Vendor lineage

Lotus Development

Forecast a million dollars in its first year, made fifty-three, and shipped software still running four decades later.

Mitch Kapor and Jonathan Sachs founded Lotus in April 1982 with backing from Ben Rosen. Kapor was 32 and had been head of development at VisiCorp, the distributor of VisiCalc, where he wrote VisiPlot and VisiTrend and was bought out of his rights for $1.7M. Sachs then spent ten months writing Lotus 1-2-3 in assembly language for the IBM PC.

1-2-3 shipped on 26 January 1983, and the name described three uses - spreadsheet, graphics, database - of which people overwhelmingly used the first. What made it win was less romantic than the name: it was fast, it recalculated quickly, and it was written to exploit machines with 256K of memory rather than the smaller ones its predecessor had targeted.

The numbers from that first year are worth stating in full. The business plan forecast $1M of sales. Lotus did $53M, and was the world's third largest microcomputer software company by 1983. Very few companies on this timeline missed their own projection by a factor of fifty in the right direction.

Its predecessor is already on this site. The Apple entry records VisiCalc as the software that made the Apple II worth buying - the first spreadsheet, and the reason a business would justify a personal computer at all. Kapor came from the company that distributed it and built the product that replaced it. Then in 1985 Lotus acquired Software Arts, the company that actually wrote VisiCalc, and discontinued it. The successor bought the predecessor and switched it off.

Jim Manzi arrived in 1982 as a McKinsey consultant, became an employee four months later, president by October 1984, and chief executive in April 1986 when Kapor stepped down. He ran the company until it was sold.

And then the product that outlived everything else. Lotus Notes came out of Ray Ozzie's Iris Associates, and it was not a spreadsheet, an email client or a database, but a thing built out of all three: replicated document stores that worked when disconnected, with application logic attached. That is an unusual design and it is exactly why it survived - organisations built their actual business processes inside it, and a business process is far harder to migrate than a file format.

IBM bought Lotus in 1995 for $3.5B, primarily for Notes, and specifically to get into client-server computing as its own host-based OfficeVision was being made obsolete. Along the way Lotus had also acquired cc:Mail in 1991 - which appears on this timeline in the Qualys entry, because cc:Mail's founder Philippe Courtot went on to run Qualys for two decades.

On 6 December 2018 IBM announced the sale of Notes and Domino to HCL for $1.8B. So the software has now had three owners across more than thirty years, and it is still sold, still supported, and still running the internal processes of organisations that built them in the 1990s and never found a reason expensive enough to justify leaving.

That is the fact worth carrying away, and it contradicts how this industry usually talks about itself. Most of the companies on this timeline were bought for a technology that was quietly retired within a few years. Lotus was bought for one that outlasted the buyer's interest, the buyer's strategy, and eventually the buyer.

Two footnotes about the founders, both good. Sachs left in 1985 to write photo-editing software and has been shipping it since 1994. Kapor dropped out of a master's degree at MIT Sloan in 1979 to go and start all this, and finished it in 2025 - forty-six years later.

Founding stories

1982

Lotus Development

Cambridge, Massachusetts · Founders: Mitch Kapor, Jonathan Sachs

One founder who had built products for the company that distributed VisiCalc, and one who then spent ten months writing the replacement in assembly language. The division of labour is the whole story: somebody who understood what the market wanted next, and somebody who could write it fast enough to matter.

1984

Iris Associates

Westford, Massachusetts · Founders: Ray Ozzie, Tim Halvorsen, Len Kawell

The separate company that built Notes, funded by Lotus and acquired by it in 1994. Its founders had worked on PLATO, the university teaching system whose notesfiles were among the first online discussion spaces, and the lineage shows: Notes is a descendant of an academic collaboration tool rather than of an email product.

The timeline

1-2-3 ships
26 January, written for machines with 256K of memory when its predecessor had been built for less - a deliberate bet that the hardware would arrive.
Notes 1.0
Group messaging, discussions, contacts, encryption, message authentication and document links, on DOS and OS/2 - a feature list that took the rest of the industry most of a decade to match.
cc:Mail
Acquired, and its founder later ran another company on this timeline for two decades.
Iris acquired
The developer of Notes brought inside, a year before IBM bought the whole thing.
Domino 14.5
An AI layer running inside the customer's own environment rather than a vendor's cloud, alongside REST and gRPC interfaces and container support - architectural work, not maintenance.

Flagship products and solutions

Lotus 1-2-3The spreadsheet that defined the category on the IBM PC, and whose macro language turned it from a calculator into a programming environment that finance departments genuinely programmed.
Notes and DominoThe client and the server: replicated document stores with application logic inside them, which is a description that sounds unremarkable now and had no equivalent in 1989.
cc:MailCorporate email, acquired in 1991 and eventually displaced by the company's own Notes mail.
SmartSuiteThe office bundle - Word Pro, Freelance Graphics, Approach - which lost to Microsoft's for reasons that had more to do with distribution than with quality.

Key innovations

The application lives inside its own database fileAn NSF file holds the documents, the forms that display them, the views that index them and the code that processes them, all together. That is why a Notes application is portable and why leaving is so expensive: you are not migrating a database, you are migrating an application whose logic has no existence outside its own storage format. The body above notes that a business process is harder to move than a file format - this is the mechanism.
Every document knows what it is everywhereA note carries a local identifier and, separately, an originator identifier that names the same document across every replica and stamps its last modification. That pair is what makes bidirectional replication tractable: two copies edited independently can be recognised as the same document and compared, rather than silently overwriting each other.
Replication built for networks that did not workBoth ends push and pull, only changes travel, and edits made offline queue locally until a connection appears. That was designed for dial-up and expensive leased lines - and it produced software that worked on a laptop on a train in 1993, which is roughly when the rest of the industry decided that was impossible.
Schema-less documents, decades earlyFields could hold varied content and documents in the same store need not have the same shape. The document database was not invented in 2009; it was shipping in 1989, in a product most people remember for its email client.

Main markets

Large organisations that built internal applications - approvals, case handling, registers, the workflows that never justified a bespoke system but ran the business anyway. Those applications are why the software outlived three owners: they were written by people who left, in a language nobody replaced, doing work nobody fully documented.

It lost the office suite market to Microsoft and the mail market to Exchange, and kept the one thing neither competitor directly replaced: a place where a department could build a working application without asking anybody's permission.

Analyst standing

There is no current analyst position to report for a company that stopped existing as one in 1995. What can be assessed is the software, and the assessment is unusual: still sold, still developed, still receiving architectural work three decades after the market declared it legacy.
The reason, in one line: what was acquired was not a product but a place other people had put their work.

Acquisitions

1985 Software Arts
The company that wrote VisiCalc, the first spreadsheet and the software that made the Apple II a business purchase.
VisiCalc was discontinued. The product Lotus had beaten in the market was bought and switched off by the company that beat it.
1991 cc:Mail reported around $50M
The dominant corporate email platform of its era, at roughly 40% market share.
Philippe Courtot, 1988 - he later ran Qualys for twenty years, and appears on this timeline in his own right.
Lotus's email business, later overtaken by Notes itself.
The purchase figure is commonly reported rather than confirmed in this session's sources; treat it as approximate.
1994 Iris Associates
Ray Ozzie's company, which had developed Notes under contract to Lotus since the 1980s.
Notes and later Domino - the products that survived two further owners and are still sold today.

Lotus Software ↗

From the company

This company no longer trades under this name. Now part of HCLTech, which acquired the Notes and Domino portfolio.

Sources