As the highways were being built, Sam Walton looked at the state of the retail industry in terms of what was changing in transportation. Walton realized that trucks on reliable highways could substitute for warehouse capacity. If he could get information about what his stores were doing to a centralized location, that location could have a supplier deliver a truck with sufficient quantities to update the inventory for multiple stores. 

 

Before Walton changed everything with Walmart, retailers contracted with warehouses, or even maintained their own warehouses, that received railcars of inventory. Inventory dribbled out of the warehouse to the stores, meaning both warehouses and stores frequently ended up with stale inventory. One of the major forces in Walmart's growth was building an infrastructure that allowed inventory arriving from the supplier to be "cross-docked"; that is, inventory was unloaded from the supplier's truck directly onto trucks headed for Walmart stores. The velocity at which the inventory moved from supplier to the store meant more inventory turns (the number of times inventory is depleted and replaced), meaning Walmart's sunk costs in stale inventory decreased, meaning the volume of orders to suppliers increased, meaning Walmart was able to negotiate better pricing. This is one of the virtuous feedback cycles that led to Walmart's dominance.

We are in a similar position when looking at the new world of cloud computing. Like the Interstate Highway System, some things are obvious: we don't have to build data centers, we are renting capacity, and overall it's a little cheaper to run our systems in the cloud versus in a data center. I call that adapting to change. Exploiting change is very different.

As highways were built, companies sent their trucks on the new roadways and reaped the benefits of smooth roads that got their trucks to their destinations faster. They adapted happily. The first Walmart opened in 1962; it took a change-exploiter like Sam Walton to realize what could be done if he restructured the whole retail supply chain, creating a pipeline from the supplier all the way to the store.

In 2016, it should be obvious that you cannot exploit the cloud unless you are willing to commit to changes of this magnitude. One obvious change is to reallocate your IT infrastructure to save money.

One of the strengths of the cloud is that you are only charged for the capacity you use, and only when you use it. This means you finally have the opportunity with your computing resources to only spend money when there's an opportunity to make money, or just as importantly, when there's an opportunity to save money. 

Thus, the first question you should ask when moving to the cloud is, "What systems can I turn off?" If a system does not have to be on 24 hours a day, or available 7 days a week, turn it off when you don't need it. When you do need it, turn it back on.

Your second question in moving to the cloud should be, "Why do I need a particular system? Is this system making me money or saving me money? Will I make or save more money if I add or delete capacity?" After all, increasing revenue or saving money are the true basic criteria for any business decision. 

These two propositions--turning off capacity that isn't needed, and adjusting capacity to the optimum level--seem like simple ideas, much like cross-docked inventory. But just like the cross-dock facilities Walmart built, you must build (or rebuild) infrastructure and change your operations to reap the benefits of the cloud.

At the minimum, you must create an infrastructure for your software systems that supports or performs these activities:

Each of these requirements is equally important. For example, there may be instrumentation for a database server, and scripts that can control the database server, but traditional relational database systems can only increase or decrease their capacity by physically changing the box they are running on. So the system cannot scale automatically without interrupting its service.

You may be looking at the cloud like most retailers looked at the new highways in 1962, thinking, "Hey, I can adapt -- I can do the same things I've always done, but quicker and cheaper now." What you should be thinking instead is, "What do I need to change to exploit the cloud?"

In future posts we'll look at the new systems architectures that the cloud supports. None of this is brand new, and none of this is specifically limited to the cloud. But like anyone who really exploits change, you can combine these architectures with what the cloud makes possible to create new savings, new efficiencies, and new business opportunities. 

If you have the imagination to do it.