Year: 2019

I started this blog in the earlyish days of blogging. I was teaching a course at Northwestern’s Kellogg School about strategic uses of information technology. The blog offered a way to share thoughts with my students. I quickly plugged into a community of like-minded bloggers in education and in knowledge work in general. In time, that network connected me to Buzz Bruggeman, CEO of ActiveWords. Buzz being Buzz, we connected and have remained friends and colleagues since. That is another story in itself.

ActiveWords is a Windows utility program. On the surface, it is a text substitution tool. Type “aw,” for example and the program will produce “ActiveWords” on the screen. It does much more than that, and there are comparable products on both Windows and Mac. When I was a Windows user, it was one of the first programs I installed on every new computer. I now use equivalents on the Mac.

But this is not a software review.

The default marketing strategy for this category of tool is to emphasize efficiency.The tools invariably come equipped with tools to calculate how much time you save by typing “aw” in place of “ActiveWords.” If you are, in fact, a reasonable touch typist, those time savings are modest. Frankly, they aren’t that great for poor typists. Yet, the people who do adopt these tools often become vocal fans and evangelists. Are theses fans simply horrible typists or are they on to something more interesting?

The marketing from efficiency argument is simple to articulate and deeply rooted in an industrial mindset. Tools are good if they make workers more efficient; Frederick Taylor opined on the size and shape of shovels to improve the efficiency of strong-backed men moving stuff from pile A to box B. Knowledge workers aren’t shoveling coal. None of us work in typing pools.

These tools and their effective (not efficient) use are better understood from the perspective of augmentation laid out by Doug Engelbart. Saving keystrokes isn’t the point; redistributing cognitive load is.

Software developers figured this out long ago and designed programming languages to handle the tedious aspects of building software so that developers could focus on the tricky bits. Bob Frankston and Dan Bricklin wrote Visicalc, the first spreadsheet program, so that they could focus on setting up the finance problem to be solved and let the computer take care of the arithmetic.

There is a conflict here to be managed between knowledge workers and conventional managers. If you are stuck in an efficiency world, you must resist the temptations to cram these approaches into an industrial frame. In an assembly line, the tools are part of the line; everyone uses the same pneumatic tools in the quest for efficiency.

Effectiveness calls for a more personal perspective. You might get away with mandating a standard set of tools —Buzz would be quite happy if Microsoft put a copy of ActiveWords on every Windows machine. But you can’t impose a standard set of abbreviations, for example, on every knowledge worker in the enterprise. That process has to be tailored to each knowledge worker’s individual needs.

Let me offer a simple example. Every time I decide to use the word “individual, ” I have to stop and think about how to spell it. That interferes with my train of thought. So, I’ve taught my Mac to transform “indv”into “individual.” The program I happed to use for this, TextExpander, will happily calculate how much time I save by typing 4 keystrokes instead of 10, but I don’t care. Maintaining my train of thought is something far more valuable than 6 keystrokes.

Philosopher Alfred North Whitehead captured this calculus long before I learned to type. He observed that:

Civilization advances by extending the number of important operations which we can perform without thinking about them. Operations of thought are like cavalry charges in a battle—they are strictly limited in number, they require fresh horses, and must only be made at decisive moments.

What this does call for is learning to observe our own work and look for the speed bumps and other opportunities to redistribute the cognitive load.

Twenty-some odd years ago, I had an opportunity to listen to Colin Powell speak. The event was small enough and I was senior enough that we had a useful conversation after his talk. The question I had was “how can you benefit from the experts working for you without some understanding of what they actually know and do?”

Secretary Powell’s response was that you had to rely on your ability to judge people on a human level and trust that the enterprise you were part of was able to maintain a level of quality in how it recruited, developed, and advanced its experts. I wasn’t happy with his answer then and I’ve only become more unhappy since.

In most enterprises, for most of their history, the nature of work changed slowly. In Powell’s case his job was the same from the time he was a young 2^nd lieutenant until he was Chairman of the Joint Chiefs of Staff; defend the U.S. against the threat posed by the Soviet Union. Only the scope changed.

That world was gone when we had the conversation twenty years ago. How do you manage a sales force using Salesforce when your selling days were built around a Rolodex and a DayPlanner? How does experience interpreting focus group results prepare you to extract insight from clickstream data? Pick your domain and the flood of new knowledge obsoletes old technique and overwhelms your capacity to evaluate results on the basis of merit.

There’s a dilemma here. You can’t manage what you don’t know. You can’t know everything you need to know. What you need to know changes faster than you can keep up.

Organizations exist to solve problems that exceed the capacity of any individual. We tend to think of that as a hierarchical process. An entrepreneur hires people to carry out their vision and design. Those people are there to execute work on behalf of the entrepreneur but the implicit assumption is that the entrepreneur could do all the tasks given sufficient time.

This was Colin Powell’s world. He might not know all the details of his subordinates work but he could. He was willing to trust their recommendations because he trusted the system. What if he can’t understand the recommendations? Or the alternatives?

The military sidesteps this problem with a useful workaround—commander’s intent. Commanders share the reasoning and goals being their orders. That allows subordinates to make local decisions applying their expertise to accomplish “commander’s intent” even if that might mean ignoring or contradicting the specifics of the commander’s orders.

In our expertise saturated environment, we need to create a bilateral form of commander’s intent. Knowledge intensive work in a volatile environment comes with the requirement to teach others how to understand and make sense of your work.

“Trust me” is not an option in either direction. Success depends on getting meaningful conversation to take place about the competing claims of all the stakeholders.

Contributing to this conversation places different demands depending on your role.

If you are an expert, you must understand the organizational decisions your expertise bears on. You must then be able to articulate the assumptions and limitations that constrain your options and your recommendations. The A/B testing of copy on your e-commerce site tells you nothing about the people who never visit your site.

If you are a leader, your goals need to include sharing the reasoning behind the answers you seek. In fact, you ought to stop thinking of experts as the people who provide answers. In time-stressed settings, it is always tempting to seek focus and clarity by seeking to confirm your intuitions. Executives ask for answers to their immediate problems. The counterintuitive yet more useful strategy in volatile environments is to collaborate with your experts to articulate and explore questions you have not thought to ask.

Application and software vendors are offering more customer support in places like Facebook in addition to their more formal channels. Done well, this enlists volunteer help from users, recreating some of the feel from the earlier days of the online world where there was often a strong sense of shared community.

In those earlier days, there was a threshold level of technical sophistication that you could rely on. As the internet began to open up, that level began to drop slowly. A community ethos and slow broadening of participation led to some excellent advice on how to take full advantage of the available expertise. For my own selfish benefit, I pulled together several of the best guidelines for seeking technical advice:

• Scripting News: Writing good bug reports
• How to Report Bugs Effectively
• How To Ask Questions The Smart Way – Eric Raymond

As I reread these guides, however, and sampled the recurring questions surfacing in some of these more recent support environments, I was struck by a disconnect between today’s questions and these guides. In tech support circles, there is an old, disparaging, term about certain kinds of user interactions—PEBKAC, which is an acronym for “Problem Exists Between Keyboard and Chair.” In other words, the problem is user ignorance .

That feels a tad arrogant. It calls to mind the late Arthur C. Clarke’s 3rd Law:

Any sufficiently advanced technology is indistinguishable from magic.

and what is perhaps a more pertinent variation offered by science fiction author Gregory Benford:

Any technology distinguishable from magic is insufficiently advanced.

We operate in a world that is magical to many of it’s inhabitants. Most are content to accept that world as a place of routine magic. But too many are also content to choose to operate at Muggles; to leave understanding the magic to others. To make that choice is to reject the notion that deeper knowledge is a path to more effective leverage of the available tools.

If we accept the responsibility of learning how the magic works, we open up the possibility of acquiring greater power over our world.

Back in April I started to play with the idea of intermediate knowledge work products. I want to push that a bit farther and argue that we ought to bring back and expand on the idea of working papers.

The Limits of Deliverables

Deliverables are a powerful idea for managing knowledge work. What artifacts—report, specification. contract, presentation, analysis, application—mark the end point of a chain of work? What can you create to demonstrate to your client or manager that you’ve completed the work you promised to do?

The problem with deliverables is that they discourage the process of discovery and innovation. You deliver when you are done and the only response to receiving a deliverable is to accept it—either graciously or reluctantly, but accept it either way. This problem is aggravated by our ability to turn out work products that look like finished deliverables regardless of their actual state.

Reclaiming Working Papers

“Working papers” is a concept I first encountered in my early years as a consultant working within an audit firm. An audit is a odd process; your job is to check someone else’s work and render an opinion about whether they followed the rules. The deliverable that justifies the auditor’s bill is an opinion letter stating your conclusion. What you most desire as a client is a very short letter that boils down to “yes, you followed the rules.”

Working papers are all the analyses and evidence prepared to support that final deliverable. In a paper-centric world, this creates a problem of organizing and cataloging all of the intermediate pieces of paper so that you can trace and recreate the analysis later if circumstances warrant.

On the other hand, there was also an advantage in having everything on paper. Add a date and your signature in the top corner of a memo or analysis, stick it in the files, add a new line to the index page, and your working papers were up to date. The old joke was that the files weren’t complete until there was a least one paper placemat covered with lunch notes in them.

It’s the notion of “working” that is important here; the medium is not. Thinking and managing in terms of working papers highlights the importance of process and dialog. Working papers only exist in the context of how they support the process of interaction, whether it is the interaction between data and analyst or the interaction between stakeholders seeking to understand a problem.

Digital Limits of Working Papers

The shift to digital work makes just about every aspect of knowledge work better. Preparing a spreadsheet by hand on actual 28-column paper is not an experience I wish to revisit. But we also produce digital work products in ways that obscure and interfere with the evolutionary nature of effective thinking and analysis.

One limitation of digital work products is that you must deal with version elaboration and control intentionally. All interesting work products go through an evolutionary process. Preliminary hypotheses need elaboration. The structure of an outline morphs with feedback from potential readers. Drafts beget revised drafts. Analog products reveal this evolution in the proliferation of copies and marked up drafts.

Our digital tools obscure this evolution. The spreadsheet or report file contains only the most current version. Previous versions that were naturally preserved in an analog world disappear; it takes a conscious effort to preserve an intermediate digital version. Think of the crude control efforts revealed in file names with random dates, initials, and version numbers littering the drives of the average knowledge worker.

Software developers have built sophisticated tools to support their work of evolving a code base from idea to working application. I am hard pressed to identify other knowledge workers taking advantage of those tools or practices as they develop equally complex final deliverables.

The reason that version control matters is that final deliverables grow out of interaction and dialog. That dialog is rarely linear. Future iterations grow out of the process of building analyses and testing alternative hypotheses.

This iterative development process is an essential element of doing meaningful analysis and getting all of the participants to agree with the results. Treating analysis as a process is the first step. Deliverable centric thinking obscures this. A working paper approach puts process where it belongs.

One additional step with knowledge work artifacts adds significant value at minimal cost. Our digital tools produce outputs that resemble finished product regardless of their actual state. It can be difficult to distinguish a back of the envelope calculation from an SEC-ready filing when both are printed in 11-point Arial. Adding a “DRAFT – FOR DISCUSSION PURPOSES ONLY” watermark is a small step in the right direction. What I’ve had more success with is to use fonts in working papers that drive home their intermediate status. This is a time when a font such as Charette or Papyrus sends a useful reinforcing message about the state of a knowledge work artifact.

Using Tools to Reinforce Your Intentions

We’ve been creative and adept at creating multiple digital tools to make knowledge work easier and more effective. We’ve been less thoughtful about how to deploy and use those tools to support all the relevant dimensions of our knowledge work processes. If we want to be more effective at knowledge work, then we need to be more intentional about how we fit our tools and processes to the tasks we undertake.

It was essentially a casting call. I was interviewing retired and semi-retired executives to fill a role in a training simulation we were building. We needed someone to play the part of the client CEO and someone had introduced me to Scott. Scott was a central casting silver fox, tanned from a recent visit to Florida.

I don’t know why I noticed it, but Scott was wearing a custom made shirt and the shirt had no pockets. I remarked on the missing pocket and Scott’s response sticks with me two decades later. Here was his logic, ”My job is to delegate work to my direct reports. If I have a pocket someone will want to hand the work back to me in the form of a question or a task. Without a pocket, there’s no place to put a note and no way for that to happen.”

That was a 20^th Century view of management and work. Just as well that Scott was about to retire. But the mindset persists; there is work and there is management.

Peter Drucker invented the term knowledge worker sometime around 1959—just another indicator of his prescience. We accept that we live in a knowledge economy but grapple with what it means to be knowledge workers. Organizations remain slow to accept all the implications of that line of thought. At the core, the distinction between worker and manager is disappearing. There may still be power dynamics, but you can’t readily discern who is working and who is managing by observing the tasks they carry out.

We are all struggling to make sense of the changing nature of work. There are grand policy level analyses on the implications for organizations, industries, and nation states. At the other extreme, there is an endless supply of tactical advice and tools for tackling very specific problems.

And, there is the middle where we spend our days trying to muddle through.

Maybe it was the time I spent in the wings at the boundary between what you see on stage and what goes on behind the scenes to make the magic happen. It gave me roots in that middle space. I chose to stay there, building connections between vision and execution. That began deeply immersed in designing and building technology and information systems to answer particular management questions about aluminum cans, soft drinks, industrial paints, or construction equipment.

Working in the technology space led to questions about business and organization that I couldn’t answer. That led me back to school several times and into multiple organizations in search of more insight. The schools gave me pieces of parchment attesting to my mastery of subjects they deemed worthy; chiefly strategy, information systems, and organizational design. The organizations I worked for and created put that knowledge to practical tests and frequently reminded me that parchment and mastery aren’t well correlated.

This has played out as we’ve all grown hardened, if not accustomed, to a world of accelerating change. The theatrical metaphor that helps me grasp what this change entails is a shift from scripts to improv. In a script world, we grow by adding to our repertoire of scripts we can call into play. In an improv world, we grow by learning to see patterns that we can play with and by collaborating with other players to create magic in the moment.

“It’s always best to start at the beginning.”
Glinda, The Good Witch of the South, The Wizard of Oz

Remember, Glinda isn’t real; nor is the opportunity to start at the beginning. We always start in the middle.

We’re better served thinking like MacGyver. Step 0 should always be to empty your pockets and look around. What do you have to work with?

Pretending to take out a clean sheet of paper is well-meaning but ultimately misleading. The goal of that clean sheet is to avoid ending up with little tweaks squeezed into “the way we do things around here.”  The only way to achieve that goal, however, is to build a comprehensive picture of the “the way we do things around here” together with an understanding of why.

That clean sheet of paper is one of those business cliches that sounds wise, yet conceals more than it reveals. The point of the clean sheet is to eliminate assumptions that no longer serve their purpose. But you can’t surface those assumptions without understanding the existing environment.

It isn’t the assumptions you see that cause problems, it’s the assumptions you miss. Better to have a fully marked up sheet of where you are actually starting and know what obstacles need to be addressed than to trip over something hiding behind the whiteness.