Photo: Susan Merrell

Entries in leadership (4)


Humanizing your emails

Recently a colleague emailed my intern group list-serve without a clear salutation, addressing them as "you." One intern replied to the email, triggering a copy to the whole list-serve, thinking the you was in fact just her, rather than the collective you.

This got me thinking about how often emails mis-fire in this way, and about the root causes.

In today’s world of mobile devices, I have noticed that people sometimes reply to emails without a full appreciation of who was cc’d on the original email, or whether their reply might be going to many people. I think this happens because mobile devices do not display all the meta-data about a message the way we typically see them on larger computer screens.

Sometimes someone who is bcc’d and receives the message on a mobile device, replies to all because their phone does not show cc and bcc lines. Even though this does not reply to all the people who were bcc’d, it DOES reply to all who were cc’d. This can be extremely problematic.

Therefore, for professionalism and safety and quality reasons, I recommend the following practices, which I try to model:

Begin your emails with a salutation that makes it clear who is being addressed, such as “Dear X”. My colleague did not do this, and the use of "you" may have inadvertently misled each recipient.

If you are cc’ing anyone, mention this explicitly after the salutation, such as “Dear X, and cc’ing Y and Z”

Never bcc anyone. If you want them to see the message, forward it to them after sending.

Be aware when replying to a message that if the message originated as part of a distribution list (list-serve), your reply may go back to the whole list-serve by default, even if it looks like you are only responding to a named sender.

Generally be aware that people may be reading your email on a mobile device with a small screen and more limited formatting options, and in different time zones on devices that may or may not adjust for time zone differences.

If you can, avoid attachments. Paste useful content into the body of the message – with minimalist formatting. People often don’t see that attachments are included, especially if they are checking email on phones. Recently I was in a meeting where someone had received a bunch of attachments but denied getting attachment #7. Their email program (Outlook) displayed two lines of attachments and then you have to scroll to see more attachments. Which this person did not know to do. Therefore they kept denying that they had received attachment 7, even as the sender insisted it was attached. Attachment 7 was a critical part of a proposal that ideally would have been reviewed before the meeting. So if you do include attachments, list them in the body of the email in a numbered list.

Take the time to compose emails that short-circuit several cycles of communication. For example, when trying to arrange a meeting with someone, propose 1 definite time and offer 3 available backup times. Specify time zone. Always include an absolute date (never “tomorrow” or “Friday”). See my blog article on fully specified requests at When a meeting is confirmed, I will typically send an email with all the pertinent details (date, time, call-in information etc) and send a calendar invite with that information in both the "location" and note fields. The huge advantage of calendar invites is that they generally appear grayed out on people's calendars until they are accepted. Even if recipients overlook the email, they will usually notice something appearing on their calendar. However, email systems are not fully interoperable, so calendar invites may not function for others the way they do for you. That's why I do send a duplicate email summarizing the calendar item for people who are outside of my organization.

Generally in all communications, do some perspective switching and anticipate how the message is going to be received, accessed, interpreted, stored, and shared. The best book I have ever read on strategic communication, in which you anticipate the first, second, and third-order effects of your messages, is the book “When talking makes things worse”. Out of print, available at your library. Along these lines, use the subject line to summarize any request and deadline, as recipients do review subject lines while screening emails. Make judicious use of any ways to flag emails. My email system allows me to add an urgent mark (!), plus request a delivery receipt and read receipt, and schedule a reminder to emails. These layers communicate to recipients that I really, really want to make sure the email does not slip through the cracks. Overusing such flags will backfire, but generally I have good results with them. However, as with other issues, you cannot assume they will work similarly on all email systems.

As for the actual contents of the email, I try to limit each email to a single topic that matches the subject line, and structure the email so that a recipient can respond with a yes or no or very simple and short response, and move our collaboration forward. Finally, as with all communication, proceed with curiosity, fallibility, and perspective-taking. The primary implication of these is that I am always testing my assumptions with my respondents ("Did I understand correctly that you are requesting X, or am I missing something?"). Look over an email and count the question marks before you send it. Typically people make 10 or more statements for every question. When I review transcripts of really productive conversations, I find ratios more in line with 3 or 4 statements for every question. How do you feel when people communicating with you include questions to test their assumptions or solicit your views? My guess is you feel included and collaborative. The philosopher Immanuel Kant pointed out that a key ingredient of humanism was treating people as ends in themselves, not as means to an end. My experience is that asking questions is inherently humanizing.

When I look over these strategies for composing professional emails, they boil down to reconstructing the context for face to face, heart to heart communication that is otherwise likely to go missing from the electronic medium. Context such as the who, what, when, where, and why surrounding the communication. Context is humanizing. What a strange word, humanizing. Why would we need to humanize anything that we are involved with? Isn't our mere presence in an activity inherently humanizing? Well, no, not in the sense of being humane to each other. Something we need to keep in mind when communicating through machines.

As I write this, I've begun to hum one of my favorite songs, Rehumanize Yourself, by The Police, on the great album, Ghost in the Machine. "I work all day in a factory/I'm building a machine that's not for me/There must be a reason that I can't see/You've got to humanize yourself." If we are to build machines through our emails, let's build them for and with each other.


Continuous improvement through critical reflection

One of the mantras I have adopted in my life is: "there is no such thing as failure, only feedback."

Corollary: Back in my days at a high tech startup, my colleagues and I would regularly try to raise money from venture capitalists. We would come back to the office after making our pitch and employees would ask, "Well, did you get the money?" Our CEO would say, "No, but we learned a lot." So the expression was born in our office: "Learning is what you get what you don't get a check."

I do like to extract maximum learning from failure or feedback or not getting a check or whatever you want to call it. I have evolved a short template for reflecting critically on my performance. After any experience (e.g. giving a talk, putting on a training workshop, writing a grant, etc), I write down my answers to the following questions:

1. Current goals? What was I trying to accomplish? (These are usually carried over from a previous attempt, see last item below.)

2. Achievements? It's important to note and celebrate the ways in which I did accomplish or contribute to my goals. As my daughter's first-grade teacher says: give yourself a pat on the back.

3. Failures? What did not go well or according to plans, hopes, desires?

4. Success factors? What did I or other people say or do, or what was happening in the environment, that contributed to the achievements above?

5. Barriers? What did I or other people say or do, or what was happening in the environment, that contributed to the failures (or inhibited the achievements) listed above?

6. Next goals? What am I going to work on next time? I carry those over to the next performance.

Just as an example, here is my reflection after conducting a workshop that I give periodically on decision making:

Goals? – experiential; shorter (6 hours); same content. Focused only on skills.

Achievements? Individual, realistic practice (e.g. with computers), focused (not distracted by sharing  a computer). Finished on-time. Students were all engaged, even at the end of a long week. Students did arrive at skills they will need (confident).

Failures? Half-trained (not a lot of process training); Not ready to initiate a phone call; Did not present the service delivery lifecycle very well; (in SF Margot took us through the clinic before, recording of Margot initiating a phone call). The context. This was presented week before.

Success Factors? Computers available: kept away until they needed them (no checking emails). Experiential worked (practice, role playing). Undergrads had very fresh perspectives. No model clash.

Barriers? Limited time availability (of medical students – almost sporadic availability).

Goals for next time? Integrate the skills and process training? Immediate follow up and practice? Pipeline of patients waiting to be served? Process training would include scripts, practice calling, etc. Add time and split between two days? Video clip of process (project for student). Course for undergrads (intense), weave in medical students. 

In addition to using this framework to reflect episodically, I use it every week with my team. Each of us responds to those six questions with reference to the week we just completed.

Here are some excerpts of my reflections from last week (redacted for privacy):

LAST TIME GOALS  -  Finalize performance reviews; Film SCOPED promo; Fix budget for Mendocino in CMS Innovation; SSU affiliate agreement; BCT paper – check calculations; SV/O2O/MAP manuscripts; Reimbursements.

ACHIEVEMENTS - Performance reviews; SCOPED promo v1; CMS Innovation grant submitted; BCT paper moving along; PANCAN; Shanti; QL WCRC; IHPS adv bd meetings - leadership summit idea; CERC idea well received. Some progress on SSU.

FAILURES - Did not get to O2O/MAP manuscripts; reimbursements; CS video and marketing materials; 

BARRIERS - Grant collaborator canceled meeting, did not complete draft on time.

SUCCESS FACTORS - Grant collaborator put more resources on project and team rallied to submit a promising proposal.

UPCOMING GOALS - Mtg with John and Jill; SV/O2O/MAP manuscripts; SSU; CHQI Feb privacy mtg; reimbursements; Feedback; BMB syllabus; Promo video to Trina

Each week I share my reflections with my team, and they share theirs with me, and we discuss all the elements. It's a powerful way of going beyond setting/reporting on goals... to reflecting on the dynamics surrounding our productivity.


Decisions and reversals of fortune

I feel fortunate that my schooling led me to discover my professional calling. Based on undergraduate and graduate work in statistics and engineering, I now help people improve their leadership, teamwork, and decision-making skills. The most challenging decisions I have encountered are those involving a family health crisis.

One of the key concepts I learned early from my teachers (especially Professor Ron Howard) was the difference between a good decision and a good outcome.

In theoretical terms, you can look at decision-making in terms of coherence or correspondence. The coherence approach says, "How coherent is this decision at the time you make it?" In other words, you judge the quality of a decision at the time you make it, along a continuous spectrum. The correspondence approach says, you can't tell how good a decision was until you see the outcome.

The implications are significant for medical decision making. But before I get into that, let me illustrate them using an example from everyday life.

In early February, 2011, we had a busy father/son day in my family. The Giants were celebrating FanFest and we were given passes to get in early. My son was also looking forward to his playoff game in a local basketball league. Then we were celebrating Chinese New Year with my in-laws. (None of us is Chinese, but we live in San Francisco where this feels like a national holiday for all to enjoy.) To conclude this busy day, we had an invitation to go to a Cal basketball game as guests of my in-laws, or stick with our plans to watch USF play Santa Clara.

As the day progressed, we faced a series of decisions. It started early in the morning. I was intent on getting my son out the door so we could catch the bus (my wife and our daughter had the car). We needed to look ahead at all the gear we would need that day. I printed out a color map of the FanFest events. My son was sure that with our early passes we would have unlimited access to autographs etc. I figured out the bus schedule online and off we went. I was feeling pretty good. Ten minutes into the bus ride, my son asked me if I had the FanFest tickets. No! We jumped off the bus and headed back home.

Now we were going to be running late for the 9:30 gates at FanFest. As we rode another bus home to get the tickets, I formulated some scenarios. First I called my wife to see where she was with the car. No answer. I didn't leave a message, figuring I shouldn't bother her. We can work this out on our own, right son?

Next I told my son we could call a cab. But cabs are notoriously unwilling to come out to our house on the outskirts of San Francisco. They almost always pick up rides off the street even after accepting the dispatch to our house. Another alternative was to call Uber. This is a unique company that dispatches airport limos that have downtime, and redeploys them as city taxis during each hiatus between booked rides. You can order them using your smartphone and see their progress on a map as they come and pick you up. But it's two or three times as expensive as a cab.

My son thought we should just call a cab. I did, and got an estimate of 5-20 minutes. Twenty minutes later, no cab. Now we were really stuck because going with Uber would cost a lot and still not get us there on time.

How am I doing, son, with my decision-making?

Just at that moment, we got bailed out. My wife came home with the car. I called and canceled the cab (still no sign of it). We all dashed down to FanFest. We got there in time for the opening of the gates at 9:30 am.

Now here's where things get interesting from a decision making point of view. The gates did not open at 9:30 as scheduled for us special pass holders. And when we got there at 9:30, as opposed to waltzing in with no line (as in previous years), we had to find the back of the line... which was a 20 minute walk down the Embarcardero, practically at the Ferry Plaza. Apparently winning the World Series changes the dynamics at the Giants FanFest.

I despaired of ever getting into FanFest. I asked the kids, "Should we bail on the line and go get an ice cream sundae at the Ferry Plaza?" I don't EVER offer food bribes as distractors, so this captured their interest. But they were more intent on FanFest.

The line started moving at 10:10 and we got in at around 10:30. It was jammed. No hope of getting autographs by noon (our departure time for my son's basketball game). My son lamented the timing of his game - without it, he would have gotten in line for the autographs and grab bags, and maybe even viewed the World Series trophy (longest line for that one). But one of my management principles, for family and work alike, is that we keep our commitments in the order we make them. We attended some Q&A, watched some video highlights of backstage at the World Series, and wandered around. Then left for my son's game. As we left the attendant reminded us, no coming back in.

As soon as we got to a quiet place, my wife checked her voicemail and indeed there was a message. Our son's game was canceled - rescheduled to Monday. Aargh! Had we known this 10 minutes earlier, we would have stayed at FanFest for another couple hours before meeting my in-laws for lunch. There was much wailing and gnashing of the teeth and bemoaning of our fate, particularly by a certain 10 year old for whom this was a big big deal.

For the second time that day (an all-time record), I offered a junk food bribe as a distractor, this time successfully, and we all went to Jamba Juice.

Of course, when we got to lunch at Shanghai Dumpling, the line was long and we then faced the decision of, should we stay or should we go? We stayed. Service was slow. The food, when it arrived, was divine.

Finally, we went to the USF game. This was a make-up game because I had been too sick to take my son to the USF-Gonzaga game earlier in the month. Which turned out to be a cliffhanger pulled out by the home team, USF. This one, against Santa Clara, was OK. Meanwhile, the game we didn't go to turned out to be a triple-overtime thriller at Cal.

So how did we do in terms of decision-making?

Let's recap. First, except to my son, there was nothing terribly high stakes on this afternoon. I would say I made a low-quality decision when I spent a lot of time printing out a color map of the Fanfest events and lost sight of what should have been a higher priority, locating and bringing the tickets. This was a low-quality decision regardless of the outcome: I would recognize that as an error even if it never cost me anything.

I also made a foreseeable error when it came to ordering a cab. At that point, my son and I believed that arriving by 9:30 would be critical to our enjoyment of an event we had been looking forward to for weeks. Given what I value in clear-headed moments, this was not the right time to save a few dollars, particularly given my long and negative experiences getting cabs out to our neighborhood in a timely fashion. The Uber airport limo service, in contrast, is a lock. I should have splurged. Note that I feel that way even though we had a happy outcome, in that my wife and daughter showed up with the car.

With the benefit of hindsight, some would say, "Fanfest was over-run. You would have wasted money on a car service, and the delay associated with returning to get your tickets turned out to be immaterial." True, but given what I believed and what I valued at the time I made those decisions, they were lower quality decisions than I could and should have made. Regardless of the outcome.

Leaving FanFest was, in contrast, a good decision even though the outcome was unhappy. We experienced much regret over leaving when we learned we did not have to be at my son's playoff game. But again, given what we believed and valued, it was the right thing to do and I would do it again under the same circumstances. Perhaps my wife could have checked her voicemail before leaving the park, but it was too noisy, and anyway we share a family value of not being obsessive about checking our voicemails and emails when we are together.

At the restaurant, we chose to wait because my mother-in-law assured us the food would be worth it, based on her prior experience. We loved it. It would have been a good decision, based on what we believed and valued at the time, even if the food had turned out badly.

We attended a decent USF game and missed a splendid Cal effort in their triple-overtime loss to Arizona. My son doesn't quite subscribe to the coherence theory of decision quality (yet!), so his consolation was that Cal lost, and in retrospect, three overtimes preceding a loss would have simply prolonged his suffering. Better for him to have enjoyed the USF win. This logic is impeccable among sports fans, and matches the correspondence theory of decision quality. As a management scientist, I feel this is indeed incoherent. Attending the USF game over the Cal game was a good decision before the games took place, based on what we believed and valued that afternoon. How could the outcome of the night-time games change the quality of our decision?

Coming back almost full circle to the Chinese New Year, I just read my daughter a Buddhist fable that illustrates the different reactions we all may have to reversals of fortune. The gist of the fable is that a villager's most prized possession, his horse, disappears one night. Others come to express their condolences, or perhaps revel in his misfortune. "What bad luck," they say. "Good luck, bad luck, we'll see." shrugs the villager. Next, the horse returns with a dozen wild horses. "Congratulations on your good luck," sings the chorus. "We'll see" says the villager. Then the villager's son breaks his leg while taming the wild horses. "Such bad luck" say the neighbors. Again, from the villager: "We'll see." As the son is limping around, the Emperor's army sweeps through the village and drafts all able-bodied men. The son escapes this fate. "So lucky," say the other parents. "We'll see" says the son, who has internalized his father's attitude.

This fable illustrates that even the quality of outcomes cannot be judged in the short term! So even if you subscribe to the correspondence theory of decision quality, you are stuck waiting an arbitrarily long time before you can judge a decision. The correspondence view of decision quality is indeed incoherent.

The coherence theory of decision quality - judging a decision based on what is known and valued at the time it is made - allows us to surf on the sea of uncertainty without drowning in regrets at every reversal. This philosophy of decision making is perfectly captured in the Serenity Prayer - "Give me the courage to change the things I can change, the serenity to accept the things I cannot, and the wisdom to know the difference."

Conversely, many in our culture tend to judge outcomes. If things turn out well, conventional wisdom supposes the happy events must have been preceded by good decisions. If things turn out badly, it was because of bad decisions. 

Over the years I have directly worked with hundreds upon hundreds of patients making life and death decisions. I have reviewed hundreds more such cases through my work with organizations implementing decision support programs. Bringing the coherence theory of decision quality to the patient bedside or examining room has been rewarding because people recognize and embrace it as the most productive way to take positive action while recognizing the limits of our control. Decision support programs encourage patients to review information, ask questions of their doctors, and reflect critically on the responses they get. In practice, this usually means slowing down, challenging authority, and getting second opinions, among the seven steps I have described previously. And, having followed this process of critical reflection, we can move forward with confidence and recognize any reversals of fortune as simply bad outcomes that sometimes follow even good bets.

Recently one of my colleagues saw a patient who had experienced a recurrence of breast cancer five years after her initial treatment. In addition to suffering from the recurrence, this patient was experiencing significant distress over her recollection that she had declined the most aggressive possible treatment the first time around, opting instead for a slightly less aggressive strategy with significantly fewer side effects and a slightly higher rate of recurrence. My colleague showed this patient the list of questions  and consultation records generated five years ago as part of our decision support program at UCSF. The patient was relieved to be reminded of the coherence of her original decision, based on what she and her doctors knew and valued five years earlier. Although no one likes to think of medicine as a gamble, she recognized that she had made a bet that was right for her at the time. With a reminder, our patient was able to separate her decision from a subsequent reversal of fortune. Indeed, she might have recurred even with the more aggressive and invasive treatment. Who knows? We will support her in moving forward with courage, serenity, and wisdom.

My personal and professional experiences leave me more resolved than ever to promote the coherence view of decision quality. Others have used fighting words such as logical or rational to describe this view of decision making. Fighting words because people can have different standards for logic or rationality, and will resist the imposition of someone else's definition. But decision quality is not binary like logical/illogical or rational/irrational. It is a spectrum. Please join me in asking this question, especially in high stakes situations: to what extent do our decisions reflect what we know and value at the time we take action?





Reframing our approach to teamwork

I rely on Argyris and Schon's work in Action Science for key frameworks that help me serve as a more effective collaborator and program leader. Lately I've been paraphrasing some of their key observations and concepts using terms that are more comfortable and natural for me.

One of their key insights was the ubiquity of our mental models favoring unilateral control. They articulated a whole model, Model 1, with governing variables, core values, predicted behaviors.

After living with (in?) their frameworks for a decade or so, I want to share what I think are some of the manifestations of Model 1 that we can all recognize in ourselves and around us:

  1. Look good
  2. Save face
  3. Cover up mistakes
  4. Defend your turf
  5. Avoid conflict

These are natural tendencies that arise out of insecurity and defensiveness. Of course I want to look good! Especially if the alternative is what, looking bad?

The problem occurs when my attachment to looking good (or saving face, covering up, etc) inhibits me from being an effective leader or collaborator with respect to other worthy goals such as, say, providing a high quality product or service.

I've come to a point where I don't try to fight these natural tendencies in myself. But I do try to subjugate them to the higher callings in my roles as a leader or collaborator.

It reminds me of the subtleties around an old expression: "money is the root of all evil." Apparently the saying actually goes, "the love of money is the root of all evil." Similarly, I don't think we should suppress our insecure and defensive tendencies, so much as recognize them and let go of them at critical times.

I've been leading my program for several years now around a different set of practices:

  1. Curiosity
  2. Fallibility
  3. Perspective-taking

These are also totally inspired by Argyris and Schon, this time their Model 2 which characterizes mutual learning.

I try to cultivate in myself and others these tendencies as replacements for the control/insecurity/defensiveness habits.

Curiosity is the notion that we should approach collaboration as a puzzle.

Fallibility is just the notion that each of us is missing pieces of the puzzle, whenever we engage in collaboration.

Perspective-taking is the need to work a little harder to see the angles that others are seeing, in order to really understand what they are saying and doing.

These simple habits have been transformative for me. They appeal to the scientist in me. Many of us with scientific training are used to using these values over the life of an initiative or project, through quantitative and qualitative data collection, analysis, writing manuscripts, etc. Argyris and colleagues taught me to practice using these habits in real time, in every conversation. When you bring scientific attention to simple conversations or meetings between colleagues, these become incredibly challenging intellectually. (For this reason I have also found parenting to be the most intellectually stimulating and challenging role in my life: if you take every interaction seriously as a potentially formative occasion for your child, it deserves your best thinking as a chess player.)

I like to emphasize, to myself and others, the power and attractiveness of the new habits, rather than beating myself up (or judging others) for abusing the old.