The Validation Trap: Why GBS benchmarking needs a revolution

"We benchmarked at top quartile last year."

How many times have you heard this in the boardroom? How many times have you said it yourself?

There's something deeply human about our relationship with benchmarks in Global Business Services. At its core, benchmarking has always been about validation—that psychological need to look back and say "I did it."

If I had to break it down to primitive instincts to rank, order and categorize - these are as ancient as the first sentient beings on earth. The best cave stories, the best crops for the hunter -gatherer. the best art, the best wheel forms etc. and possibly while benchmark metrics may evolve a need to rank performance will continue to stay in any industry, Take for instance LLM benchmarks such as accuracy F1 Score, etc. this industry didn't exist / was not prominent just a few years ago but we quickly got around to ranking the most advanced intelligence around us . take for example the future of camera performance. its likely that it may evolve from metrics such as resolution, image quality, low light sensitivity etc. to something more nuanced like how well can a camera aid with "creating" an image represented by the best combination of hardware benchmarks and AI tool benchmarks.

We've built entire measurement ecosystems around proving we've achieved something, compared to someone, somewhere. But what if this validation-seeking behavior is now holding us back?

The Comfort of Looking Backward

For over two decades, GBS leaders have found comfort in benchmarking's rearview mirror. Those quarterly reports showing we're "above median" on cost per transaction or "best-in-class" for processing times have provided the validation we desperately needed. They've been our professional security blanket—proof that the massive organizational changes we championed were working.

This wasn't misguided. In GBS's formative years, validation was essential. We were pioneers in uncharted territory, and benchmarks provided the evidence base needed to secure continued investment, calm nervous stakeholders, and build organizational confidence. They helped us prove that consolidation worked, that global delivery models were viable, and that shared services could indeed be "better, faster, cheaper."

But here's the uncomfortable truth: We've become addicted to this validation.

The Addiction That's Killing Innovation

Today's C-suite conversations reveal a troubling pattern. We've elevated benchmarks from decision-support tools to decision-making authorities. Instead of asking "What should we achieve?" we ask "What do the benchmarks say we should achieve?" We've handed over strategic thinking to statistical comparisons.

Consider this scenario that plays out in boardrooms globally:

"Our cost per invoice processed is at the 60th percentile. We need to get to top quartile."

But why? What business outcome does that improvement actually drive? What customer value does it create? What competitive advantage does it provide?

Too often, the answer is simply: "Because the benchmark says so."

The Agentic AI Reckoning

The emergence of agentic artificial intelligence isn't just changing how we work—it's exposing the fundamental flaws in how we measure success. When AI agents can process thousands of transactions simultaneously, learn from each interaction, and continuously optimize their own performance, what does "cost per transaction" even mean?

When the marginal cost of processing approaches zero, being at the "75th percentile" becomes not just irrelevant—it becomes counterproductive. You're optimizing for metrics that have lost their connection to business value.

The metrics that got us here won't get us where we need to go.

This June, we're going deep on the most critical conversation facing GBS leadership today: How do we evolve from validation-seeking measurement to value-creating intelligence?

Share your thoughts and how benchmarks have helped your journey and any thoughts on a reset that you can think of.