Usually, a team or individual starts testing—either because it’s recognized as an opportunity at the tactical level or because it’s mandated by management—and the program evolves organically as resources, scope, and impact expand. Sometimes, this process leads to a fit that’s natural for the program and business as a whole.
More common, however, the testing program struggles to maintain efficiency and inclusivity while trying to find a place between other teams. And even if it’s easy to spot the problems within your own program’s structure, it can be hard to imagine a different approach.
This dilemma is common and testing managers have developed a number of different approaches for dealing with the problem. Here are a few of the most popular solutions:
Centralized Testing Teams
One obvious approach is the creation of a single, centralized team that oversees and executes testing across an entire organization. For this method to work, the central team must be fully staffed and have adequate access to key resources. This means a team will require, at a minimum, an analyst, developer, project manager, creative lead, and perhaps an experienced strategist.
- A central team can easily store and curate a repository of testing knowledge
- The team develops extensive experience in testing
- Velocity and efficiency can be increased easily
- Preventing test collisions is easier
- Prioritization can be based on the needs of the entire business in addition to each individual group
- Scaling a fully centralized team is very difficult
- A central team requires dedicated resources
- Total volume may decrease
- Balancing the needs of teams across a business may be challenging and political
- Cultivating an organization-wide testing culture may be more difficult
Decentralized Testing Teams
The opposite of a centralized approach is a decentralized team, in which each individual group or department manages testing for themselves. This is the model that emerges in many organizations that have adopted a grassroots approach to testing—one team starts the process, word spreads, and other departments launch their own efforts.
- Scaling the program is easy
- Each department can assign the resources necessary for their specific testing goals
- Since lots of people have the opportunity to participate in the process it is easy to build a testing culture
- Each group can prioritize tests according to their unique goals and capacity
- Volume will likely increase
- Standardization is difficult or impossible
- Collisions between tests are likely
- Process will evolve differently within each group
- Quality assurance is difficult or impossible to maintain
- Resources must be duplicated
- Test results may not be centrally archived limiting the impact of each learning
- Test velocity may be very difficult to increase
While centralized and decentralized approaches represent the poles of the organizational spectrum, there are many hybrid approaches that blend the benefits of the two. These include:
Hub and Spoke
The hub and spoke model is a clear hybrid of both centralized and decentralized testing teams. A central hub group, commonly referred to as the Center of Excellence, oversees organization-wide testing efforts, ensures collisions are prevented or minimized, and collects and manages results in a single testing repository. At the same time, each individual department can develop and prioritize tests according to their own goals.
- Balances scalability and a need for standardization
- Minimizes test collisions
- Allows for a central testing repository
- Still encourages the growth of a testing culture
- Scalability can be a challenge
- Training is required within each spoke
- Some resources will still need to be duplicated
- Teams may still have competing interests within some funnels
Within a hub and spoke model, testing can be organized in two different ways, one leaning toward a more centralized approach and one leaning toward a more decentralized approach.
Testing from the Hub
Taking a more centralized approach allows the Center of Excellence to launch and manage tests. This consolidates critical resources and testing knowledge. It also allows testing programs to actively manage factors that influence test velocity and maintain complete oversight of quality assurance assessments. With this centralized control, however, it is still a challenge to rapidly increase test volume and a testing culture may be slow to form since individual teams don’t have direct experience with the process.
Testing from the Spokes
Leaning toward a decentralized approach allows each spoke to launch and manage tests themselves. In this case, the Center of Excellence fills more of a consulting role, offering advice and guidance for testing efforts in each spoke. This approach makes scaling easier and allows programs to rapidly increase test volume. It also helps foster an organization-wide culture of testing by allowing a larger number of people to participate in the testing process. Velocity, however, is difficult to improve since each spoke must address the critical factors individually. Resources, too, will be duplicated across spokes which may pose a serious challenge in some organizations.
Which model works best for your testing program will depend on a lot of factors including organizational buy-in, available resources, and business goals. Understanding how each model functions, however, can help streamline your testing process—even if you can’t change the way your program is organized immediately.