Martin Rand and the Value of a Conversation That Never Happened

Martin Rand had five reasons to say no. The company across the negotiating table had five demands he could not meet. He was representing The Climate Corporation in Europe, the discussion involved millions of dollars, and he gave the answer that seemed available. The other side walked away. The deal was finished before he had found out what might have made it possible.

In Rand’s telling, the useful part came afterward. He did not settle for blaming an unreasonable counterparty. He began examining his own response. What had their demands been trying to achieve? What else could he have offered? A business conversation had ended, but the questions it left behind would eventually help produce another business.

Before Martin Rand helped build Pactum AI, he had to admit that having responsibility for a negotiation did not mean he knew how to conduct one. A management title, it turns out, does not arrive with the instruction manual.

First, make the paperwork disappear

Rand’s earlier work had given him a different kind of education. At Skype, he was a product manager. Later came VitalFields, the farm management software company he co-founded in 2011. The change of subject was considerable: from internet communication to the records, plans and reporting that accompany a working farm.

VitalFields helped farmers organize field activities and track crop inputs, including information needed for European environmental compliance. Here was a software problem with mud on its boots. Time spent reconstructing records was time taken from running the farm. Making the administration easier had a practical value that needed little explaining.

The company went through Startup Wise Guys’ first accelerator batch. Its beginnings were less predictable than its eventual market: the idea that became the farm management system started as a surfing app. Anyone demanding a perfectly straight founding narrative may wish to avoid early-stage companies.

By November 2016, VitalFields was available in seven European countries. The Climate Corporation, then a Monsanto subsidiary, acquired it that month. Rand subsequently became Climate’s commercial lead for Europe. The founder who had been selling software to make agricultural work easier now had a larger organization’s negotiations to manage.

A calendar full of different rules

He has described weeks moving between negotiations with Ukrainian, Spanish and French businesses. Geography brought different expectations about trust, timing and bargaining. A commercial lead might carry the same product into each discussion and still discover that the conversation worked differently once he arrived.

That experience helps explain why Rand became interested in the mechanics beneath negotiation. People bring assumptions to a table. They misunderstand priorities, react to demands and sometimes leave before discovering a workable trade. Across enough transactions, those missed possibilities become an economic problem.

He pursued formal negotiation training, including Harvard’s Negotiation Master Class, completed in 2020, and Stanford’s Negotiation Mastery program. His eventual proposition was that negotiation expertise could become part of a software system. A useful conversation could be designed with more care than the reflexive exchange of demands and refusals.

Rand’s failed deal is instructive because he places responsibility on himself. The other party had asked for things he could not supply. He could still have asked questions. He could have explored what sat behind the requests. He remembers the cost of treating the first set of positions as the whole negotiation.

The overlooked end of the supplier list

In 2019, Rand co-founded Pactum with brothers Kaspar and Kristjan Korjus. Kaspar had led Estonia’s e-Residency program; Kristjan brought an AI background that included Starship Technologies. Rand became founding CEO. The combination joined commercial experience with technical and institutional experience from Estonia’s technology community.

The opening they pursued was easy to miss precisely because it was so ordinary. A large enterprise has more supplier agreements than its buyers can examine closely. Attention goes to consequential relationships. Smaller contracts carry on, sometimes for years, with terms that nobody has found the time to revisit.

Calling every supplier has a cost. A buyer can reasonably conclude that a modest saving does not justify hours of work. Repeat that sensible decision across thousands of agreements and the unattended opportunity becomes substantial. The constraint is the price of having the conversation.

Rand described an attractive early customer as a business with many similar negotiations. Repetition matters: a system can be prepared for a category of deals, understand its permissible trades and handle them repeatedly. A miscellaneous collection of unrelated, one-off transactions offers much less of that advantage.

“Strategy for humans and implementation for machines.”

Martin Rand

Price is only one of the things people want

The useful insight is that counterparties can attach different values to the same terms. One supplier may care about receiving payment sooner. A buyer may be willing to offer that in exchange for a discount. Contract length, delivery arrangements and other conditions can create further room to agree.

These are illustrative trades, rather than a promise that every negotiation has a happy ending. They explain why reducing the task to demanding a lower price misses much of its substance. Before software can explore a trade, the business has to decide what it wants and what it is prepared to exchange.

In his early explanations of Pactum, Rand described analysts interviewing clients and building a value function: a representation of the objectives, limits and relationships between negotiation terms. There is considerable human homework before a machine can start being useful.

This detail says something about Rand’s approach to AI. He has cautioned against assuming that switching on a tool automatically creates value. Data quality matters. So does the company’s procurement strategy. A fragmented collection of records and an unclear mandate remain management problems, however capable the interface looks.

An €80 beginning

Pactum’s first reported autonomous negotiation was agreeably small. In June 2019, the software negotiated with a Spanish taxi driver for a limo rental business. The client saved €80. It was a test, with no fee paid to Pactum, rather than proof of a mature business model.

For a young company, the point was that the exchange could take place at all. A supplier could use a chat interface, discuss terms and reach an agreement. There was something concrete to examine afterward. The theory had acquired a transaction.

Rand later spoke publicly about deployments at businesses including Walmart and Maersk. The settings were larger, but the neglected conversation remained central. Equipment, services and transport rates offered repeated opportunities to reconsider commercial terms. Suppliers could respond without requiring a buyer to attend every exchange personally.

Pactum now lists more than 50 global enterprise customers and $55 billion in spend under management. Those are company figures describing its current business, years after the first experiment. They give a sense of the distance traveled without turning one taxi negotiation into an explanation for every later result.

The machine needs a mandate

Rand has stressed the importance of explainable, rules-based negotiation technology. Authority to conclude a commercial agreement requires constraints. A fluent answer alone does not tell a company whether its objectives have been met or whether a proposed concession was allowed.

The appeal of autonomy therefore depends on preparation and accountability. People establish the strategy and acceptable limits. Software handles repeatable work within them. The arrangement still needs people who understand the supplier relationship and can recognize exceptions. Delegating execution makes the mandate more consequential.

Rand’s public career has included explaining that distinction as a speaker and writer. His Forbes work has covered negotiation, leadership and AI, including what the history of computer chess might suggest about technological change. In 2024, he was publicly identified as Pactum’s president. He also joined GScan’s supervisory board.

There is a personal detail that fits a little too neatly to ignore: he is a licensed IFR pilot. It would be extravagant to turn that into a theory of his leadership. Still, someone interested in flight has at least chosen a pastime in which preparation, instruments and operating limits deserve attention.

A different kind of first flight

By 2026, Rand’s public work included First Flight Ventures. The philanthropic initiative supports early-stage founders in underserved regions through micro-grants. Its grant program that year offered selected ventures up to $4,000. The amount belongs to a different financial world from enterprise procurement, and to a different moment in a company’s life.

An August 2026 UPVESTOR interview described Rand as Pactum’s previous CEO and discussed his travels after stepping down, his new initiative and the effect of AI on professional skills. His public LinkedIn profile now foregrounds First Flight Ventures. Pactum’s current leadership page names Kaspar Korjus as CEO.

Rand has explained the initiative in terms of unequal access to opportunity: talented founders need capital, mentorship and networks, while geography can determine which of those they encounter. Earlier, when asked what movement he would start, he answered, “I would start the movement of kindness.” Helping without an immediate return was part of his answer.

That aspiration has a practical counterpart in his advice to founders. Discuss the idea. Meet people. Prepare. Give others a chance to improve what you are trying to build. His own career has crossed product teams, accelerators, an acquisition and a founding partnership; it offers ample evidence that a company is a social undertaking.

The connecting thread is an interest in what gets passed over: paperwork that absorbs a farmer’s day, a supplier agreement nobody revisits, a founder outside the usual network. Rand once ended a negotiation because he could see no acceptable answer. Much of his work since has concerned making room for another question.