It is difficult to explain my opinion on IBM DOORS Next; the usability is not as good as I expected, and it is very complex and complicated. It is not a bad tool if you understand how it works, but from the perspective of engineers who only use IBM DOORS Next approximately several times a month but not permanently, it is not very comfortable or intuitive to use. The implementation, migration, and configuration need more user-friendly usability, perhaps through on-site guidance or intuitive use with push button functions, which might be more comfortable, because at the moment, it looks very complex, and ordinary engineers often mention that they have to work with this tool but would not choose to. Simplifying IBM DOORS Next would not be a bad idea. From my perspective and connections with friends at IBM in Switzerland, I gain access to very good background information that helps me satisfy my clients. However, if I had not had these contacts, I might have felt lost inside the tool chain. I am really satisfied as long as I can get help, but I believe it would be a great benefit if the tool itself offered more intuitive push-button functions and similar enhancements. The pricing of the tool itself does not actually matter because the power, performance, and accuracy of this tool are excellent, and that is not the point of contention. All clients agree that the tool is not bad, but the complexity is an issue since it creates a situation where you feel lost while working with it. The intuitive usability that we learned from Classic DOORS is simply not the same. I understand that the complexity has grown, yet I believe it would not be a bad idea if IBM considered splitting or breaking down IBM DOORS Next into two options or, better yet, developing a modular architecture that suits smaller and mid-sized projects. For larger projects with a lot of subsystems, it makes sense to use the full range of the tool, but for startups or mid-sized companies, it would be beneficial if they could select modules according to their needs. More visible on-site automatic help would be beneficial. For instance, if you need to move something, as you use the mouse cursor, an automatic message could pop up asking what you would like to do so that you can select within that context, and it would automatically perform the task. Modern software development recognizes that this type of modifying usability makes life much easier for users. Many have mentioned that whether it is Rhapsody, DOORS, or IBM DOORS Next, the issue is they work only a few times a month and are not professionals with these tools, which leads them to contact me for assistance. It would not be a bad idea for IBM to make this tool more handy, efficient, and user-friendly since most users do not work full days or even months on these tasks and are not familiar with the complete usability.
Senior Mechanical Design Engineer at Pricol Limited
Real User
Top 10
Jun 12, 2025
We are not interested in integration with lifecycle management tools. If required, we will connect to the local India team for dashboard reporting tools or additional features. I am not an exact user or end user, so I would need to check with my team regarding drawbacks and improvements.
The improvement for IBM DOORS Next is related to the exportation of documents. We had to install an additional tool to export Word documents. We needed additional tools, and the whole exploitation process was also difficult for us because we used Word templates that we integrated into the process so that IBM DOORS Next included it in the export process. That was not ideal. When I use Jira, I just need to go to the menu and click to export, and then I have all the things that I want. That is the big difference. It depends on what you want to do. In the past, we worked in the classical way, and today we are working in Agile. In my opinion, IBM DOORS Next does not have any Agile support, and that is why for requirement analysis, IBM DOORS Next is correct, but for someone who is working in an Agile process, IBM DOORS Next is not the solution because it is not integrated into the Agile working process.
There is a need for improvement in user experience, as the UI is too complex and outdated. Updates should be more frequent to ensure security and functionality, especially in addressing vulnerabilities. The lack of built-in AI functionalities is noted, with existing solutions not officially supported.
System Engineer at a tech services company with 10,001+ employees
Real User
Top 5
Apr 12, 2024
IBM Rational DOORS Next Generation has room for improvement compared to other tools like Polaris and Jama Connect. These tools offer more flexibility and options for developers, which IBM Rational DOORS Next Generation lacks. For example, you can define your link rules in Jama Connect, but you can't do that in IBM Rational DOORS Next Generation.
Senior Technical Product Manager at a manufacturing company with 10,001+ employees
Real User
Top 10
Dec 26, 2023
IBM Rational DOORS Next Generation is not a very user-friendly product. From an improvement perspective, IBM Rational DOORS Next Generation needs to be made a very user-friendly product.
Project System Engineer at a aerospace/defense firm with 10,001+ employees
Real User
Dec 13, 2022
It does have a tendency to condense the requirements. It kind of puts them in a tree format. Sometimes those trees are a little difficult. Unless you're used to using it daily, how it truncates everything and kind of rolls things up can be a little bit challenging. That said, I don't know if there's an easier way to do that.
The user interface of IBM Rational DOORS Next Generation needs to be improved, there is a lot of scope with this. Even though there have been improvements from versions five, six and seven, the UI still has a lot of bugs. There is room for improvement in the APIs that they have exposed for integration. The solution relies on OSLC for integration APIs, but those APIs do not support all the capabilities.
System Engineering Manager at a wholesaler/distributor with 10,001+ employees
Real User
Aug 18, 2022
I would say the only additional feature would be if it had dynamic linking to other MBSE tool sets or industry-leading tools. That way you could have that link and not have to do a manual insert or translate the data to an Excel-based product and then use the Excel to load into DOORs.
Technical Sales Specialist at a computer software company with 1,001-5,000 employees
Real User
Jan 4, 2022
This solution is currently having some performance issues because they changed the data storage in the back between version 6 and version 7, so it needs some performance improvement. The data storage is the relational database in the background, and it has changed and needs performance improvement. Performance is a very big thing that is needed when people are writing requirements. They have to wait too long for the page to return and then they don't like it. They are looking for different ways to do and complete their work. Good performance is very essential in this tool, same for DOORS Classic. There is also a lot of possibilities to improve reporting.
Software Engineering Consultant at a manufacturing company with 1,001-5,000 employees
Consultant
Jun 18, 2020
It offers a bad user experience and the usability is poor. A lot of engineers do not use it, primarily for two reasons. First, it is very abstract and hard to understand. Second, the operation seems to be very complex. For example, the setting is distributed in a lot of section and it is not easy to find where to configure it. The baseline is not as good as it is in DOORS because you can't create a baseline for a module. Rather, you can only create on for the project. The speed of technical support needs to be improved. The DOORS Web Access (DWA) is not stable.
Associate Director Systems Engineering & Safety Assurance at AECOM Technology Corporation
Real User
Apr 2, 2020
I find the tool clunky, user experience is very dependent on the configuration of the tool for which you require a facilitator, and the setup/configuration fees can be expensive. Overall, it seems like DOORS Rational and DOORS NG, have been written by software developers for software developers. They haven't thought about the standard project team that is actually going to use the system. In comparison ComplyPro it's completely different. ComplyPro interface is easier to use, it's almost like using a spreadsheet and is more intuitive. The whole layout of DOORS could be improved, it's too rigid and requires professional support to change. DOORS is very difficult for novice users to use. For example its common to export requirements to a spreadsheet for designers to insert verification / validation evidence and then import the data, this creates extra work for our requirements management team that could be avoided. DOORS needs to be a more intuitive product. It needs more spreadsheet-like features. The DOORS team need to understand how real world users are using the tool and make configuration and data entry easier.
CEO at a manufacturing company with 10,001+ employees
Real User
Dec 4, 2019
The tool is quite new, so it may not be a mature product, but I haven't spent enough time with it to have a list of improvements that it needs. When you are in JIRA or Confluence, you have some freedom in how you type in the text. That's also a weakness of Confluence, however, as it opens the doors to sloppy work. In DOS Next Generation, the text is very rigorous, but it might be difficult for people who don't have the discipline. Having a way to quickly enter requirements could help. It might already be in there, but I don't know. I don't have enough experience with the tool yet.
CIO at a comms service provider with 501-1,000 employees
Real User
Aug 12, 2019
An area that could be improved is user interface which is not very intuitive. Product is also Java based and we have a lot of issues with performance (at beginning). Resolution of all problems takes a lot of time. ALso upgrades of product are from our experience very painful.
IBM Engineering Requirements Management DOORS Next (DOORS Next) is a requirements management (RM) software designed to help organizations manage engineering project requirements throughout the development lifecycle. It offers a central location for capturing, defining, and organizing project requirements, facilitating collaboration among stakeholders like engineers, system designers, and customers. Key features include requirements traceability, version control, and impact analysis. DOORS...
It is difficult to explain my opinion on IBM DOORS Next; the usability is not as good as I expected, and it is very complex and complicated. It is not a bad tool if you understand how it works, but from the perspective of engineers who only use IBM DOORS Next approximately several times a month but not permanently, it is not very comfortable or intuitive to use. The implementation, migration, and configuration need more user-friendly usability, perhaps through on-site guidance or intuitive use with push button functions, which might be more comfortable, because at the moment, it looks very complex, and ordinary engineers often mention that they have to work with this tool but would not choose to. Simplifying IBM DOORS Next would not be a bad idea. From my perspective and connections with friends at IBM in Switzerland, I gain access to very good background information that helps me satisfy my clients. However, if I had not had these contacts, I might have felt lost inside the tool chain. I am really satisfied as long as I can get help, but I believe it would be a great benefit if the tool itself offered more intuitive push-button functions and similar enhancements. The pricing of the tool itself does not actually matter because the power, performance, and accuracy of this tool are excellent, and that is not the point of contention. All clients agree that the tool is not bad, but the complexity is an issue since it creates a situation where you feel lost while working with it. The intuitive usability that we learned from Classic DOORS is simply not the same. I understand that the complexity has grown, yet I believe it would not be a bad idea if IBM considered splitting or breaking down IBM DOORS Next into two options or, better yet, developing a modular architecture that suits smaller and mid-sized projects. For larger projects with a lot of subsystems, it makes sense to use the full range of the tool, but for startups or mid-sized companies, it would be beneficial if they could select modules according to their needs. More visible on-site automatic help would be beneficial. For instance, if you need to move something, as you use the mouse cursor, an automatic message could pop up asking what you would like to do so that you can select within that context, and it would automatically perform the task. Modern software development recognizes that this type of modifying usability makes life much easier for users. Many have mentioned that whether it is Rhapsody, DOORS, or IBM DOORS Next, the issue is they work only a few times a month and are not professionals with these tools, which leads them to contact me for assistance. It would not be a bad idea for IBM to make this tool more handy, efficient, and user-friendly since most users do not work full days or even months on these tasks and are not familiar with the complete usability.
We are not interested in integration with lifecycle management tools. If required, we will connect to the local India team for dashboard reporting tools or additional features. I am not an exact user or end user, so I would need to check with my team regarding drawbacks and improvements.
The improvement for IBM DOORS Next is related to the exportation of documents. We had to install an additional tool to export Word documents. We needed additional tools, and the whole exploitation process was also difficult for us because we used Word templates that we integrated into the process so that IBM DOORS Next included it in the export process. That was not ideal. When I use Jira, I just need to go to the menu and click to export, and then I have all the things that I want. That is the big difference. It depends on what you want to do. In the past, we worked in the classical way, and today we are working in Agile. In my opinion, IBM DOORS Next does not have any Agile support, and that is why for requirement analysis, IBM DOORS Next is correct, but for someone who is working in an Agile process, IBM DOORS Next is not the solution because it is not integrated into the Agile working process.
There is a need for improvement in user experience, as the UI is too complex and outdated. Updates should be more frequent to ensure security and functionality, especially in addressing vulnerabilities. The lack of built-in AI functionalities is noted, with existing solutions not officially supported.
IBM Rational DOORS Next Generation has room for improvement compared to other tools like Polaris and Jama Connect. These tools offer more flexibility and options for developers, which IBM Rational DOORS Next Generation lacks. For example, you can define your link rules in Jama Connect, but you can't do that in IBM Rational DOORS Next Generation.
IBM Rational DOORS Next Generation is not a very user-friendly product. From an improvement perspective, IBM Rational DOORS Next Generation needs to be made a very user-friendly product.
It does have a tendency to condense the requirements. It kind of puts them in a tree format. Sometimes those trees are a little difficult. Unless you're used to using it daily, how it truncates everything and kind of rolls things up can be a little bit challenging. That said, I don't know if there's an easier way to do that.
The user interface of IBM Rational DOORS Next Generation needs to be improved, there is a lot of scope with this. Even though there have been improvements from versions five, six and seven, the UI still has a lot of bugs. There is room for improvement in the APIs that they have exposed for integration. The solution relies on OSLC for integration APIs, but those APIs do not support all the capabilities.
I would say the only additional feature would be if it had dynamic linking to other MBSE tool sets or industry-leading tools. That way you could have that link and not have to do a manual insert or translate the data to an Excel-based product and then use the Excel to load into DOORs.
This solution is currently having some performance issues because they changed the data storage in the back between version 6 and version 7, so it needs some performance improvement. The data storage is the relational database in the background, and it has changed and needs performance improvement. Performance is a very big thing that is needed when people are writing requirements. They have to wait too long for the page to return and then they don't like it. They are looking for different ways to do and complete their work. Good performance is very essential in this tool, same for DOORS Classic. There is also a lot of possibilities to improve reporting.
It offers a bad user experience and the usability is poor. A lot of engineers do not use it, primarily for two reasons. First, it is very abstract and hard to understand. Second, the operation seems to be very complex. For example, the setting is distributed in a lot of section and it is not easy to find where to configure it. The baseline is not as good as it is in DOORS because you can't create a baseline for a module. Rather, you can only create on for the project. The speed of technical support needs to be improved. The DOORS Web Access (DWA) is not stable.
I find the tool clunky, user experience is very dependent on the configuration of the tool for which you require a facilitator, and the setup/configuration fees can be expensive. Overall, it seems like DOORS Rational and DOORS NG, have been written by software developers for software developers. They haven't thought about the standard project team that is actually going to use the system. In comparison ComplyPro it's completely different. ComplyPro interface is easier to use, it's almost like using a spreadsheet and is more intuitive. The whole layout of DOORS could be improved, it's too rigid and requires professional support to change. DOORS is very difficult for novice users to use. For example its common to export requirements to a spreadsheet for designers to insert verification / validation evidence and then import the data, this creates extra work for our requirements management team that could be avoided. DOORS needs to be a more intuitive product. It needs more spreadsheet-like features. The DOORS team need to understand how real world users are using the tool and make configuration and data entry easier.
The tool is quite new, so it may not be a mature product, but I haven't spent enough time with it to have a list of improvements that it needs. When you are in JIRA or Confluence, you have some freedom in how you type in the text. That's also a weakness of Confluence, however, as it opens the doors to sloppy work. In DOS Next Generation, the text is very rigorous, but it might be difficult for people who don't have the discipline. Having a way to quickly enter requirements could help. It might already be in there, but I don't know. I don't have enough experience with the tool yet.
An area that could be improved is user interface which is not very intuitive. Product is also Java based and we have a lot of issues with performance (at beginning). Resolution of all problems takes a lot of time. ALso upgrades of product are from our experience very painful.