The 58-Year-Old Printer That Turned the Deal Room Into an AI Operating System

The most interesting thing about Datasite is not that it uses artificial intelligence. Every software company with a login screen now says that. The interesting thing is that Datasite began in 1968 as a husband-and-wife typesetting operation in St. Paul, Minnesota. Its work arrived on paper, under deadline, carrying the sort of financial information that makes lawyers hover over commas. Nearly six decades later, the medium is a cloud platform, the documents can be questioned by an AI assistant, and the deadline still rules everything.

Datasite sits in the anxious middle of mergers and acquisitions. A seller must show confidential contracts, accounts, employee records and intellectual property to potential buyers without losing control of them. Buyers need to search, question and compare that material. Bankers want to know which bidder is engaged. Lawyers need an audit trail. Nobody wants a spreadsheet named FINAL_v7_reallyfinal.xlsx deciding the fate of a billion-dollar transaction.

The company's answer is a suite built around the virtual data room. Diligence handles secure review. Prepare stages documents before launch. Outreach manages buyer communications. Acquire organizes buy-side work. Pipeline tracks possible targets. Archive preserves the record. Around that core now sit Grata's private-company intelligence, MergerLinks' relationship data, Sherpany's board software and Blueflame's agentic AI. The product map follows a deal from the first promising name to the last saved file.

First, the old machine stopped making sense

When Rusty Wiley arrived from IBM in 2014, Merrill Corporation was less a company than a family reunion of businesses: financial printing, communications, regulated content and a useful SaaS product called Datasite. Several divisions faced secular decline or technical disruption. Leadership could not sensibly fund roughly a dozen separate futures. Product releases could take a quarter. The sales force wanted improvements it could actually sell.

That was the first failure: not a spectacular crash, but the slow collapse of portfolio logic. A business assembled for the paper era had too many competing needs for the capital and attention available. Wiley has recalled trying to explain the company to his mother and getting a blank stare. She asked him to email a description she could forward to friends. It is a funny anecdote with a merciless strategy lesson. If a company needs forwarding instructions, simplification may be overdue.

“We had to make some very difficult decisions and start a journey to simplify the business.”Rusty Wiley, on the Merrill transformation

The team chose the crown jewel. Over four years, Merrill sold legacy divisions and redirected investment into its M&A platform. In 2018 it completed the major divestiture of its financial-printing operations. A disclosed sale of the Print Logistics business brought $60 million in cash, or $43.9 million after costs and adjustments, and helped reduce borrowings. Other transaction values were not broadly disclosed. The full price of the pivot therefore cannot be reduced to a tidy number. It included businesses sold, employees transferred, years of organizational strain and a 50-year corporate identity retired.

The product became the name

By March 2020, “Merrill Corporation” described the past better than the business. The company renamed itself Datasite, borrowing the brand customers already used for its platform. The timing was uncanny. Days later, work that had long relied on war rooms, couriers and face-to-face review went remote. A secure digital room moved from useful infrastructure to obvious necessity.

CapVest acquired Datasite later in 2020 for an undisclosed price. The private-equity owner backed an expansion that looked more deliberate than acquisitive sprawl. Firmex added another established data-room business. MergerLinks pushed upstream into finding advisors and relationships. Sherpany moved sideways into boards and meetings. Sealk added AI search. Key Ansarada businesses deepened coverage in Australia and New Zealand.

The bigger turn came in 2025. Datasite acquired Grata, an AI-native database of private companies, and CapVest promised $500 million for further intelligence expansion. Datasite then bought Blueflame AI, whose agents were designed for investment research. Sourcescrub and, in May 2026, Sweden's Valu8 enlarged the data pool. Valu8 alone brought information on 70 million European companies and more than 600 customers. The company's ambition had moved beyond storing a deal after someone found it. Datasite wanted to help find the deal.

AI can enter, but the permissions stay

Generic AI created an awkward problem for transaction teams. The fastest way to summarize a contract was often to download it from the secure room and upload it somewhere else - making an uncontrolled copy and punching a hole through the audit trail. Datasite's April 2026 MCP server attacked that gap. It lets assistants including ChatGPT, Claude, Copilot and Blueflame work with live room content while existing permissions remain in force. Actions are logged. A user can only reach documents that user was already allowed to see.

Integrations with legal-AI platforms Legora and Harvey extend the same idea. Lawyers can pull approved transaction material into analysis and drafting workflows without rebuilding an access system by hand. This is where Datasite is genuinely different from ordinary cloud storage. Its useful unit is not the file. It is the governed action around the file: who saw it, what they asked, what was redacted and what happened next.

“AI belongs where deal work lives.”Rusty Wiley, Datasite president and CEO

Scale sharpens the advantage. Datasite reports more than 16,000 new transactions a year, 1.8 million-plus users since 2020 and activity in more than 180 countries. It says four of the five largest global M&A transactions by value in 2025 used its platform. Because projects begin before deals are announced, Datasite also sees the market backstage. In the first half of 2026, global deal kickoffs on the platform rose 31 percent year over year, led by a 52 percent increase in the Americas.

A premium tool for expensive moments

Datasite sells enterprise software through demos and custom statements of work; it does not publish a standard price list for the core room. That makes “what does it cost?” frustratingly deal-specific. Buyers need to ask what counts as a page, how storage and overages work, how long the room stays open, what Archive costs and which services are included. Industry pricing surveys report that enterprise data rooms can run into five figures for a small deal and far more for huge document sets, but those reports are not Datasite quotes.

The trade is straightforward. A startup raising a modest round may prefer DocSend, SecureDocs or another self-service tool with visible pricing. A global bank running a contested sale cares more about permissions, multilingual support, redaction, Q&A at scale and a person answering at 3 a.m. Datasite offers 24/7/365 support in 20 languages. It is priced and staffed for transactions where delay, leakage or a missed document can cost much more than the room.

Its closest enterprise alternatives include SS&C Intralinks, Ideals, DFIN Venue, DealRoom and Imprima. Datasite's pitch is breadth: deal intelligence before diligence, workflow during it, and a governed record after. That breadth can also become the risk. Customers do not benefit merely because one owner has assembled many products. The advantage appears only if identities, permissions, data and navigation actually move cleanly across them.

The copyable move is subtraction

The popular version of digital transformation says to modernize every department. Datasite's experience suggests a harsher sequence. Identify the product customers already pull from you. Sell or close businesses that consume leadership attention without strengthening it. Rebuild teams around the product rather than internal habit. Then expand one painful customer workflow at a time.

That play works only under specific conditions. There must be a true crown jewel with customer trust and attractive economics. Divestitures must buy enough time to invest. The new product culture needs authority, not a renamed IT department. And acquisitions must share users or data with the core. Without those conditions, “focus” becomes a slogan attached to a shrinking company, while expansion becomes another collection of unrelated logos.

Datasite kept one durable idea from the printing era: high-stakes information needs careful handling by people who understand the deadline. It discarded almost everything else - presses, portfolio clutter, eventually the Merrill name. The bet now is that the same trusted room can become the place where a deal is discovered, analyzed and executed. That is a much larger claim than secure file sharing. At least this time, the company is easy enough to explain without forwarding an email.