The best personal information managers match your work style while reining in the chaos on your desk

The business history of computing has so far reflected a seesaw struggle between the forces of innovation and of standardization. During brief experimental periods, dozens of companies propose scores of solutions for emerging needs and markets. This happened in the hardware business 15 years ago, when Eagle Computers, North Star, Xerox, Digital Equipment Corp. (DEC), and many other firms tried to make their products the industry standard for personal-computer architecture and operating systems. The same gold-rush mentality applies today to applications involving the Internet. This stage comes to a close when one system gains a small edge over the others, which very quickly becomes a big edge. Eventually we are left with a powerful standard-setter -- Microsoft today, conceivably Netscape tomorrow -- and a bunch of beaten competitors with their orphaned offerings. Then the creative period begins again in some other part of the industry.

Most of the software used in small- and medium-scale businesses went through this cycle in the 1980s and became standardized, even "commoditized." There are small differences in "fit and finish" among the major word-processing programs, for example. But in fundamental ways they are all similar to one another, and are all capable of handling nearly any user's needs. The same is true of the main spreadsheet programs and database managers, and the leading communications programs. Like automobiles from different manufacturers, all get you where you're going; if you know how to operate one of them, you can very quickly figure out the others.

Only one important software category has yet to go through the standardization cycle. This is the deceptively trivial sounding field of personal information managers, or PIMs. In entrepreneurial terms, the PIM business is still wide open. While the word-processing market is dominated by well-known programs from three big companies -- Microsoft's Word, Novell's WordPerfect, and Lotus's WordPro and AmiPro -- at least two dozen programs, many from tiny companies, still struggle for serious consideration as PIMs.

Symantec's Act!, Netmanage's Ecco, Cognitech's Sharkware, Telemagic's Telemagic, Borland's Sidekick, Modatech Systems International's Maximizer, Elan Software's GoldMine, Tracker Software's Tracker Two, askSam Systems' askSam, Micro Logic Corp.'s Info Select, CrossTies Software's CrossTies, Jensen-Jones's Commence, Microsystems' CaLANdar -- these and many other programs are all thought of as PIMs (some also call themselves "contact managers" or simply "organizers"), suggesting that they offer some roughly similar service to the user. In reality, the strongest similarity among them is the weird punctuation style of their names. You can master Info Select and yet not have the slightest clue about how Ecco or Commence operates. The data presentation that one program features as its central asset might be left out of others altogether. The entire PIM product line resembles financial-modeling software before the appearance of the first real spreadsheets, the now-defunct VisiCalc and Lotus's 1-2-3, or the word-processing business in the days before WordStar (now put out by SoftKey International) and WordPerfect supplanted mainframe-style editors like EMACS.

This crowded, unstandardized market -- which may seem promising if you are hoping to launch a new product but confusing if you are trying to find the right tool -- cannot last forever. Apart from the normal battle fatigue that will drive weaker competitors from the market, sooner or later some PIM designer somewhere will make a breakthrough comparable to what Dan Bricklin accomplished in 1979 with VisiCalc. That is, he or she will come up with a product that ends further debate about how PIM programs should look and what they should do because it is so clearly superior in what it offers the user. From that point on, rival companies will concentrate on refining and elaborating the design, as Bricklin's competitors did when producing ultimately far more successful spreadsheets like 1-2-3 and Microsoft's Excel. When that moment comes, the right PIM will be as indispensable to business as spreadsheets and word processors are now. Indeed, the ideal PIM could be the most valuable of all software products to professionals and managers, because it is the one that in principle would best approximate the juggling of different tasks with varying time horizons that makes up the manager's day.

But that moment has not arrived. The breakthrough has not yet occurred, and the reality of today's PIMs is far short of the ideal. How, then, can managers and entrepreneurs best employ these flawed products? In the short run, they can look for the program (or programs) from today's vast array that best suits the quirks of their personality and their business. In the long run, by knowing how PIMs should work, they can be ready for the breakthrough when it comes.

* * *

The term "personal information manager" was popularized in the late 1980s by Mitchell Kapor, founder of Lotus. In today's market "PIM" has been applied to products as narrowly focused as computerized datebooks or telephone lists, but Kapor had a much broader conception in mind. In their daily lives, he said, busy people had to cope with an enormous and ever-changing variety of information: the phone call to be returned 10 minutes from now; the business plan to present three months from now; the records of a legal battle last year. Ideally they should have a tool that would let them view data in isolation when that is called for -- just the phone calls to be placed today, just the reports from subordinates that are overdue -- but that would also let them see deeper connections among their data. With the tools then available, notably Borland's Sidekick, Kapor knew he could call up simple to-do lists and phone records. He knew of no program that would just as easily let him see larger patterns.

Out of such concerns grew a program named Lotus Agenda, which Kapor developed with Ed Belove and Jerry Kaplan. Agenda has legendary status in any discussion of PIMs. To financial analysts it stands as an unmitigated disaster, a cumbersome product for which Lotus never found more than about 100,000 customers and which it eventually pulled from the market. To software engineers who worked on it -- and even more, to the relatively few customers who discovered and learned to use it -- Agenda represents an impressive, almost artistic achievement, comparable to the early, failed airplane flights in France that preceded the Wright brothers' success. The details were off, but the concept was truly there. Like the pioneering French experiments in aviation, Agenda's record is worth examining because it has implications for the PIM that will someday fly.

Agenda embodied several smaller insights and one big idea, all of which should play their part in the ideal PIM of the future. A small-seeming step of great potential power was "text parsing," also known as an "inference engine." If you entered, say, "Meeting with Bob about India contract, tomorrow at 4:00," you did not have to tell the program that this concerned your associate Bob or your operations in India. When you asked to see all dealings, past and future, with Bob, the entry would show up -- as it would in a record of your Indian operations and in your schedule for tomorrow afternoon. Apparently text parsing is difficult to pull off; no PIM that has appeared in the four years since Agenda was pulled from the market has fully incorporated it. And since Agenda allowed users to fine-tune the phrases they wanted matched -- you could decide, for instance, that any item containing the names "Gandhi" or "New Delhi" is to be included in the list of items related to India -- it amounted to a kind of programmable artificial-intelligence system.

More than any PIM now on the market, Agenda demonstrated the power of "filtered views" as a way to isolate chosen pieces of information. The traditional database program, like most of today's PIMs, relies on the "search" as its main isolation tool. You issue a query -- show me all meetings involving India -- and then you leaf through the results. If you change the data, you must start another search to find the new results. "Filtered views," by contrast, are perpetually up-to-date and eliminate the need for searches altogether. And while these views were allowing users to find selected information, Agenda's "inheritance" concept was helping them see the connections and larger patterns. The manager of a company with worldwide operations, for example, might want to see an overview of information about all his Asian projects. If he defined India in his Agenda setup as being part of Asia, then data about India would automatically be included in any Asian overview. The advantage of being able to amass detailed information in overviews seems obvious: it is the way people naturally think. Yet this ability is a rarity in today's PIMs.

The large insight that Agenda's architects applied was that a personal information tool should, above all, be customizable. Everyone who has used a normal database program knows its power -- and its limits. The first step is to define, ahead of time, every possible type of data you might ever want to work with or keep track of. If you change your mind later on, it is often simplest to redo the whole project from scratch. By contrast, every aspect of Agenda's data structure could be changed at any time and from any point in the program. The number of data categories, their type, their relationship, what you wanted to call them, which items belonged in which groups -- all could be redone, and undone, and redone again, as the user's needs evolved. Only with this flexibility, Kapor argued, could a manager consider the program a tool rather than a constraint.

Despite its conceptual importance, Agenda failed on the market for several reasons. Its appeal was difficult to explain in advertisements. Early versions tended to produce corrupted data files. Lotus steered its internal resources to a Windows version of 1-2-3 and away from an upgrade to Agenda. And like other projects started by Kapor, Agenda lost in-house luster during the years Jim Manzi, Kapor's successor, served as Lotus's CEO.

Diehards willing to try a program that is no longer supported by its manufacturer, and that runs only under DOS, can obtain Agenda and add-ons called Partner's Planner and President's Planner from Phase III Computing, in Toronto (416-925-8760, Those interested in a systematic argument about the importance of improved PIMs for managers and professionals can contact the PIM Power Users of the New York PC Users Group, which a year and a half ago issued a widely circulated manifesto urging software designers to fill this niche. (The document can be obtained from Bruce Frumerman, of the marketing firm Frumerman & Myers, in New York, has been the organizer of this effort. He has argued extensively that an improved PIM is the crucial tool that will let small-business managers outcompete larger, richer firms. You can reach him at the same address.)

* * *

The full package of innovations that Agenda offered -- text parsing, filtered views, inheritance, and limitless customizability -- cannot be found in any PIM now on sale. Individual elements, though, can be found here and there. Info Select is among the most customizable of the programs; Maximizer applies inheritance; the latest version of Ecco offers a limited version of text parsing, which can be expanded through an add-on program called Catalyst Tools, also from Netmanage. Judging which of these features matters most to you -- and which others, like GANTT charts (which help you chart the progress of a complicated project), are important in your work -- involves both practical and philosophical assessments.

To begin with the practical, the most important lesson is not to buy any of these programs based solely on a description -- no, not even Agenda itself. You can get used to any of today's major spreadsheets or word processors, but the differences in look, feel, and operating philosophy among the PIMs are so great that it is crucial to actually try a program before making an investment. Some PIMs rely heavily on mouse clicks, while others provide keyboard alternatives for virtually all commands. Know your tastes -- I avoid using a mouse when at all possible -- and try each program to be sure it matches them. In practice this means either trying a friend's version or buying only those PIMs that offer a full money-back guarantee. Commence is relatively mouse-dependent; askSam, comparatively mouse-free -- but don't take my word for it without trying them yourself.

A second important nuts-and-bolts tip is to make sure you choose a PIM that allows you to enter crucial information only once. When you type in an associate's address and phone number, for example, the program should provide internal links that allow it to process that information for every dealing: using the phone number in an autodialer, listing it alongside appointments in your schedule book, or addressing E-mail. This may seem elementary, but it is beyond the capacity of some otherwise appealing PIMs.

Third, be sure to test your PIM's data-recovery system. The more you enjoy working with a program, the sooner it will become the repository of your most important business information, from phone contacts and schedules to long-term strategic plans. Losing such data is far more alarming than the normal computer glitch. Weak damage-prevention systems were an Achilles' heel of the original Agenda. Even more than with word processors or spreadsheets, lean toward PIMs that offer ways to extract data from corrupted files.

Fourth, as part of your trial, see how hard or easy it is to move data from on-line sources into your PIM. With each passing month a larger share of important information arrives on-line. PIMs that make a transfer clumsy, with fixed-length data files or awkward "paste" protocols, will become less and less attractive.

On a more abstract plane, you need to ask yourself the "cake mix versus cookbook" question. Some PIMs come custom-made for quite specific applications. Sharkware, for example, is a world built around the business card. The phone calls you've made, the meetings you've held, the birthday cards you're supposed to send -- all are tied to the business cards of your associates. Ecco presents you with a screen that looks like a legal pad, yellow background and all. Commence, askSam, and Info Select, on the other hand, start out more or less with blank screens. The structure you want to create is up to you. The "cake mix" approach, epitomized by Sharkware, is obviously easier to start with. Before deciding, ask whether you're more likely to welcome its simplicity or chafe at its inflexibility in the long run.

You can also ask whether, given the limits of today's PIM programs, you're likely to even try to use them for the long-range planning and strategic overview that Agenda's designers were reaching for. If so, look for PIMs with as little built-in structure as possible. After all, it's unlikely that a software engineer who has never seen your business will have thought of the exact strategic thought-plan that makes the most sense to you. You might even consider Inspiration, a quasi-PIM (by Inspiration Software) that lets you convert normal outlines into flow charts. I find it gimmicky, but that may reflect the peculiarity of my mind and work. Friends who are managers say they value it highly. If you're not interested in testing the limits of PIMs in this way -- if, in other words, you know you do your best planning with a pen in your hand or while out for a run or a swim -- then you can save yourself heartache by selecting a PIM with a straightforward purpose, like coordinating group schedules.

What I use for now is a little of everything: doughty Agenda for my scheduling and journalistic plans and research; the equally orphaned and antique GrandView, by Symantec, for making outlines; Info Select to retrieve addresses and phone numbers; and askSam for large data dumps. It may be messy and inelegant, but that is the nature of life among the PIMs of the day. Like other users, I'm making do until the millennium.

* * *

James Fallows ( is Washington editor of The Atlantic Monthly. His latest book is Breaking the News: How the Media Undermine American Democracy (Pantheon, 1996).


Because PIMs vary so much in look, feel, and operating philosophy, it's important to try out any program before purchasing it to make sure that it not only has the features you want but also suits your tastes. Once you have the PIM in hand, consider the following checklist:

Decide whether you want a program that defines how information will be organized or one that enables you to set up the structure yourself. If you plan to use the PIM for long-range planning and strategic overviews, look for programs with as little built-in structure as possible.

Choose a PIM that allows you to enter information only once.

Test the PIM's data-recovery system, and lean toward models that provide ways to extract data from corrupted files.

Check how hard or easy it is to move data from on-line sources into your PIM.