So, Coupa’s bought another AI company. This time it’s Tonkean, a workflow automation outfit. This is their second AI acquisition in under two weeks, hot on the heels of grabbing Rossum, which does something with documents. And the big pitch? Making supply chains and procurement ‘autonomous.’
Look, as someone who’s watched Silicon Valley churn out shiny new toys for two decades, ‘autonomous’ is a word that usually means ‘more complicated than it needs to be,’ and ‘agentic trade network’ sounds like something dreamt up by a committee whose primary goal was to confuse everyone.
What does this actually mean for the poor soul wrestling with an invoice, or the buyer trying to wrangle a PO before a deadline? Coupa’s spouting off about reducing cycle times by 50% and saving operations teams 30 hours a week. Those are nice numbers, if you believe them.
Let’s not forget, Coupa has been sitting on a mountain of spend data – over $10 trillion, they say. That’s the fuel for their AI engine, or so the story goes. They’re building what they call the “#1 agentic trade network,” which, by my count, sounds an awful lot like just… a better enterprise software system. Who’s actually making money here? Usually, it’s the folks selling the software, not necessarily the ones trying to use it.
Is This Just More Corporate Jargon?
This whole push into AI agents and orchestration feels like a sophisticated way of saying they’re trying to make their existing spend management platform smarter. Tonkean’s no-code platform is supposed to glue disparate systems together without tearing everything apart. That’s the promise, anyway. No-code is great until you need to do something slightly outside the box, and then suddenly you’re back in the weeds.
Coupa CEO Leagh Turner is quoted as saying this is “game-changing.” That’s what they always say. She also added, “With the acquisitions of Rossum and Tonkean, in short order, we have now amassed all the assets to make this promise possible for both buyers and suppliers at a time when buying and selling is getting more complex and costly.” It’s a mouthful, and it conveniently glosses over the fact that the complexity and cost are often created by the systems these companies sell.
Think about it. For years, we’ve been told ERP systems and procurement software would simplify everything. Now, we need AI agents to make those systems work? It’s like adding more layers to an already complicated cake, hoping the extra frosting makes it taste better.
Why Is Coupa Buying So Much AI Stuff?
This isn’t just about making things run faster. It’s about control and intelligence. Coupa wants to own the entire procurement process, from the initial request to the final payment, all orchestrated by their AI. This allows them to gather even more data and refine their algorithms, which, in turn, makes their platform more indispensable. It’s a classic tech company feedback loop: collect data, build AI, sell AI, collect more data. The question is whether the end-user, the actual business trying to buy stuff, sees a tangible benefit beyond buzzwords and impressive-sounding percentages.
Coupa claims Tonkean’s tech offers over 250 connectors and can coordinate agents. This sounds like an IT department’s dream – less custom coding. But for the business user, it still means navigating another layer of complexity. The real test isn’t the number of connectors; it’s whether that invoice gets paid on time, or whether that urgent order actually arrives when needed, without a human intervention cascade.
This isn’t entirely new territory. We saw similar consolidation in the CRM space years ago. Companies bought up specialized tools to create a more integrated, ‘intelligent’ whole. The results were… mixed. Some integrations worked, others felt like forced marriages.
Coupa’s not alone in this AI gold rush. Every enterprise software company is scrambling to slap an AI label on their products. The question remains: are these acquisitions genuine leaps forward in capability, or just expensive ways to reposition existing technology in a hotter market?
For the people on the ground, the hope is that this wave of AI means less drudgery, fewer errors, and maybe, just maybe, a slightly less painful procurement process. But given my track record of watching these cycles, I’ll believe it when I see it actually saving someone time and money, not just the company selling the system.
