The value provided by enterprise architecture (EA) is in the identification and understanding of the links between operational execution and strategic intent and their influence on business outcomes. In fact, better business outcomes are the only reason to take on something like EA; any other focus detracts from the main purpose of the organization. In this Advisor, I use “enterprise architecture” to mean both business architecture as well as the traditional enterprise-wide technology architecture.
There is now general agreement that EA, as a practice with a set of tools, has value and the conversation has shifted to how to measure it. Since the main purpose of investing in EA is better business outcomes, it makes sense to start there. Measuring changes in business outcomes alone isn’t sufficient, however, as an increase in gross margin could have been the result of EA work, some other cause, or some combination. This is the age-old problem of correlation not equaling causality. We need a way to trace business outcome changes back to work done under the banner of EA. It turns out that such tracing isn’t as difficult as it seems.
All EA work results in change projects being added to the investment portfolio, which are then executed as time and funding permit. These projects, because they are the work of enterprise architects, change one or more structural measures of various processes used by the business. Five measures are common to all processes: capacity, activity cost, defect rate, cycle time, and the cost structure (fixed versus variable costs), and all require investment in the process to change. A project must change at least one of these values for at least one process, usually by changing the underlying capability, in order to have any hope of changing a business outcome.
Let’s look at an example of how that might work for each of the five structural measures.
Project 1 increased the transaction capacity for the online sales channel for your business. You know from prior research that as online sales grow, the percentage of packages returned for credit grows as well, and you expect it to reach nearly 50% by the time you achieve your sales targets. (Note to readers: this actually happened, so watch for it). This means that in order to achieve your growth target of doubling your online sales, you need a four to five times increase in capacity to support the transaction volume. After the project, sales continue to grow so that you finally achieve your sales target two years after completion. The growth in sales is directly linked to the change in capacity, and in fact could not have been attainable without it.
Project 2 decreased the time for one of the activities in the payroll department to free up resources for other activities without increasing the size of the department. The activity in question was executed many thousands of times per month, so even a small change made a big difference. The project resulted in a decrease in the time for the activity by about two minutes per instance, or about 20%. Over the course of a month, that 20% reduction netted over 1,000 hours now available for other activities. If the project had not been done, new resources would have been required for those other activities, so the benefit derived from EA in this case is the time-value, in dollars, of those 1,000+ hours each month.
Project 3 reduced the defect rate of one of the key processes by 30%. Fewer defects meant a number of additional changes could be made: reduced customer support staff overtime by 80%, reduced warranty costs by 50%, and an increase in customer satisfaction of about 5%, which translated into 3%-4% additional sales. The value of this project is the sum of all of the cost savings and the potential new revenue, which just goes to show you how much defects actually cost.
Project 4 reduced the cycle time for processes that handle custom orders but had no effect on non-custom orders. The margin on custom orders is about three time what it is for other orders, so an increase in custom orders would be a huge benefit. It turns out that by reducing the time required to deliver custom orders, the volume of custom order increased by 20% almost immediately because of latent demand for custom orders, resulting in $1 million per month in new margin. Like Project 2, the increase in custom orders simply would not have happened had the cycle time not been reduced, so the entire $12 million per year in new profit is due to the project.
Project 5 modified several of your manufacturing processes to trade variable costs for fixed costs, with the result that for the current level of activity, total costs (fixed plus variable) were reduced by 15%. Projections indicate that activity will increase by 100% over the next two years. At that new level of activity, the difference in total costs is more than 60%. This difference in total costs is a direct result of, and therefore a benefit of, the project. In fact, had the project not been done, the increase activity would not have been profitable and could not have been supported.
The architects among you will recognize a two-step pattern: first, determine the change in structural performance measures resulting from a change project, then determine the business outcomes that are either caused by or allowed by the structural changes. These two steps will always lead to giving credit where credit is due, regardless of the prowess of management or changes in the economy. After all, even a great executive cannot perform better than the capabilities allow.
It isn’t difficult to see that even a small EA project can have huge impacts on business outcomes. In a large corporation, it isn’t hard for a single enterprise architect to generate 10 times their annual salary or more, every year. I once asked the VP to whom our EA group reported if we could work on commission, maybe 1% of the resulting cost savings or new revenue resulting from our work. He was intrigued, but after a brief conversation, decided it would mean too much money and rejected the idea. He was right; after just one year, I, myself, had generated enough cost savings to retire a multimillionaire. The other members of my team had similar results. I haven’t given up on that idea, as you can well imagine.