Most portfolio managers are drowning in data and struggle to get any meaningful insights into the state of the projects and programmes under their control. Dashboards show you what’s happening now, reports tell you what’s happened in the past but what about the patterns hiding in plain sight?
After working with hundreds of portfolio managers over the years, I’ve noticed something interesting. The best ones don’t just collect data. They ask the awkward questions and dig into the “why” behind the numbers. Until recently though that has been a time consuming and complex exercise, with very few people having the time to do a thorough job.
With AI now capable of analysing data at scale, it’s become much easier to interrogate portfolio data to get insights that might never have come to light.
The benefits of AI for project management are well documented, and portfolio management is no exception
Here’s my list of five questions that I think every portfolio manager should be asking as well as some insights that you might not want to hear.
1.Which of my ‘green’ projects are actually in trouble?
This is the question that makes people VERY nervous but it’s one that needs to be asked. RAG status is often more art than science and project managers have every reason to keep things looking positive for as long as possible.
AI can spot the warning signs that humans miss or ignore:
- Projects with a green status but its burning through budget faster than schedule progress
- Teams consistently missing internal milestones while maintaining positive external reporting
- Resource utilisation patterns that don’t match the reported progress
- Risk registers that haven’t been updated despite clear schedule or budget pressure
By asking this question one of our customers discovered that 40% of their “green” projects were heading for significant delays. The early warning gave them time to intervene rather than react to crises.
2. What patterns predict which programmes will fail?
This is where AI really shines – spotting patterns across your entire portfolio history that would take humans years to identify or be missed completely:
- “Programmes sponsored by Department X have a 60% higher failure rate”
- “Projects over $2M with more than 8 stakeholder groups always go over budget”
- “Infrastructure programmes starting in Q4 consistently face resource conflicts”
These aren’t comfortable truths but they’re solvable problems. You can change governance structures, adjust programme timing or provide additional support based on historical patterns.
3. Are my data quality problems hiding real risks?
Having poor quality project data isn’t just a nuisance, it’s a strategic risk. Time and again it’s been shown that projects with missing or inconsistent data tend to be the ones that end up with schedule overruns and cost blowouts.
AI can identify the gaps that matter:
- Projects claiming no risks but showing clear budget or schedule variance
- Teams that consistently update some fields but ignore others (usually the important ones)
- Data patterns that suggest teams aren’t engaging with your governance processes
- Projects where the reported status doesn’t match the underlying metrics
One portfolio manager told me they found that projects with incomplete risk registers were three times more likely to miss major milestones. The data quality problem was really a project management problem in disguise.
4. Which resource allocation decisions am I getting consistently wrong?
Resource conflicts are the bane of every portfolio manager’s existence. It’s not uncommon for allocation decisions to be made based on incomplete information or political considerations rather than data.
AI can analyse your historical resource allocation patterns and outcomes:
- Which resource combinations lead to the best project outcomes
- How team composition affects delivery speed and quality
- Which skills gaps consistently cause delays across programmes
- Whether your star performers really do improve project success rates or just get credit for easier projects
The answers might challenge some assumptions about your best people and your resource allocation strategies.
5. What’s the real cost of my portfolio management overhead?
Here’s a question that keeps finance directors awake at night. How much time and money are you spending on governance, reporting and management processes – and is it improving outcomes?
AI can help you understand:
- Which governance activities correlate with better project outcomes (and which don’t)
- How much time your teams spend on reporting vs delivery
- Whether your approval processes reduce risk or just slow things down
- Which portfolio management activities provide genuine value vs those that just make everyone feel busy
One organisation discovered they were spending 30% more time on governance than their most successful competitors with no improvement in project outcomes. That’s an expensive insight but something that’s an easy fix.
The hard truth about asking the difficult questions
Asking difficult questions often results in answers that you don’t want to hear. For example, you might discover that your best programme manager is consistently choosing easy projects. Your rigorous governance processes might actually be hindering delivery without improving outcomes. Your “strategic” initiatives are no more successful than random selection.
The good thing is, once you know about these issues you can do something about them.
Getting started with AI-driven portfolio analysis
You don’t need to become a data scientist to start asking better questions of your portfolio data.
What you do need is:
Clean, consistent data: AI is only as good as the data you feed it. Start with data quality basics and remember GIGO. Garbage in, garbage out.
The right questions: Focus on questions that, if you knew the answers, would change the decisions that are made.
Courage to act on insights: There’s no point uncovering issues if you’re not prepared to do anything about them.
A safe environment to experiment: Start with historical analysis rather than live project decisions.
The portfolio managers that embrace AI and use it to support their work aren’t doing it because they love technology. They’re doing it because they want to make better decision using the data they already have,
Your data is trying to tell you something. Are you listening?
Ready to start asking better questions of your portfolio data? Our AI Analyser is designed specifically for portfolio managers who want genuinely useful feedback rather than a generic regurgitated summary of what you already know. Get in touch if you’d like to see what patterns are hiding in your data.
