What do you actually do? When Women in BIM asked 461 professionals across 47 countries this simple question, the answers revealed something urgent.
There is a quiet identity crisis happening across BIM departments worldwide and it started a long time before AI entered the conversation. AI doesn’t create this weakness; it exposes it. It will automate all the work that is rule-based, not judgment-based. To survive and thrive, you must focus on your information foundations, develop critical thinking, and secure your value.
Through Women in BIM’s 2024 Global Work Survey, we asked 461 professionals across 47 countries in the AEC industry a simple question: what do you actually do?
The answers revealed something we needed to address:
- 87 different job titles appeared: BIM Manager. BIM Coordinator. Digital Construction
Lead. Information Manager. Virtual Design Coordinator. The list goes on.
- Many dramatically different titles described nearly identical day-to-day work.
- Many identical titles described completely different jobs, often with pay bands that varied by tens of thousands of pounds.
In short: there is a fundamental lack of clarity, consistency, and fair valuation of expertise within digital construction roles.
With AI entering the conversation, this problem is becoming urgent.
The big question is: what can we do about that? And why are data quality and EIR so important?
This post was written by BIM Corner’s Guest Author, Willow Williams.
Table of contents
Survey Results That Made Us Pause
First, let’s have a closer look at the survey itself. When we analysed the data, three things stood out:
- The “Manager” title is inconsistent. Our survey showed “BIM Manager” was one of the most common job titles, but the responsibilities attached to it varied widely, from leading digital transformation to producing clash detection reports. Some managed teams of 20. Others worked entirely alone. The only constant was the word itself.
- Experience doesn’t equal progression. The data highlighted that even those with postgraduate or doctoral qualifications often remain in technical roles, while others with fewer years of experience move quickly into senior positions. Something other than expertise is being valued, but what?
- Training is informal. Thirty-seven percent received no formal training for their current role. Of those who did, over half said it focused mainly on software operation rather than methodology, problem-solving, or leadership. YouTube, LinkedIn Learning, and colleagues filled the gaps.
These patterns existed long before AI. But AI makes them more visible and more consequential. To understand why, we need to look at what AI actually does beyond the vendor presentations.
What AI Actually Does
Software vendors frame AI as a productivity tool: automating repetitive tasks so humans can focus on “higher-value work.”
That’s true, but incomplete.
AI tools now appearing across prominent vendors and new startups aren’t just about clash detection or drawing production. They can recognise patterns, synthesise information, and make basic decisions.
So the question isn’t can AI automate technical tasks? It can.
The real question is: how much of our daily BIM work is rule-based and therefore automatable, versus requiring human judgment?
The uncomfortable answer: more than most job descriptions acknowledge.
Which brings us to the skills gap that was already creating problems before AI arrived and will determine who thrives as AI becomes standard.
The Skills Gap That Preceded AI (And Will Outlast It)
Over half of survey respondents said they train others as part of their role. Often it’s the technical specialists, the ones with the least influence, who carry the most current knowledge.
They’re training managers, leading software adoption, and yet sit at the bottom of pay scales.
This mismatch already created inefficiency. With AI, it becomes a liability.
The real differentiators now are:
- Critical thinking. Not skepticism for its own sake, but the ability to ask the right questions, weigh context AI doesn’t have, and make judgment calls rooted in project realities.
- AI literacy. Not just “using AI tools,” but recognising what AI can and can’t do reliably, spotting when outputs are wrong, and validating against project requirements.
Right now, many professionals struggle to evaluate human work, let alone algorithmic outputs. AI doesn’t create that weakness. It exposes it.
There is another factor that determines whether AI helps or harms your projects, one that most discussions completely overlook.
The Part We Should Talk More About: Data Quality
AI is only as good as the data it works with. If project information is inconsistent, incomplete, or poorly structured, AI won’t fix that. It’ll simply scale the problems.
Why it matters:
When humans encounter messy data, they improvise. When AI encounters messy data, it confidently delivers wrong answers that look professionally formatted. That gap between looks right and is right is where risk lives.
Where data quality usually breaks down:
- Inconsistent naming conventions across disciplines
- Metadata so incomplete models can’t be searched
- EIRs and BEPs that exist as PDFs nobody references
- Classification systems applied differently by each consultant
Sound familiar? These aren’t new problems. But they’re about to become expensive.
Which is why the most practical place to start preparing for AI isn’t with the technology itself. It’s with the information foundations AI will rely on.
Start Here: Making Your EIR Work
Exchange Information Requirements shouldn’t be documents you produce once and file away. They should be living foundations for how project information gets structured, validated, and delivered.
A practical approach looks like this:
- Build it modular. Break your EIR into clear, standalone sections: naming conventions, LOIN by stage, software and file requirements, coordination procedures, delivery milestones.
- Make it searchable using AI tools.
Upload your EIR suite into an AI-powered assistant like Microsoft Copilot, integrated with your Microsoft 365 environment. This allows project teams to query documents in natural language.
Check with your IT team to confirm which tools are approved for securely handling project documentation.
- Embed validation checkpoints. Define not just the requirement, but how compliance is verified at each stage.
- Tie requirements to roles. Don’t just write “models shall be coordinated.” Spell out who does it, with what tool, when, and how results are tracked.
- Version control with change logs. Document why requirements change and use your AI system to identify cascading impacts automatically.
This sounds like a lot. But you don’t need to rebuild everything at once. Here’s where to begin.
Five Actions You Can Take This Week
Here’s what you can do immediately. No new software required.
1. Audit one deliverable
Check one recent model against your EIR. Document gaps between stated requirements and what actually happened.
2. Interview three team members
Ask designers, modelers, and co-ordinators the same questions: Where do you waste time searching for information? What requirements are unclear or contradictory? What documentation do you wish existed?
3. Create a one-page “Quick Reference EIR”
Extract the five most commonly needed items. Make it visual. Put it where people actually work.
4. Test AI on Your Documents
Upload your EIR and BEP into your enterprise-approved AI assistant tool (e.g. Microsoft CoPilot, Gemini etc).
Ask the kinds of questions your team would, check what it gets right, what it misses, and how clearly your requirements are written.
5. Track your judgment calls for a week
Write down three decisions daily that require human context rather than following a rule.
These actions prepare you for a larger shift that’s already visible in how BIM careers are evolving and how AI will accelerate that evolution.
What This Means for Careers
Forget job titles. Focus on the work you actually do.
- How much of it is rules-based versus judgment-based?
- Do colleagues mainly value your software proficiency or your problem-solving ability?
- Are you the one training others, but with little influence to shape strategy?
- Are you the quiet enforcer of data quality in your team?
If you’re rule-based, expect automation to touch your role.
If you’re judgment-based, you’re harder to replace.
If you ensure data quality, your moment of recognition is coming.
Our survey revealed regional salary differences that compound these challenges. North America reported the highest salaries, while South America reported the lowest. These disparities highlight how economic pressures may accelerate where digital construction work actually happens.
What Comes Next
The ambiguity around BIM careers isn’t a side note to AI. It’s the central variable.
Whether AI strengthens or destabilises our industry depends on how honestly we address roles, skills, and value right now.
We can have those conversations proactively. Or we can wait until technology makes the decisions for us.
Either way, clarity is coming.
When it does, it won’t just reshape BIM. It will reshape how we understand expertise in the built environment.
This post was written by BIM Corner’s Guest Author, Willow Williams.
References
