Billable Utilisation Hits Historic Low: Why Certinia Is Betting Acquisition Can Fix Consulting's Admin Drag

The Number That Should Embarrass Every Consulting Partner

Sixty-six point four percent. That is where billable utilisation sat across professional services in 2025, the lowest reading in the entire history of SPI Research's benchmark survey. The target most firms work towards is 75%. The gap between those two figures represents hours lost to chasing notes, assembling statements of work, and building status updates that nobody particularly wanted to read.

The 2026 SPI Research Professional Services Maturity Benchmark, drawn from 509 real organisations managing $63 billion in PS revenue, confirmed what most operations directors already suspected: AI tool adoption has not translated into recovered capacity. Generative AI use in projects rose to 27.1% in 2025, a 40% increase year-on-year, yet utilisation fell anyway. Buying tools and actually extracting billable time from them remain two entirely different exercises.

What Certinia Is Actually Buying

On 21 July, professional services automation vendor Certinia announced the acquisition of Moonnox, a Chicago-based AI-native delivery platform founded in 2022 by three alumni of Salesforce integrator Bluewolf. Financial terms were not disclosed. It is Certinia's first significant acquisition in more than a decade.

Moonnox's technology does something distinct from the AI assistants most consulting teams have bolted onto their existing stacks: it extracts operational context from unstructured content. Documents, conversations, decisions. The thesis is that the manual co-ordination burden persists precisely because AI tools trained on clean, structured data cannot handle the messy reality of how project work actually happens. The relevant intelligence lives in meeting transcripts, email threads, and shared drives, not in the PSA system.

The deal extends Veda, Certinia's AI System of Action, adding agents designed specifically for proposal development, statement of work creation, implementation planning, meeting documentation, risk monitoring, and post-project knowledge management. The target integration spans Salesforce, Microsoft 365, Google Workspace, Jira, Confluence, and Zoom. Certinia describes the combined platform as the professional services sector's first fully autonomous services delivery system. That is marketing language, but the underlying architecture point is substantive: context-aware automation across the full arc from proposal to renewal, without rebuilding data mappings at each handoff.

Clients already on both platforms include PwC and Spaulding Ridge. Certinia itself is backed by Haveli Investments, Salesforce Ventures, TA Associates, and General Atlantic.

The Wider Problem Moonnox Is Being Asked to Solve

The utilisation crisis is not simply a technology failure. It is a structural one. As Certinia's own analysis of the SPI data notes, the gaps that erode utilisation typically open before a project begins: sales seeing demand that delivery cannot yet plan for, resource teams knowing who is available but not who is actually suited, project scopes shifting faster than staffing plans can adjust. By the time those disconnects become visible in a margin report, the damage is already done.

Standalone AI add-ons can automate isolated tasks. Scaling those capabilities across a services enterprise requires the kind of unified, context-aware data layer that most firms have not built. According to SPI Research's 2026 Impact of AI on Professional Services report, data quality has ranked as the greatest barrier to AI adoption for two consecutive years. When finance, delivery, and client systems do not share a common data model, AI has no solid foundation on which to operate.

That is the gap Moonnox is positioned to close. Whether it can do so at enterprise scale, with the governance requirements that PwC-tier clients demand, remains the unanswered question.

What the Trademark Signal Tells You

UK Class 41 trademark filings (covering education, training, and knowledge services) reached just 2,818 in 2026-Q3, a decline of 66.8% against the prior period, according to AIBD analysis of IPO trademark data as of July 2026. Class 41 is the primary filing class for corporate training and L&D brands. A collapse of that magnitude suggests that the cohort of new entrants who were branding and registering AI-enabled training businesses through 2024 and early 2025 has thinned sharply. The market is consolidating, not expanding. Firms that cannot demonstrate ROI on AI-assisted delivery are not building brands around it. They are exiting.

The Certinia/Moonnox deal is a product-level response to exactly that consolidation pressure. If the utilisation gap cannot be recovered through point solutions, the only viable move is a platform that owns the full workflow context.

The Uncomfortable Conclusion

Consulting leadership has spent three years telling clients that AI would transform delivery economics. The SPI utilisation data suggests it has done so unevenly, and mostly in the wrong direction so far. The administration burden has not shrunk; it has shifted, fragmented across more tools, and in many cases grown.

Acquiring Moonnox does not solve that by itself. Integration timelines in PSA are rarely as clean as press releases suggest, and getting consultants to trust AI-generated project context enough to act on it without manual verification is a culture problem, not a software one. But the underlying logic is sound. If the industry's utilisation problem is fundamentally a context problem, the answer has to reach into the unstructured layers where most working knowledge actually lives.

Sixty-six point four percent is the number that puts everything else in perspective. Every percentage point below 75% is margin that no AI licence has yet recovered.