May 5, 2023
What is Data Provisioning in Test Data Management?
If your company has taken the time to master test data generation—including steps to ensure that your test data is free from personally identifiable information, is suitable for different tests, and is representative of your data as a whole—data provisioning might feel like an unimportant step. But like a runner who trips a few feet before the finish line, companies who struggle with data provisioning will face delays and other issues at one of the last steps in the test data management process, wasting much of their hard work. The good news is that getting data provisioning right is a straightforward process, though it will require businesses to have a strong inventory of their data management needs.
What is Data Provisioning?
Data provisioning is taking prepared datasets and delivering them to the teams responsible for software testing. That process might sound deceptively simple at first, but data provisioning faces similar challenges to last-mile logistics in package delivery. Moving packages in bulk from San Francisco to Dallas on time and at a low cost is relatively easy. It’s much more challenging to achieve a low price and on-time delivery when taking those same packages and delivering them to thousands of homes across the DFW metro area.
In the same way, creating one or more high-quality datasets that help testers identify issues before launch is not that complicated, relatively speaking. But doing it when multiple teams may be testing different parts of an app, or even testing across multiple apps, can be a big lift. And if your company is using an agile software development approach , there could be dozens of different teams doing sprints, potentially starting and stopping at different times, each with its own unique testing needs. Those teams may start on an entirely new project in as little as two weeks, which means those managing your test data could receive dozens of requests a month for very different datasets.
Why Does Data Provisioning Matter?
Failing to deliver test data on time can have severe consequences. For example, a lack of test data could mean that the launch of a critical new feature is delayed, despite being essentially complete. Data that’s even a day or two late could lead to developers being pulled off their new sprints to resolve bugs revealed in testing. When that happens, other teams are potentially disrupted as personnel are moved around to keep things on track, or else the issue can potentially lead to cascading delays.
In other scenarios, the consequences could be smaller. The test data could exist, but not be stored in a way that testers can easily access. That could mean that your test data managers are spending time in a “customer service” role, where they have to spend time ensuring testers have what they need. If the friction of this process grows too large, testers might start reusing old datasets to save time, which can lead to bugs and other issues going undetected. The data provisioning challenge for businesses is ensuring that testers always have what they need, when needed, to ensure that testing catches bugs before they go live and becomes much more expensive to fix.
Strategies for Effective Data Provisioning
Does that mean that an IT-style approach is right for data provisioning? For the typical IT department, as long as there is enough capacity to support all needs on the busiest days, there won’t be any significant IT problems. However, data provisioning is significantly different from IT needs. IT needs are unpredictable, with some days having heavy demands and others producing very few requests. Data provisioning needs are tied to the development process and are nearly 100 percent predictable. Because of its predictability, companies can be efficient in resource usage for data provisioning, aiming for a “just-in-time” style process rather than maintaining excess or insufficient capacity.
Of course, achieving a just-in-time process is easier said than done. One of the most effective steps companies can take to streamline their data provisioning process is to adopt a self-service portal. While it will vary from company to company, a significant portion of test data generally needs to be reused in multiple tests. This could be for features in continuous development or applications where the data structure remains unchanged, even as the front end undergoes transformations. Enabling developers and testers to grab commonly needed datasets on their own through a portal frees up your data managers to spend more time on the strategic decision-making needed to create great “custom” datasets for more challenging use cases.
Test data sets, whether in a self-service portal or used on a longer project, need to be regularly refreshed to ensure the data they contain is up-to-date and reflective of the business. Maintaining these portals can be a very time-consuming task for your data managers. Automating the process so that this data can be regularly refreshed, rather through a request in a self-service portal or by regular updates on the backend based on rules set by the test data managers, can help ensure that data is always available and up to date.
How Mage Helps with Data Provisioning
The reality of data provisioning is that your process may not look anything like anyone else’s, and that’s a good thing, as it means that you’ve customized it to your specific needs. However, getting to that point by building your own tools could be a long and expensive process. At the same time, off-the-shelf solutions may not meet all your needs. With Mage, companies can have the best of both worlds. With its suite of powerful tools, Mage gives companies just about everything they need for data provisioning and test data management as a whole right out of the box. However, everything is customizable to a company’s specific needs, allowing you to obtain the benefits of customized software without the price tag. To learn more about what Mage can do for you, contact us today to schedule a free trial.