In the last couple of years, many speaking requests and podcast invitations I received revolved around the same topic: the future of business analysis, and how it is impacted by the accelerated use of artificial intelligence (AI).
This article is a summary of many discussions and a list of resources for a business analyst seriously considering which way to look and what the right direction for their career development is.
While the article will refer to many conversations and interviews, I will present it in a question-and-answer format, as if speaking with a single interviewer.
As you read this article, listen to the podcasts, or refer to these or other resources, don’t forget to apply your critical thinking. In the face of ever-greater reliance on AI, including helpful AI overviews offered to us at every step, we must make a conscious effort to practice thinking for ourselves. Consider your strengths, competencies, interests, and aspirations in light of the overall industry and technology trends to decide what is right for you.
Let’s start with the burning question.
Can AI replace business analysts?
The answer is no. AI cannot replace a good business analyst.
We can train AI tools to complete tasks such as documentation, summarizing information, research, and data analysis. However, there is much more that is required of a good BA:
- Fostering a collaborative environment for requirements analysis
- Facilitation of thinking meetings
- Listening to stakeholders with different needs and navigating conflicting interests and agendas
- Creating a shared understanding of requirements among all stakeholders
- Applying critical thinking to all the information received as inputs, including the output from AI tools
Staying Relevant as a Business Analyst: Where Humans Outperform AI (my Medium article)
How can AI make a business analyst’s job easier?
It depends on what we mean by “AI” here. Most of the time, this question implies using Large Language Models (LLMs) such as ChatGPT, Copilot or Gemini, even though the AI concept is much broader. LLMs can take away many tedious tasks or make them easier:
- Creating business analysis documentation
- Capturing meeting notes and summarizing key points
- Extracting answers to questions from large amounts of information
- Validating text, lists, and tables against specific criteria
Freeing analysts from these tedious tasks enables them to focus on higher levels of thinking and analysis activities, determining and adjusting analysis goals, building relationships, and supporting change management and adoption activities.
To get good results and helpful drafts from an LLM model, a business analyst needs to practice and learn how to prompt, how to provide context, and how to get improved answers with additional prompts. While using generative AI tools for these purposes, business analysts must consider confidentiality and cybersecurity aspects. Using public models is not secure and may create serious problems. When using generative AI for work, always consult with your cybersecurity team first and follow your organization’s guidelines.
How will BA’s role evolve with the use of AI and other sophisticated tools?
In the past, some leaders and companies held a somewhat shortsighted view of business analysts as those who “create documentation”, or order-takers who can be given any task on a project which doesn’t have a clear owner.
On the other hand, if we are constantly occupied maintaining the backlog, updating JIRA tasks, and feeding the sprint machine, sometimes there is no time left, and business analysts remain too distant from business to help them with optimizing their processes.
When the time we need to spend on generating documentation goes down, a business analyst has an opportunity to build their expertise as a trusted advisor to the business. An advisor who understands the technology enough to help stakeholders articulate their needs and understand how technology can alleviate the pain, minimize inefficiencies and increase customer satisfaction.
To become these trusted advisors, business analysts should immerse themselves more in business – learn about the operations, understand the nuances of business processes, and become experts in their organization’s products and services.
Watch my conversation with Joe Newbert, “Dreaming About Business Analysis – The Future Business Analyst with Yulia Kosarenko“:
What is the goal of business analysis if requirements documentation can be done by AI?
The goal of business analysis was never just to create documentation.
Now, even more than ever, we have to think of the role of business analysis as increasing organizational efficiency and reducing waste.
The analyst has the training and skills necessary to observe business operations, ask the right questions to discover inefficiencies, help stakeholders analyze and agree on required changes, and describe these needs (requirements) to the party that will build the solution.
And to make it all possible, the business analyst is responsible for creating and maintaining a shared understanding of requirements by all stakeholders – the vision of the future solution.
Finally, a business analyst able to bring this value to the organization will need to apply their skills and competencies to new types of projects. These projects will be more data-driven, will make use of new emerging tech, and will incorporate different types of analytics to support business process optimization and automation.
So, business analysis is changing not only because AI tools are now available to automate some of the tasks, but also because we need to apply analysis to new types of projects.
How is business analysis changing for new types of projects aimed at implementing machine learning algorithms and AI technologies?
Of course, there is nothing new in the idea that business analysis must also cover data management requirements – every solution deals with data, every system stores, retrieves, and uses data to support its functionality, and these requirements must be understood and captured during requirements analysis.
Reporting requirements are also an old staple, even if sometimes they were added as an afterthought. Stakeholders start thinking about reports, dashboards, and monitoring performance after the main rush to figure out what the system is supposed to do is over.
However, advanced analytics requirements require a paradigm shift. Instead of having business users look for reports, navigate charts or wait for a daily progress report email, we must think about embedding analytics in the workflows.
What is needed is to deliver analytics insights to people doing their job at the right time and in the right place – embedded analytics. A business analyst must help business people determine what analytical insights they need to make better decisions, to take the right next step in the workflow, with the goal of optimizing business outcomes.
We are talking about optimization tasks and real-time analytics requiring that operational system workflows are tightly integrated with the organization’s analytics platforms.
Requirements for these solutions need to take into consideration how analytics can be generated, what factors must be collected from the context and assessed in real time, and what insight, prediction, or recommendation is needed to make the process move in the right direction.
This is where business analysts need to have a deep understanding of their company’s analytics solutions and capabilities to become trusted advisors to businesses and help them formulate their requirements.
Listen to our conversation about embedded analytics with Oleksandr Moskalyuk on the BA-Day Afterwork Talks Podcast “Improving Business Processes with AI”:
What should business analysts learn to work on data science and AI projects?
Learn more about data, data types and how data is prepared and used to develop AI solutions.
Learn how machine learning works and which types of machine learning algorithms are suitable for different types of problems.
Read about generative AI and LLM to better understand how this particular type of AI works, and to realize that generative AI is only one of many possible applications of machine learning, focused on generating content in different formats (text, video, audio, and pictures).
BA Mindset and the main BA competencies are still key and necessary when working on new types of projects.
Explore the BA Mindset concept.
Data Management & Analytics Fundamentals course
Watch our discussion with Markus Udokang, “AI for Business Analysts: with Yulia Kosarenko”.
What can help business analysts build resilience and flexibility to adapt to the changes in the job market?
The best defence against the threat of being replaced by AI is a continuous development and sharpening of your independent and critical thinking. Make it a habit to evaluate LLM outputs critically. It’s not just about having a list of the “best ChatGPT prompts for business analysis”.
What is more critical is to understand the principles on which LLM models are designed and their resulting limitations. Each type of AI solution has its areas of strength – the use cases where it can excel, and other use cases where its performance will always be inferior, by design. If you want to learn more about it, read this blog by Gary Marcus.
To be prepared for the future that relies more and more on AI technologies, you must continue to educate yourself and learn more about the disciplines that are related and sometimes tightly coupled with the development of AI solutions.
Future-Proofing Your Skills in the Age of AI (my Medium article)
Diversifying your skills is also key; while specialization can open up opportunities, it is also a risk if the particular area of your expertise becomes easily automated by emerging technology.
I have one more tip, another way for a business analyst to get invaluable experience, to gain confidence and a deep understanding of solution development that many BAs no longer have these days. This is an experience of working on a team building a new software system “from scratch”. Supporting in-house system development is a special experience. You get to observe and participate in building something new: from the first discovery discussions with the business to brainstorming solution options, battling with design challenges, coming up with the best ways to solve problems, arguing about the user experience, and discovering through testing the mistakes you made during business analysis.
I shared my experience in this long-read article: How to Design and Implement an Innovative Solution in Nine Months.
We also talked about this special experience and other ways to prepare for the future in a conversation with Deirdre Caren, “Preparing For The Business Analysis Future With Yulia Kosarenko”.
What would be your advice for someone who is not a business analyst but has strong technical and data analytics skills?
Earlier, we focused on business analysis professionals and their career prospects. What about data analysts, dashboard designers, machine learning engineers, and data scientists? While their technical skills have been much in demand, is the tide changing? Will AI tools be able to replace many coding tasks? Can ML models be trained to monitor and optimize themselves without an engineer’s supervision? Will data engineers engineer themselves out of jobs?
Of course, we can only guess. However, there are a few things that these specialists can do to strengthen their career prospects:
- Learn more about the business domain, their company’s business model. When you understand how your company makes money, you are in a better position to propose new ideas for monetizing its data and improving business outcomes through better use of data.
- Learn about business processes in the company. What is not working? What are the bottlenecks? Where are the most mistakes made? How can the process be optimized through embedded analytics and prescriptive algorithms?
- Work with business stakeholders to reframe business problems as prediction problems. What prediction(s) can help them make better decisions? What variables can be used as features in predictive models? What accuracy is needed? What will business stakeholders need to be able to trust a machine learning model (call it “AI” if this helps)?
Building these relationships with businesses is key for data people. Without stakeholder trust and acceptance, without the adoption of analytics solutions, your hard work may end up as a waste of time and effort. It is not good for either your confidence or your career.
We touched on some of these topics and more in this podcast with Ganna Pogrebna, “How to Get into Business Analytics and Business Strategy of AI: Key Capabilities”:
To explore this subject further, check out my new book:
To be continued.
Training | Mentorship | Courses | Books | Downloads | LinkedIn | YouTube |BlueSky | Medium
What a compelling read! The future of business analysis is clearly in great hands with AI, automation, and embedded analytics driving meaningful transformation across industries.