For most of my career, business analysts did two things at once without ever naming them separately.
We wrote documents. And the process of writing those documents created shared understanding and judgment. The interviews, the workshops, the back-and-forth on a requirement, the moment in a review when two stakeholders realized they meant different things by the same word. The document was the visible output. The understanding was the real product, and it was built in the making.
AI has pulled those two things apart.
Now a document can appear in seconds, complete and polished, with none of the conversations that used to produce it. The output exists. The understanding may not.
[Your moment: a document you wrote early in your career that was “done” and approved, where you later realized the stakeholders never actually agreed. What gave it away?]
What a Document Can Hide
Here is what I see happening more and more:
- A BRD is complete while stakeholders still disagree about the problem.
- A solution spec looks convincing while describing a solution nobody has validated.
- A detailed spec makes an assumption look like an established fact.
Each of these was possible before AI. What AI changes is how easy and fast it is to produce the polished version. A well-formatted, well-written document signals rigor. When that polish arrives without the underlying work, it can create false confidence at exactly the moment a team needs to be asking hard questions.
What Business Analysis Actually Contributes
The work was always about understanding what matters, exploring alternatives, resolving conflicting needs, and making tradeoffs explicit. The document was a container for that work.
AI can genuinely help. It can draft options, surface questions, summarize input from dozens of stakeholders, and stress test a problem statement. Those are real accelerators.
Generating an answer is a different act from establishing that it is the right answer for this business, these customers, and these constraints. That second act requires people in a conversation, evidence, and judgment. It is the part of business analysis that deserves our full attention now.
We need to get much clearer, as a profession, about what we contribute. The question I keep coming back to: are we producing evidence of understanding, or documents that look like understanding?
Tip 1: Separate the Draft From the Decision
When AI produces a draft, treat it as the start of the analysis. Label it that way. Share it as “here is a starting point to react to,” and make clear which parts are assumptions and which parts are confirmed.
A simple practice that helps: tag key statements in any AI-assisted artifact.
- Confirmed: agreed by the right stakeholders, with evidence.
- Assumed: plausible, not yet validated.
- Open: stakeholders disagree or nobody knows yet.
When a stakeholder sees how much is still assumed or open, the conversation changes. The document stops being a finish line and becomes a map of what still needs to be decided.
Tip 2: Use Prototypes and Scenarios to Test Understanding
People agree with documents they skim. They react honestly to things they can see and touch.
One of the best ways to find out whether understanding is real is to put a working prototype in front of stakeholders and walk through concrete scenarios together. What happens when a customer does this? What should the system do when that data is missing? Who decides in this exception case?
Those run-throughs surface disagreement that a written spec would have hidden. Two leaders who both approved the BRD suddenly realize they expect different outcomes. AI makes this dramatically easier, because a prototype that once took weeks can now exist in an afternoon. Use that speed to test understanding, early and often.
Tip 3: Make Tradeoffs and Decisions Visible Artifacts
If documents no longer prove understanding on their own, capture the understanding directly.
Keep a short, living decision log for any meaningful initiative. What did we decide? What alternatives did we consider? What tradeoff did we accept, and who accepted it? What evidence did we use?
This is a small artifact with outsized value. It shows the thinking behind the solution. It gives new team members and AI tools the context they need. And six months later, when someone asks why the system works the way it does, there is an answer grounded in real decisions made by real people.
The Work Behind the Words
I find this moment exciting for our profession, even though it is uncomfortable. For years, BA value was often judged by the documents we produced. That measure is fading fast, because anyone can produce a document now.
What remains is the harder, more human work: helping people understand each other, surfacing the conflicts, making tradeoffs explicit, and building judgment together. That work has always been the heart of business analysis. Now it gets to be visible.
The BAs who thrive will be the ones who can show evidence of understanding, in every room and every artifact.
Build the Practice Behind the Artifacts
My Maven course series helps BAs shift from producing documents to producing understanding, including facilitation with prototypes, scenario-based validation, and making decisions and tradeoffs visible.
Visit www.maven.com/angela-wick to explore current courses and upcoming cohorts.
Anyone can generate the document now. You can build the understanding.