Proserv & PE Insights · Thursday, September 17, 2026

Hi, this is Koko from Koko Knows.

Let's get right into it, because the sharpest signal since I last talked to you is coming out of IBM Consulting, and it's the kind of move that should get every partner's attention. IBM has now productized its hybrid-AI delivery platform across roughly two hundred client accounts. This isn't a pilot anymore. It's running behind a hundred and fifty thousand consultants and more than four thousand digital workers, and here's the part that matters most: they're pricing it on productivity, not hours. That is a direct shot at time-and-materials billing across the entire industry. If IBM can show a client a fixed, outcome-based price for work that used to be metered by the hour, every competitor sitting on a T-and-M book is suddenly exposed.

And it's not happening in isolation. PwC just launched an AI strategy accelerator that compresses six to eight weeks of senior partner-led work into under a week, built on proprietary data layers and engagement memory. So think about what's happening here. IBM is squeezing the billable hour from the delivery side, and PwC is squeezing it from the senior-advisory side. Two different firms, two different attack angles, same target: the premium pricing model that's funded partner compensation for decades. If you're an MD running a practice that still leans on hourly or FTE-based staffing models, the runway to rethink that is shorter than it looked even a few weeks ago.

The second big development is Accenture's new tokenomics research, and I'd put this right up there in importance because of what it reveals about your own clients, and frankly about your own portfolio companies if you're sitting on the PE side. They surveyed seven hundred fifty executives and found that most CFOs still can't get a straight answer on what AI actually costs to run. Six blindfall spots, five disciplines needed to make token spend financially legible. This tells you something uncomfortable: after two years of enterprise AI spending, a huge number of organizations still can't produce clean ROI evidence. That cuts two ways for you. On the services side, it's a genuine advisory opportunity, you can sell governance and financial-legibility work against this gap. But on the PE side, if you've been underwriting AI-driven margin expansion in a portfolio company's model, this is your signal to go back and stress-test whether that spend is actually being tracked with any rigor at all. There's a real parallel here to the Auditoria data floating around right now, showing that two-thirds of finance teams are increasing AI investment while only about one in five can show measurable results. That's not a technology problem. That's a capital allocation problem, and it's exactly the kind of thing a diligence team should be pressure-testing before the next round of AI capex gets rubber-stamped in a portfolio company board deck.

Third, and I'll be quicker here, Cognizant is deepening its Anthropic relationship, embedding Claude into its Flowsource and Neuro AI and IT Ops platforms with spec-driven development running alongside human engineers. This is a clear bet that hybrid human-plus-AI delivery, not full automation, is where the differentiation lives, at least for now. Worth watching against the backdrop of the broader concentration risk story, too. With OpenAI pushing toward a valuation north of a trillion dollars amid open safety disputes and a documented agent security incident, every firm deepening a single-vendor AI relationship, whether that's Cognizant and Anthropic or anyone else, should have a real multi-vendor exposure map ready for the board, not just a roadmap slide.

One bright spot worth a mention: BCG cut its Scope 1 and 2 emissions by ninety-two percent against a 2018 baseline while nearly doubling revenue, using internal carbon pricing in the thirty-to-three-hundred-dollar-a-ton range. That's a useful proof point, both for firms managing their own sustainability commitments and for PE partners looking at portfolio companies where the assumption has always been that emissions discipline caps growth. BCG's numbers say otherwise, and that's a benchmark worth citing in your next portfolio review.

So pulling it together: pricing models are under pressure from two directions at once, AI spend accountability is a live gap you can either sell into or get burned by, and vendor concentration is a governance question that needs a real answer, not a slide. Watch where the next pricing-model announcement comes from, because IBM and PwC won't be the last.

That's the rundown for today. I'm Koko, this has been Koko Knows, and I'll be back with more as soon as there's something worth your time.