter pilot style,” and can be found over at GigaOm. Read that post first and then come back to read this one.

In “Business agility, fighter pilot style,” I explained how Col. John Boyd, a fighter pilot in the US Air Force, gave us a framework for thinking about agility. Boyd identified a series of processes that everybody goes through all the time when reacting to the external world, something that has come to be known as the “OODA Loop.” The OODA Loop has four processes:

  1. Observe — Observe events and the state of the world around you
  2. Orient — Determine the meaning of that state and those events
  3. Decide — Make a decision about what your reaction is going to be
  4. Act — Attempt to carry out that decision

The OODA Loop is happening all the time, often without your conscious knowledge, whether you’re making a big decision like buying a house, trying to figure out your order in a restaurant, or battling to the death in an air-to-air dogfight. Individuals have OODA Loops, as do large organizations. Boyd identified that during competition and conflicts, participants could generate a greater chance of success if they:

  • Increase tempo — Speed up the rate of decision making such that the adversary is always “behind,” reacting to the last thing you did. My friend, Alistair Croll, just wrote a great piece for O’Reilly Strata describing how incremental improvement and organizational learning, applied at a fast tempo can outstrip the competition
  • Generate what Boyd called “fast transients” — Do the unexpected, confusing the adversary and causing him to question his own strategy and decisions

Finally, I said that organizations that want to take advantage of Boydian thinking should:

  • Create small loops — Break teams down into smaller, semi-autonomous units. Fewer people means faster decision making, leading to increased tempo
  • Communicate high level intent — Small, autonomous teams could devolve into chaos if not well-managed, so provide those teams with a “high level intent,” a set of strategic goals that they should pursue, but otherwise leave them free to choose appropriate tactics for reaching those goals

So, how do cloud computing, big data, and other agile technologies fit into Boyd’s framework? The obvious answer is “Automation!” Take manual processes and turn them into automated workflows. By doing so, we increase the organizational tempo and we get more agile. Easy peasy. That’s certainly one piece of the answer, but it’s way too simplistic.

Take a look at the OODA Loop diagram. Notice that there are four phases: Observe, Orient, Decide, and Act. When you think about it, most IT automation systems really just target the final step — action. Given that the organization has decided to do something, somebody pushes a button and the automation system makes it happen, quickly and efficiently. But that’s a big “given.” What about the other three phases? It turns out that agile technologies also have an effect on the rest of the OODA loop.

For example, one halmark of cloud computing is the availability of a self-service portal that allows users to consume IT services on-demand. Because IT services are cheaper and can be provisioned and deprovisioned quickly, there is reduced need for a long, complex approval process. In other words, we’ve reduced the size of the Decision phase of the loop.

Further, big data technology coupled with personal visualization systems like Tableau Desktop can “democratize data,” targeting the Observation and Orientation phases of the OODA Loop. The number of people that need to be involved with a given OODA Loop shrinks still further, making your teams smaller and more autonomous. While ascribing true meaning to a state or event can still require a human being, additional monitoring and management systems can help alert teams to anomalous and interesting conditions. Tempo increases as small teams become more autonomous and individually empowered.

These technologies also increase the odds that teams will generate “fast transients,” unusual market moves that will throw off competitors. While the creation of fast transients is a uniquely human skill (fast transients are not random moves, but rather calculated but unexpected moves that still lead toward the main goal), tools like cloud computing and big data provide teams with the tools to identify opportunities and create experiments quickly. They support a “fast-failing” culture of experimentation that can iterate quickly  to adapt to changing market conditions and generate the compounding effect that Alistair talked about in his article.

The upshot here is that if you want to create a culture of agility in your enterprise, you need to think about more than simple automation. While automation certainly has value once decisions are made, there is even more value in targeting the OOD portions of the OODA loop. Using cloud computing and big data, you can create smaller, more empowered teams that can iterate faster, increasing tempo and generating more surprising market moves.


  1. Hendy Appleton says:

    Thanks for this, Dave. As we transition from Vyatta-the-small-startup to Vyatta-a-Brocade-company, maintaining agility is a huge concern. I’m going to share both articles up my new management chain.

  2. Duane Gassman says:

    Dave, great articles and a good practice to follow. I have went by one simualr for years called: IPDE stratagey; Identify, Predict, Decide, Execute. We all know that organizations are not going to abandon current investments in infrastructure or resources, but are looking for ways to improve thier ROI.

    Customer Portals are a GREAT way to get started in Cloud and test the water. My only caution here is to monitor and prevent ‘sprawl’. Organizations still need to maintain control of thier infrastrucutre no matter where it resides. The ability to react (Fast Transits) is one of the best reasons to choose cloud which offers flexible, fast, scalable solutions at almost a moments notice. Management of the cloud must also be addressed early and often.

    I have found my experience with customers entering into cloud tend to follow a crawl, walk, rocket ship pattern! Quick response from the team level is vital from a response (procure) level as you indicate, and also agility in managing the new resource abilities!

    Governance of what the teams are up to (allocating, developing, testing, etc.) in the cloud will make the transition from traditional infrastrucutre and big data practices, to the cloud a much more seemless and cost effective way to go.

    Thank you for the great articles.


Speak Your Mind