Lotus Agenda

» see lotus.com as it lived, on the wayback machine

~ the eulogy ~

agenda was the kind of product that needed an office full of knowledge workers with budget and patience. that was not the demographic in turkey in 1988. i was four years old and the only pc in our neighborhood was an amstrad cpc 464 someone's older brother had brought back from germany. the dos personal information manager was a category for american white-collar professionals who already had a 286 on the desk and were trying to organize meeting notes across project deadlines. it was nowhere near the orbit of a kid in ankara who would not see a pc until almost a decade later.

mitch kapor had already left lotus by the time agenda shipped, having handed the ceo seat to jim manzi in 1986. he stayed involved as designer on agenda alongside ed belove and a third collaborator, jerry kaplan, who would later go on to start go corporation. the trio worked on the concept of a free-form database where information could be entered before any structure was decided. you typed notes the way you talked, and agenda would categorize them by patterns it detected and rules you wrote later.

the first release shipped in may 1988 at $395. agenda 2.0 followed in 1990 with several sample applications bundled in for common tasks like contact management and appointment scheduling. about 200,000 copies sold across the four-year run, most of them at the $195 academic and discount price. agenda 2.0b in 1992 was the last release. lotus pulled the plug and shipped lotus organizer instead, a calendar app with a leather binder skin that did almost nothing agenda had pioneered. agenda was released as freeware long after.

~ the dossier ~

~ what it actually did ~

the trick was the inversion. every other database in 1988 wanted you to design the schema first, then enter records into pre-declared fields. agenda let you type the records first and assign categories afterward by writing rules that pattern-matched the text. if you typed "call back jerry on friday about the q3 budget", agenda could be told that anything ending in "on friday" was a friday item, anything mentioning a name from a contact list was a person item, anything mentioning a quarter was a finance item. the same line could live in multiple views without being copied.

in 2026 this looks ordinary, every notes app from notion to roam to obsidian works on some version of this idea. in 1988 it was so alien that lotus's own marketing department could not explain it. reviewers wrote pieces that started with "i am still not sure what this software does" and ended praising it anyway. mitch kapor later said in interviews that the explanation problem was the real ship blocker.

~ the unraveling ~

people bought agenda, opened it, hit the learning curve and put it down. the conceptual model was different enough from a calendar or a spreadsheet that there was nothing to map from. you had to invest a weekend learning the categorization rule syntax before any of it paid off. most users never got that far. the ones who did wrote long ecstatic letters to the company about how they could not work without it anymore.

lotus saw the sales numbers and made the obvious call. organizer in 1992 was a much simpler product, a tabbed calendar that looked like a filofax. it sold orders of magnitude more copies than agenda ever had, because customers in the early 90s wanted a calendar that looked like a calendar. the windows transition was also already in motion. lotus 1-2-3 was getting a windows port. building a windows agenda would have required a full rewrite that nobody at lotus believed would pay back.

~ the long tail ~

mitch kapor spent the next two decades trying to build agenda's successor. chandler was the most ambitious attempt, started in 2003 at the open source applications foundation with millions in kapor's own funding. it shipped a 0.7 release in 2008 and was effectively abandoned by 2009. scott rosenberg wrote an entire book, dreaming in code, about how chandler failed to ship. agenda's other revival, beeswax, last released in 2008 by a small team, kept the dos engine alive on windows for the diehards.

agenda's actual descendants are not the agenda revivals. they are notion and obsidian and tana and every tool-for-thought product after 2018 that lets you tag and query free-form text. the idea was correct. the 1988 dos environment was too narrow and too early for the audience that would have made the idea pay. nothing about that lesson has stopped anyone from trying again.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »