Your Deltek Utilization Is Lying to You

Your utilization dashboard says 61%. You've benchmarked it. You've reported it. You may be managing your firm carefully— and the number might still be measuring something other than your firm's actual performance.

That's not a Deltek problem. It's a data problem. And when you're reading a number from inside the system that generated it, you can't always tell from that view whether the data underneath it is reliable. The dashboard is only as honest as what feeds it.

The median billable utilization rate across U.S. architecture firms is 61%, meaning nearly 40% of paid labor hours generate zero revenue². That benchmark is real. Whether your 61% reflects the same underlying discipline as the firm that calculated a true 61% is a different question entirely.

Deltek itself is candid about this. Inaccurate time tracking costs professional service firms more than $60,000 per year per employee¹— not because the software is broken, but because the data feeding it usually is.

Three structural problems undermine AEC utilization numbers:

• Time entry accuracy— hours logged late, reconstructed, or skipped entirely
• Definitional inconsistency— "billable" means different things to different project managers on the same team
• Feature underdeployment— most firms run Deltek's financial core while leaving resource planning and forecasting modules untouched

Before looking at solutions, it helps to understand exactly where the number breaks down. There are three distinct places. Once you can name which one applies to your firm, there's a three-question audit you can run this week— in Deltek, using what you already have— that will tell you whether your utilization data is worth trusting.

Why AEC Utilization Metrics Break Down

Each root cause compounds the others— inaccurate hours flow into inconsistently-defined categories and produce a figure no one is planning against. Nearly 40% of professional services firms report challenges with time tracking compliance and accuracy⁵— if you haven't audited yours, there's a real chance you're in that group.

Root Cause 1— Time Entry Isn't Happening Right

The most common reason Deltek utilization numbers are wrong is simple: people aren't logging time accurately. Delayed entries, reconstructed timesheets, and rounding mean the hours flowing into your dashboard reflect memory, not reality.

Think about the tasks that don't get logged. The five-minute client call. The quick document review before sending. None of these feel worth stopping to record in the moment. They add up. Monograph research shows that firms lose 15 to 25% of billable hours annually to delayed or inaccurate entry— and for a 10-person team billing at $150 per hour, a 15% shortfall translates to over $200,000 in unrealized revenue per year³.

Deltek acknowledges this directly. Manual and legacy systems struggle to accurately capture the time employees spend on each project— and this is a root cause of utilization inaccuracy, not a configuration issue¹. The system isn't failing. The discipline around it is. And just because batch entry on Friday afternoon is easier than logging as you go doesn't mean it produces accurate data.

The most common time entry failure modes in AEC firms:

• Batch entry on Friday afternoon— reconstructing a week of work from memory systematically undercounts small-increment tasks
• Rounding to the nearest hour— individually minor, collectively enormous across a team
• Skipping non-project administrative time— leaving overhead hours uncategorized creates false utilization signals

As Monograph puts it, the enemy is reconstruction³. The increments that get lost are individually trivial and collectively enormous. No dashboard feature fixes that.

If you're tracking the hidden costs of AI projects in your firm, untracked time is the same class of problem— invisible erosion at the margins that looks fine from the top until someone pulls the numbers.

Root Cause 2— Everyone Defines "Billable" Differently

Utilization is billable hours divided by available hours— but both variables are defined inconsistently across most AEC firms. When "billable" means different things to different project managers, your aggregate utilization rate is averaging apples and oranges.

Kantata, which specializes in professional services analytics, puts it plainly: while the formula for utilization rate seems straightforward, the nuances in defining both "billable hours" and "available hours" can significantly impact calculations. The operational definitions you use for the variables are what truly matter⁶.

Three definitional inconsistencies that show up regularly in AEC firms:

• Business development time— some teams code BD as overhead, others as non-billable, others omit it entirely
• PTO and holidays in the denominator— whether these are deducted from "available hours" changes the utilization percentage without changing a single hour of real work
• Administrative overhead— firm meetings, training, unbilled coordination time all land differently depending on who's entering

But there's a deeper technical issue. Stambaugh Ness, an accounting firm with a dedicated AEC practice, is direct: utilization rate calculation must be based on cost from the general ledger (the firm's accounting system of record) instead of hours from a timesheet⁴. For a salaried exempt employee who works 45 billable hours in a 40-hour week, hours-based utilization shows 112%. That's a number, not a measurement. Cost-based utilization— derived from actual labor cost data— is the accurate method for AEC firms with salaried staff.

Most firms default to hours-based because it's easier to track in Deltek. That choice can systematically overstate or understate actual performance depending on how overtime is handled.

Try this right now: ask five project managers to write down, in one sentence, what counts as billable time. Collect answers independently. The variance will tell you more about your data quality than any report.

Root Cause 3— You're Not Using the Whole System

Deltek's utilization data is only complete when the resource planning and forecasting modules are actively in use. Most firms run Deltek for financials and project accounting while leaving capacity management features underdeployed— which means they're measuring billable time but not planning it.

This isn't a software critique. Deltek Vantagepoint implementation commonly takes 6 to 12 months⁸. By the time a firm is live and stable, there's understandably little appetite for another sprint. Resource planning modules get pushed to "phase two." Phase two rarely arrives.

The Deltek Clarity Study— Deltek's annual AEC industry benchmark— names this directly: organizations face the defining challenge of closing the gap between adoption and execution⁹. The gap is the space between having a system and using what it's actually built to do.

Commonly underdeployed features in Deltek implementations:

• Resource planning— forecasting which staff are available for upcoming projects before the pipeline commits
• Capacity forecasting— predicting utilization patterns before projects stack or gap out
• Utilization dashboards— reporting that surfaces performance trends over time, not just point-in-time snapshots

Without forecasting active, utilization is entirely backward-looking. And 59% of professional services leaders struggle to predict resource needs in advance⁷— they aren't missing a better dashboard, they're running a system they never fully turned on. Running Deltek without its capacity modules keeps you in that group. And it's the easiest number in this article to change.

The adoption pattern here isn't unique to Deltek. Building a culture that actually uses the tools covers why the default drift toward minimum viable usage is universal— and what actually reverses it.

How to Audit Your Own Utilization Data

You can assess the reliability of your Deltek utilization data this week with three questions. The answers won't fix the problem— but they'll tell you which problem you actually have.

Utilization rate measures how time is allocated, not output or quality⁷. A clean utilization number is a precondition for good decisions. It isn't a decision itself. Nearly 40% of professional services firms have time tracking accuracy problems⁵— which means if you haven't audited yours, you may be one of them.

1. What percentage of your team's hours are entered within 24 hours of the work?

Pull a Deltek report on entry timestamps vs. work dates for the past 30 days. This report exists in Vantagepoint. If more than 20% of entries are logged 48+ hours after the work occurred, time entry fidelity is your primary problem. That's where to start.

2. Ask five project managers to define "billable time" independently.

Don't brief them first. Collect answers without letting them compare notes. If there are meaningful differences in how they categorize business development, administrative coordination, or overhead tasks— your teams are computing utilization differently without knowing it. You're averaging incompatible numbers.

3. Which Deltek modules are generating regular operational reports?

Ask your Deltek administrator. If the answer is primarily time entry and billing— and if resource planning and capacity forecasting are not producing weekly or monthly reports that inform staffing decisions— you're running the system at a fraction of its diagnostic capability.

Most firms that run this audit find at least one of these issues. Finding all three is common. They compound: inaccurate hours feed into inconsistently-defined categories that produce a utilization figure nobody is planning against. That's when a number becomes a habit.

Where AI Fits Into This Problem

AI tools can address all three root causes— but only after you understand which ones you're dealing with. The audit comes first.

According to Rocketlane's analysis of McKinsey's 2024 State of AI research, 67% of professional services firms have implemented at least one AI-powered workflow, compared to 51% across all industries. But only 18% of organizations formally track ROI from AI implementation¹⁰. That gap should be instructive. Most firms adopt AI tools without knowing whether they're working. And that's the same structural problem as Deltek— tools that are present but not producing clean signal.

Done right, AI addresses the utilization accuracy problem at three points:

• Automated time capture— calendar integration and mobile tools log small tasks as they happen, eliminating reconstruction; Deltek's own ZeroTime product is positioned for this, with third-party integrations available for Vantagepoint
• Intelligent classification— AI can flag time entries that appear miscoded, surfacing definitional inconsistencies before they corrupt weekly reports
• Utilization forecasting— AI can predict utilization by resource or project based on historical patterns and current pipeline, turning the metric from a rearview report into a planning tool

But the caution stands: AI doesn't compensate for absent discipline. If weekly batch entry and inconsistent definitions persist, AI tools inherit those problems. Fix the foundation first.

For firms thinking through how to measure whether AI tools are actually working in their practice, the same logic applies: you need a reliable data foundation before any tool can tell you something true. And when you're ready to think about deciding where AI fits in your practice, the utilization audit is a natural starting point— it tells you where your data is clean enough to act on.

FAQ

These questions come up regularly when AEC firm leaders start auditing their utilization data.

What is a good utilization rate for an AEC firm?

The median across U.S. architecture firms is 61%². Top-quartile firms achieve 82.4%; bottom-quartile firms average 71.1%². That 11-point gap is significant— and it suggests most firms in the middle don't fully understand what's driving their number. If you're sitting at 61% and haven't audited the data quality underneath it, you don't actually know whether you're performing at the median or just measuring at it.

How do I know if my Deltek utilization data is accurate?

Run the three-question audit in this article: check what percentage of hours are entered within 24 hours of the work, ask five project managers independently to define "billable time," and confirm which Deltek modules are generating regular operational reports. If any of the three reveal problems, you've found the source of your data quality issue.

What's the difference between hours-based and cost-based utilization?

Hours-based utilization divides billable hours by total hours— which can show above 100% for salaried staff working overtime. Cost-based utilization uses actual labor cost from the general ledger, which is the more accurate method for AEC firms with salaried employees⁴. Most firms default to hours-based because it's easier to track; the result can be a figure that systematically overstates or understates actual performance.

How much revenue could we be losing to time tracking errors?

For a 10-person team billing at $150 per hour, a 15% shortfall from delayed or inaccurate time entry translates to over $200,000 in unrealized revenue per year³. At the firm level, this scales to seven figures for mid-sized practices. The math isn't dramatic— it's compounding loss that accumulates without appearing in any report until someone pulls the entry timestamp data.

Conclusion

If your Deltek utilization rate is 61%, that number may be accurate. But if your team logs hours in batches, if "billable" means different things across projects, or if resource planning modules are sitting unused— it's measuring something other than your firm's actual performance.

Deltek's own research is candid about where the problem lives. The system is honest. The data feeding it often isn't. And the three root causes— time entry fidelity, definitional consistency, feature adoption— are all addressable once you know which one is actually yours.

If the audit surfaced issues in all three areas, that's not a sign of a broken firm— it's a sign the data layer hasn't been treated as infrastructure yet. That's a fixable problem. Building the AI layer on a more reliable data foundation is exactly the conversation I do best at Dan Cumberland Labs. Worth a conversation.

References

1. Deltek, "The Hidden Impact of Inaccurate Time Tracking" (2025)— https://www.deltek.com/resources/articles/hidden-impact-of-inaccurate-time-tracking/
2. Deltek, "8 Key Performance Indicators for Architecture Firms" (2025)— https://www.deltek.com/resources/articles/kpis-for-architects/
3. Monograph, "Utilization Rate Guide for Architecture and Engineering Firms" (2025)— https://monograph.com/blog/utilization-rate
4. Stambaugh Ness, "Common Misconceptions About Utilization Rate for AE Firms" (2025)— https://www.stambaughness.com/blog/misconceptions-about-utilization-rate-ae-firms/
5. Velosio, "Best Practices for Tracking Utilization for Professional Services Organizations" (2025)— https://www.velosio.com/blog/best-practices-for-tracking-utilization-for-professional-services-organizations/
6. Kantata, "How to Measure and Improve Billable Utilization" (2025)— https://www.kantata.com/resource-article/how-to-guide-measuring-improving-billable-utilization
7. Accelo, "Calculating Resource Utilization and Billable Utilization Rate" (2025)— https://www.accelo.com/post/calculating-resource-utilization-billable-utilization-rate
8. Noloco, "Deltek Alternatives for Modern Architecture & Engineering Firms" (2025)— https://noloco.io/blog/deltek-alternatives
9. Deltek Clarity Study, "The Latest Deltek Clarity Industry Studies" (2026)— https://www.deltek.com/company/news/latest-deltek-clarity-industry-studies-highlight-ai-challenges/
10. Rocketlane, "AI-Powered PS Management: The 2026 Guide" (2026), citing McKinsey 2024 State of AI report— https://www.rocketlane.com/blogs/ai-professional-services-management