Testing of information systems and applications plays an important role in bringing a quality end product to the market. Even well-designed and well-developed systems inevitably have shortcomings or errors, and testing is responsible for finding and eliminating them. The complexity of systems, integration, resource constraints and deadlines force QA to seek ways to address these challenges. The opportunities offered by test automation may well be the solution to these challenges, but what needs to be considered when implementing test automation.

Here are a few tips to keep in mind to avoid the pitfalls of test automation.

1. Take time to plan and prepare

The quality of an IT project is best ensured when sufficient time is spent understanding the customer’s business processes and planning the project to be developed. A successful testing strategy and implementation is often created in projects where testing is involved from the project preparation stage. The preparatory phase includes review and consideration of existing tools and the adequacy of the expertise and resources dedicated to testing. It is also important to consider the phasing and timeliness of manual and test automation.

2. Choose your technology carefully

There are several technologies suitable for test automation. The systems, applications and environment to be tested will influence which testing tools and technologies are best suited to the IT project in question. If these requirements are not well understood, the benefits of test automation may be partial or, in the worst case, it may remain a mere “network weight”.

The choice must also take into account the versatility and continuity of the technology. When an ongoing project comes to an end, the production phase is entered, when the test automation works best for running regression tests. Changes should be planned and implemented in such a way that they are easily transferred to the regression test systems that can be maintained. Particular attention must also be paid to maintaining regression tests so that they do not “degrade” (remove unnecessary tests) over time.

Technology choices for test automation should also follow in the development of systems. Ideally, test automation can be integrated into the system development process, so that quality assurance and test automation solutions are an integral part of the whole system development life cycle.

3. Test to standards, but leave room for creativity

Testing should be carried out in accordance with industry standards. At the same time, independence is enabled. By adhering to standards, solutions are widely used and expertise is widely distributed – among many different actors. Nor is the solution dependent on a particular test automation supplier.

Testing is perhaps less often the phase of an IT project that innovation is traditionally associated with. However, test automation offers a huge range of solutions and opportunities where there is room for innovation and it is worth using them. By actively monitoring the test automation market and exploring the opportunities offered by new solutions, they can provide cost-effective, and even off-the-shelf, solutions to achieve your goals.

4. Don’t forget documentation

Documentation related to test automation must be provided for both development projects and maintenance. When a development project is finally completed, the testing documentation must be at a sufficient level to ensure that the test automation can be used and maintained by new test automation experts or by a change of supplier. Good and comprehensive documentation will ensure that tests are performed and maintained, and that there are no additional disruptions or interruptions to the service.

5. Map the savings that will be made

The benefits of test automation are many, so it can also be sold to the customer without much thought being given to the use of the technology or the payback period. Therefore, when designing test automation, it should be considered how much economic value it will ultimately add.

It is difficult to put a direct monetary value on the quality of a system or application, but when you consider that all changes and maintenance cost money, you can calculate a rough efficiency ratio for quality and testing. This already gives an idea of the profitability of test automation and the investment in it. As a rule of thumb, the larger the size and complexity of the system, the greater the savings that test automation can achieve.

Complexity increases the number of testing steps, increasing the robot’s strengths in both speed and quality compared to humans. In any case, the design and implementation of test automation is preceded by manual testing in several stages, followed by the construction of the test automation itself, which also needs to be tested. Naturally, there are costs associated with all of this.

Markku Kestikievari
Leading Testing Consultant

Markku is an experienced testing professional. He has worked in various testing roles in customer projects and has led several testing projects. Markku has decades of extensive experience in the IT industry.