Michael T. Thieme, Assistant Director for Decennial Census Programs, Systems and Contracts, U.S. Census Bureau
Like the style of the clothes in your closet, the conventional wisdom on what works and how business or government organizations should acquire and use technology repeats in cycles.If you don’t think it will be effective to buy an off-the shelf solution from an external vendorin the current environment, just wait awhile,and building homegrown systems will again be the way to go. The same applies to the perennial question of which technologyapproach is best for your enterprise. If generalized, one-size-fits-all platforms providing shared services don’t work for you today, a new era of purpose-built applications tailored to the needs of separate internal business units may be right around the corner. Of course, none of us has time to wait for the cycle to come around to our way of thinking. We buy or build, generalize or tailor, as current circumstances require. To help avoid getting caught up in popular trends that may not serve your organization well, this article suggests a structured, data-driven approach to IT planning.
The information technology cycle has a significant influence on how organizations innovate, and the U.S. Census Bureau has a long history of innovation, both in statistics and in technology. We at Census have also grappled with the cyclical nature of technology, having delivered the Decennial Census every ten-years since Thomas Jefferson ran the first one in 1790.
Some examples: A Census employee, Herman Hollerith, invented the cardpunch machine to tabulate the 1890 Census. He went on to be one of the founders of IBM. Continuing into the twentieth century, the Census Bureau built innovations that led the government and private sector in computerization, statistical sampling, interviewing techniques, data processing, quality control, and geographic information systems. We pioneered the first large-scale use of the computer with UNIVAC 1 after the 1950 Census, and developed (in-house) optical character recognition to digitize data for the 1960 Census. We developed the first “wall-to-wall” digital street map of the United States, which gave rise to the many private sector Geographic Information System (GIS) innovations that brought us Google Maps and Waze. This long history of change and innovation provides the backdrop for many of us here at the U.S. Census Bureau, who have had the opportunity to watch and learn in recent decades as the IT cycle repeats.
The figure below suggests a spectrum of IT approaches that organizations face in our changing times, along with some high-level characteristics at each end of the spectrum. For simplicity, I use the terms Centralized and Business-Driven to describe each end of the spectrum.
Further characteristics of the Centralized orientation include things like a shared resource pool, a standards-driven approach, a high level of repeatability, a high level of structure, and processes oriented toward effective use of technology. In contrast, theBusiness-Driven approachemphasizesdedicated resource pools, mission/purpose-driven strategies, a high level of flexibility and responsiveness, and processes geared toward specific business goals.
The information technology cycle has a significant influence on how organizations innovate, and the U.S. Census Bureau has a long history of innovation, both in statistics and in technology
Each end of the spectrum offers advantages, but going too far either way can be problematic. In a purely Centralized approach, managers can become frustrated with a bureaucratic structure governed by people separated from the reality on the ground, and by one-size-fits-all solutions that do not allow for variations in customer requirements. On the other hand, a purely Business-Driven approach can lead to needless duplication of effort, higher costs and risks without increased benefit, and may even result in areas working at cross-purposes.
Additionally, where the Centralized approach can provide cost savings and greater consistency across the organization, being Business-Driven can foster speed, creativity and innovation. Certain characteristics from each end of the spectrum offer compelling advantages. I will use the rest of this article to describe a structured approach to help decide what’s best for your organization.
A Structure for Making Decisions
With an awareness of the spectrum, it makes sense to consider a number of options that fall on different points along that spectrum to frame thediscussion on which way your organization will go. There are doubtless many other organizational configurations that could be considered, but I suggest the following as a starting point.
• Option 1 – Purely Business-Driven organization: This option represents the right end of the spectrumemphasizing dedicated resource pools, mission/purpose-driven strategies, a high level of flexibility and responsiveness, and processes geared toward specific business goals.
• Option 2 – Exception-based Business-Driven organization: This option is primarily Business-Driven, but allows for Centralized exceptions that take advantage of enterprise services where it makes most sense for each particular business unit within the organization.
• Option 3 – Hybrid: Functional organization: This option falls in the center of the spectrum attempting to balance the trade-offs that increased centralization brings for the business unit and the enterprise.
• Option 4 – Exception-Based Centralized Organization: This option is primarily centralized, but allows for Business-Driven exceptions to centralization for high-profile, unique programs. This option acknowledges that supporting the unique systemrequirements of a single program could be disruptive to all other programs in the enterprise.
• Option 5 – Purely Centralized Organization: This option represents the left end of the spectrum requiring a shared resource pool, a standards-driven approach, a high level of repeatability, a high level of structure, and processes oriented toward effective use of technology.
With the options defined, we can now add assessment criteria and score each option against those criteria. An example of assessment criteria could bePeople, Processes, and Technology, with various sub-criteria that support the discussion. To demonstrate the process, I am using an analysis structure we used in early preparations for the 2020 Census, and showing only the Technology criterion. I suggest the graphic approach below, scoring elements in relation to your organization’s capabilities from “relatively less able” to meet the criteria (clear circle) to “relatively more able” to meet the criteria (filled-in circle). This technique should reflect your organization’s historical or research-based indications of relative ability. Ultimately, both filled-in and clear circle options are potentially valid choices from an organizational point of view.
Technology Options Assessment Example
This criterion seeks to address the ability of the organizational model to manage technology and adhere to standards while balancing the responsiveness to individual business area demands.
Any number of criteria can be considered and included in this approach. The point is to provide a stable foundation for the best possible discussion and decision in the context of your organization’s strengths, weaknesses, and goals – and I believe we can all use a little more stability in an IT world that, first and always, promises changing times.