What is our primary use case?
My main use case for Appium is to work on mobile applications to automate them, focusing on end-to-end functionality, end-to-end automation, and UI automations for mobile applications across different kinds of mobile applications such as hybrid, native, and mobile web app applications.
A quick specific example of how I used Appium in one of my projects is that I am currently automating a healthcare smart care application which is used for healthcare professionals, including hospital staff such as nurses and doctors, where they can communicate. I have used Appium to automate this end-to-end application functionality across all scenarios.
I have automated various workflows within that healthcare app with Appium, specifically focusing on messaging, alerts, and calling, including one-to-one calling, broadcast calls, urgent calls, and notifications.
Appium's major usefulness is for mobile application testing.
How has it helped my organization?
Appium has positively impacted my organization by increasing productivity, which was vital before the introduction of Appium MCP. Previously, we wrote all the end-to-end mobile application functionality automation manually, but with Appium MCP, our productivity has increased tenfold as it can generate scripts, debug, and auto-heal itself, reducing the time spent in scripting and debugging test cases.
I have seen improvements in test coverage because previously, writing a single test case took hours, involving understanding functionalities, navigating through Appium's DOM structure, writing page objects, converting them to behavioral attribute methods, running tests, debugging failures, and repeating the process. With the advanced Appium MCP AI agent, I can now write about 10 to 15 test cases in a single effort.
What is most valuable?
The best features Appium offers include support for different kinds of mobile applications including hybrid, native, and mobile web apps. There is no better alternative in the market as an open-source tool, which is its biggest advantage. In Appium version 3.0, many things have been made decoupled, which is really helpful with different drivers and plugins for different platforms, providing a lot of freedom to an SDET or any QA members automating mobile applications.
I find myself relying on all three great features from Appium in my day-to-day work. The decoupling of features such as plugins and drivers helps avoid dependency on a single platform, whether it be iOS or Android. Appium's open-source nature provides a level of support unmatched by competitors in the market, and its capability to support all types of mobile applications means no other tool in the market comes close to what Appium provides.
Since the introduction of Appium MCP, automation has become more robust and considerably easier, enhancing productivity significantly.
What needs improvement?
One painful point that still exists with Appium, especially in iOS mobile application testing, is its reliance on WebDriverAgent, which creates a lot of stress during setup for beginners. Although many improvements have been made since MCP, that area still needs attention. Additionally, improvements in the locator strategies are needed to adapt to modern technologies for locating mobile applications developed outside traditional Apple or core scripting technologies like Swift, Java, or Kotlin.
A community needs to emerge, driven by a set of industry leaders who can help people and contribute to the community, especially since platforms like Stack Overflow have diminished in relevance due to the rise of AI.
If I could change or improve one thing about Appium, I would focus on enhancing Appium MCP to ensure it delivers what it promises reliably.
For how long have I used the solution?
I have been using Appium since its 1.0 release in 2015 and 2016, which is quite a long time.
What do I think about the stability of the solution?
Appium is much more stable now compared to the 1.0 version.
What do I think about the scalability of the solution?
Appium's scalability for my projects has significantly increased our speed of execution and delivery, achieving up to 10x faster results, with just a few people being sufficient to write automation.
How are customer service and support?
Since Appium is open-source, I do not observe any formal customer support; however, there is a development team available who address questions online, and I appreciate their efforts.
Which solution did I use previously and why did I switch?
I have not used a different solution before Appium; it has been my main go-to tool, and I haven't utilized anything else.
What was our ROI?
I have indeed seen a return on investment, as the direct cost of tooling investment is reduced. With the introduction of Appium MCP, productivity has increased to 10x, meaning very few QA personnel are now needed to write automation.
What's my experience with pricing, setup cost, and licensing?
My experience with pricing, setup cost, and licensing for Appium is straightforward since it is open-source, and I currently have no concerns about using it.
Which other solutions did I evaluate?
In the very early days, I evaluated UI Automator by Android SDK, which only supported Android automation, but since then, I haven't looked back after adopting Appium.
What other advice do I have?
Regarding Appium's AI capabilities, I am confident about the governance and security it provides. It complies with our test data and governs everything effectively. However, when using Appium MCP, there can be concerns about sensitive data being shared through client scripting, which needs to be managed carefully.
From my current experience with Appium's AI capabilities regarding accuracy and reliability, I find the AI-generated scripts and suggestions have reached a stage where they can be reasonably relied upon.
My advice to those looking into using Appium is that if you are seeking an open-source tool to automate mobile applications of any type within your organization or projects, Appium is an excellent choice as it supports cross-platform automation. I would rate this product a 9 out of 10.
Which deployment model are you using for this solution?
On-premises