# Romain Libeau

Romain Libeau is a founder of Pivot, a procurement software company. The posts describe Pivot’s work on AI-native procurement software and focus on how data, system architecture, and implementation costs affect the value of procurement technology.

- Profile: https://b2binfluencer.marketing/creators/romain-libeau/
- LinkedIn: https://www.linkedin.com/in/romainlibeau
- LinkedIn followers: 30,560
- Topics: artificial-intelligence, startup, ai-agents
- Based on 12 dated posts, latest 2026-09-09

## Summary

Romain Libeau is most useful for practical buyer tests on procurement AI claims, data readiness, and total cost.

- Best for: A procurement technology vendor selling on data readiness, faster implementation, or total cost of ownership.
- Not for: A legacy suite launching an autonomous-procurement campaign led by agent count or AI branding.
- Check first: Confirm recent post-level results, audience composition, and sponsorship availability before contacting them.

### Takeaways

- AI in procurement is capped by the foundation, not the model.
- Subscription fees are about 23% of modeled three-year Coupa costs.
- Agent washing warns buyers against paying extra for autonomy they do not need.

## Main ideas

- **AI in procurement is capped by the foundation, not the model**: Libeau treats procurement AI as a systems problem, not a model problem. Responding to a State of AI in Spend finding that 17% of procurement, finance and IT leaders see clear ROI from their procurement technology and AI investments, they read the figure as a scoreboard, not a group to imitate: after two years of agent and autonomous-procurement launches, most respondents still cannot point to clear returns. The reasoning is that AI cannot manufacture clean supplier data, create adoption, reconcile fragmented workflows or repair bad integrations and missing context. Stacking more agents on the same foundation does not change the outcome. They make the same case from Pivot's history: starting during the AI era, they say, allowed the company to connect directly to the system of record and keep data structured, giving AI the context to do work rather than merely perform well in a demo. They acknowledge that the models are improving; the criticism is about what they are plugged into.
- **Most of a procurement software bill is not the subscription**: Libeau argues that buyers compare licences when they should compare total cost. Drawing on a Coupa-commissioned study that models roughly $24M in costs over three years, they note that only about 23% goes to Coupa fees, with the rest split between implementation, planning and deployment, and training and ongoing management. These costs are neither hidden nor surprising, they argue, but they matter when calculating ROI: returns should be measured against everything invested to deliver value, not just the subscription. They frame this as a habit inherited from legacy enterprise software that buyers rarely question.
- **"Agent" is a vocabulary problem before it is a technology problem**: Libeau argues that the language is moving faster than the tools: assistants became agents, agents became superagents, and every automation is one rebrand away from being called autonomous. They borrow Gartner's term "agent washing" for a market flooded with rebranded automation, then spell out three buyer risks: paying a premium for agentic vocabulary, buying autonomy when plain automation would do the job, and starting with an agent before looking for a problem to solve. Their question for buyers: does the problem actually call for an agent, or did the hype make the choice? Ran Shapira of Lemonade, Joël Collin-Demers of Pure Procurement, and Pivot's Marc-Antoine Lacroix were announced separately as panellists discussing what agent washing looks like and where an agent genuinely made sense.
- **Hypergrowth turns a young disruptor into an incumbent ahead of schedule**: Libeau argues that legacy status no longer takes decades; sometimes six or seven years is enough. They describe scope creep from an original idea: more products, services, packaging, specialists and AI branding than the software was designed to carry. They credit the first generation of orchestration tools with solving a genuine procurement problem by giving employees a cleaner place to start and procurement a better way to coordinate requests. Then hypergrowth knocks down the load-bearing walls, stretching the architecture and embedding complexity. Their practical advice to buyers: a recent founding date, polished interface or fleet of new AI roles will not rescue an ageing foundation, so examine what sits beneath the modern story. The argument also has a self-interested edge: it comes from a newer company.
- **Every argument routes the reader somewhere concrete**: Libeau's recurring format: a counterintuitive or self-deprecating opening, one argument developed in short paragraphs, a specific number or named example, and a clear next step — a full read, event registration, calculator or product site. Humour is the entry point, not the payload. The posts serve as doorways to longer assets, such as a study walkthrough, panel or interactive tool, rather than as complete pieces on their own.

## Work to read

- [Agent washing: what "agent" actually buys you](https://www.linkedin.com/posts/romainlibeau_everyone-seems-to-have-an-agent-now-which-activity-7503438581646962689-k5MF)
- [The 17% ROI figure and what the other 83% suggests](https://www.linkedin.com/posts/romainlibeau_i-keep-seeing-this-stat-from-a-recent-state-activity-7495829245705834496-SU6G)
- [Why the subscription is the cheapest part of procurement software](https://www.linkedin.com/posts/romainlibeau_pov-you-realize-the-subscription-might-be-activity-7496197992874586113-vvC2)
- [How a disruptor becomes an incumbent in six or seven years](https://www.linkedin.com/posts/romainlibeau_it-used-to-take-decades-to-become-legacy-activity-7493640484750290944-bfON)
- [AI-native versus agents bolted onto an old system of record](https://www.linkedin.com/posts/romainlibeau_turns-out-the-riskiest-part-of-making-an-activity-7503083199753932800-7qBM)

## Brand relationships

- **Pivot** (owned_business)
- **Coupa** (mention)
- **Lemonade** (mention)
- **Pure Procurement** (mention)

## Partnerships

Romain Libeau's posts promote Pivot's own products and content rather than paid placements for other brands. They link to Pivot's new site, a Pivot-built Coupa ROI calculator, an article on legacy software and hypergrowth, and a webinar registration for a session with guests from Lemonade and Pure Procurement. The posts include no sponsorship policy, media kit, rate card, audience specification, or partnership inquiry route. Ask directly about sponsorship availability, scope, deliverables, disclosure, and timing; the LinkedIn profile linked on this page is the only visible public contact point, but it is not a verified partnership channel.

## Alternatives

- David Blonski: Choose David Blonski over Romain Libeau when the campaign targets the legacy suite itself and the sponsor can put named customers and hard savings beside the claim. Blonski's posts make the replacement case in vendor terms: cut the legacy SaaS stack, move that money into product and people, and attach an arithmetic claim—a stated 50-70% reduction in total cost of ownership—to named companies such as Sanofi, Under Armour, and Snowflake. They hit the same procurement incumbents Libeau treats as architecture problems: Salesforce, ServiceNow, Coupa, Workday, and SAP. They also add a pricing angle Libeau leaves implicit: agent tolls, consumption fees, data lock-in, and an executive quoted on agent-pricing upside. If the campaign needs to argue that "the incumbent stack is a growing tax" and the sponsor can publish customer outcomes, these posts are closer to that argument than Libeau's readiness framing.
- Beth Thorson: Choose Beth Thorson when the buyer decision is about scaling a go-to-market motion rather than procurement software economics. Thorson's posts cover the shift from founder-led selling to team-driven growth, the case for a supply-chain operating system that unifies data and automates execution, and the human side of adoption: AI and automation simplify workflows, but people still carry relationships, innovation, and retention. A sponsor that needs a first-person operator voice gets a longer career anchor here: more than 30 years leading growth at major logistics companies, a firm Thorson founded, and a newly announced VP of Strategic Partnerships role at Auger. Romain Libeau's product-economics posts do not offer that.
- Arie Barendrecht: Choose Arie Barendrecht when the campaign goal is field presence with procurement leaders rather than a written argument. Barendrecht's posts sit inside the procurement software world at Omnea and read as event and audience work: a London summit with named procurement, finance, and operations leaders; a summary of how practical the AI conversation has become; and a framing of supplier spend as a $5 billion lever rather than a $10 million one. The line—go after the spend you control and not just the cost of running your own function—is a usable campaign hook for a spend-management sponsor. Barendrecht also writes to hiring, with an open call for sales engineers, account executives, and growth associates in New York.

---
AI-assisted summary of public posts. Audience demographics are not available. Engagement appears only where we checked recent posts. Method: https://b2binfluencer.marketing/methodology/