Global Pragmatica LLC®

Recently on LinkedIn, Asit Rairkar asked about how to parse a tricky date string into a numeric date column in JMP. He has date/time data as character data, but unfortunately it’s not formatted in one of the many date formats that JMP supports directly. He wrote:

I have found that informat does not work with all sorts of texted dates. e.g. I have had some tables with date codes as YYMMDDHHMMSS and I have had to manually seperate the 2 digits each and combine them again into YY/mm/dd hh:mm:ss format to feed in. Not sure if there is a better way to deal with this type of input.

Writing JSL to handle this situation is conceptually simple, but getting the JSL just right is tricky. The format JMP supports that comes closest to this format is “yyyy-mm-ddThh:mm:ss,” so that’s what I decided to work with, but overcoming the little differences between JMP’s format and Asit’s string takes some work. We have to doctor up the input string so that it matches the format JMP can handle. We’ll need to:

1. change the two-digit years to four-digit
2. add the “-” delimiter between year, month, and day
3. add the “T” delimiter between date and time
4. add the “:” delimiters between hours, minutes, and seconds

Once we’ve gotten the input string to match a JMP-supported format, we put that inside Num() as the Formula() for a new numeric column. We also have to specify one of the date/time formats for the column, and to keep things simple, I’ll just use the “yyyy-mm-ddThh:mm:ss” format again, but you could choose any format.

Here’s how it works

Suppose your data table looks like this:

You can generate my example by running this script:

dt = New Table( "tricky times",
  Add Rows( 4 ),
  New Column( "event",
    Character,
    Nominal,
    Set Values(
      {"Erin starts researching Asit's question", "Erin starts this blog post",
      "Erin wants a beer", "This blog post better be done"}
    )
  ),
  New Column( "when",
    Character,
    Nominal,
    Set Values( {"100726174324", "100728154315", "1007281730", "110728154315"} )
  )
)

So our first programming task is to convert a string like "100726174324" to a string like "2010-07-26T17:43:24", by performing the steps listed above. We could probably do this most efficiently with a clever pattern-match or regular expression statement, but both of those strategies present some disadvantages for a casual scripter:

• they’re hard to learn, and it’s probably too much trouble given how seldom a casual scripter needs to do this sort of thing
• they’re hard to read and decipher later—somebody who’s got a lot of practice at this kind of thing can write the patterns or regular expressions fairly easily and quickly, but figuring out a pattern or regular expression that someone else wrote (or that you yourself wrote a week ago) is much harder
• they’re hard to debug
• they’re hard to “borrow from” later on when you encounter a problem that’s slightly different

For these reasons, I’d rather solve this problem with a formula that is perhaps inefficient but is easy for me to read and easy for me to adapt later on. It won’t run as fast, but the time I really want to save is my own—I don’t care so much, in this case, if JMP has to work a little harder. So, let’s take it one step at a time.

1. change the two-digit years to four-digit

This is potentially tricky. Are all of the years in this century, e.g. 20-something? Or do we have a mix of centuries? If they’re all the same, it’s easy. We just Concat()enate a “20” in front of the two year digits we do have. I like the shorthand || operator better than the function Concat(). We get those two year digits by taking a Substr()ing of :when starting on the first character and continuing for two characters:

"20" || Substr( :when, 1, 2 )

If they’re a mixture of centuries, we’ll need to add some logic for guessing the correct century. Using a transition point is one way to do it. For example, if none of the dates are expected to be in the future or more than a hundred years in the past, we can safely assume that anytime the yy portion of the string is greater than the current year, we should prepend “19” rather than “20.” To do that, we need to compare the string’s yy, Substr( mydate, 1, 2 ), against today’s yy. But what’s today’s yy? Start by getting today, as a year: Year( Today() ) is the numeric value 10. Now convert that to a string with Char() so that we can use Substr() to grab just the last two characters:

Substr( Char( Year( Today() ) ), 3, 2 )

Now do the comparison inside an If(), and make the result "19" when yy is greater than the current year and "20" otherwise. Concat() that result onto the yy:

If( Substr( mydate, 1, 2 ) > Substr( Char( Year( Today() ) ), 3, 2 ),
  "19",
  "20"
) || Substr( mydate, 1, 2 )

Whew! Fortunately the next few steps are easy.

2. add the “-” delimiter between year, month, and day

We’ve already got the yyyy piece. Now we use Substr(:when, x, 2) a few more times with x being the first character of the mm and dd pieces. We insert a string "-" between each piece, sticking it all together with ||s:

 || "-" || Substr( :when, 3, 2 ) || "-" || Substr( :when, 5, 2 )

3. add the “T” delimiter between date and time

That’s easy:

|| "T" ||

4. add the “:” delimiters between hours, minutes, and seconds

Same old same old! Just more Substr() and || action.

Substr( :when, 7, 2 ) || ":"
|| Substr( :when, 9, 2 ) || ":"
|| Substr( :when, 11, 2 )

Put it all together!

Those four steps have produced a string formatted like "2010-07-26T17:43:24", and if we put all that inside Formula( Num( ) ) for a New Column command, we’ll be creating numeric date values.

New Column( "time",
  Format( "yyyy-mm-ddThh:mm:ss" ),
  Formula(
    Num(
      // convert what you have to something like "2010-07-26T17:43:24"
      // assume dates only in past, within last 100 years
      If( Substr( :when, 1, 2 ) > Substr( Char( Year( Today() ) ), 3, 2 ),
        "19",
        "20"
      ) || Substr( :when, 1, 2 ) || "-" || Substr( :when, 3, 2 )
        || "-" || Substr( :when, 5, 2 ) || "T" ||
      Substr( :when, 7, 2 ) || ":" || Substr( :when, 9, 2 )
        || ":" || Substr( :when, 11, 2 )
    )
  )
);

Now we have:

Notice that this technique is even forgiving of strings that are too short. I intentionally left off the ss part of a few of those strings, but since Substr(:when, 11, 2) just produced an empty string "", the formula still worked. That’s an advantage we’ll lose in the next strategy.

Uh-oh! There’s a problem in that last row! When I wrote my example time strings, I was thinking of 2011, but our algorithm above assumes that all data are from the current year or earlier, so that date got converted to 1911 instead. My event description needs improvement.

What if the dates are not in a data column?

Suppose the date value strings aren’t in a character column of a data table—suppose they’re just string values. In that case, we can use the In Format() function to convert them. The guts of it are still the same, but this time we’re parsing a string stored in a global called mydate instead of rows in a column.

mydate = "100726174324";
mydateNum=Informat(
  // convert the string to "2010-07-26T17:43:24",
  // assume dates only in past, within last 100 years
  If( Substr( mydate, 1, 2 ) > Substr( Char( Year( Today() ) ), 3, 2 ),
    "19",
    "20"
  ) || Substr( mydate, 1, 2 ) || "-" || Substr( mydate, 3, 2 ) || "-"
    || Substr( mydate, 5, 2 ) || "T" || Substr( mydate, 7, 2 ) || ":"
    || Substr( mydate, 9, 2 ) || ":" || Substr( mydate, 11, 2 ),
  "yyyy-mm-ddThh:mm:ss" // or another format
);

Running this script returns the numeric value 26Jul2010:17:43:24 and shows that in the log. Notice, however, that when mydate = "1007281730";, missing the ss part of the string, it doesn’t work, and we get a missing value. Our ugly hand-made parser is actually more forgiving than JMP’s In Format() command. But it’s not much better—if we were missing any other part of the string than the part at the end, we’d get mixed-up results.

What pesky date, time, or duration strings have vexed you?

The good folks at JMP keep adding support for more and more date/time/duration formats, but it seems like there’s a never-ending supply of crazy formats that analysts encounter. Typically the strings are coming into JMP from other programs and from data-streaming devices, but they might also arrive when peeling data off web pages or dealing with data entry performed by other people.

What are the crazy formats that you’ve had to crunch into numeric values in JMP?

What are the tricks you’ve developed for crunching them?

Please share your experiences in the Comments! I’ll try to answer any questions you leave there, too.

Sudarshan Krishnan asked a question on the LinkedIn JMP Professional Network about how to change a string column into a numeric date column:

What is the JSL command to convert a character data type (“hr:m:s” –eg:”23:59:59″) to a Numeric, Continuous, duration (“hr:m:s”) data type?

Dates and times in a character column

Let’s assume you have a data table includes date, time, or duration values as strings, in character columns. You might run into this situation frequently when importing data from text files or from other programs such as Microsoft Excel. The columns might look like this: Here’s a simple JSL script you can run to create a data table like this one:

New Table( "sample",
	Add Rows( 4 ),
	New Column( "timeChar A",
		Character,
		Nominal,
		Set Values( {"23:59:59", "10:14:01", "22:01:03", "08:01:05"} )
	),
	New Column( "timeChar B",
		Character,
		Nominal,
		Set Values( {"23:59:59", "10:14:01", "22:01:03", "08:01:05"} )
	)
);

Convert date/time strings to numeric

You have two options for how to handle this, depending on what your goals are. You can either make a new, numeric column that converts the strings to date/time values using a formula, or you can change the character column directly into a numeric column. Let’s try it each way.

Option A: make a new numeric column

To make a new numeric copy of the original column, use the New Column() operator with the arguments specifying the new column’s name (“timeNum A”), data type (Numeric), the new format desired (in this case “hr:m:s”), and a formula. The formula is simply the original column, :timeChar A, inside the Num() operator.

New Column( "timeNum A",
	Numeric,
	Format( "hr:m:s" ),
	Formula( Num( :timeChar A ) )
);

Now you have the original string column, unchanged, plus a new numeric copy of that column. Note that the new column is linked to the original with a formula, so if you want to delete the original column, you’ll get an error:

Cannot delete column “timeChar A” while it is referenced by the formula of column “timeNum A”. Removing a formula leaves the data unchanged. Removing references replaces each reference with an empty value. These effects are permanent and cannot be undone.

Just click “Remove Formula” to proceed. Of course, if you no longer want the original character column, you might as well go with Option B instead.

Option B: change a character column to numeric directly

To change an existing character column into numeric, just send a series of messages to the column, first changing the type to numeric, then specifying a format. You might also want to change the modeling type to continuous, and you might want to change the column’s name.

Column( "timeChar B" )
	<< data type( numeric )
	<< Format( "hr:m:s" )
	<< modeling type( continuous )
	<< set name( "timeNum B" );

After

Now the data table with two character columns should have one character column, a second column that was character and has been changed to numeric, and a third column that is a numeric copy of the first column:

Answering a LinkedIn group question about object oriented programming in JSL from Philip Brown, the always helpful Mark Bailey wrote, “The best scripts exploit JMP objects such as data columns and platforms.”

I don’t quite agree.

In my opinion, Mark’s both right and wrong in saying that the best scripts exploit JMP objects like data columns and platforms. Yes, if a JMP data object or analysis platform already knows how to do something you need, you shouldn’t be reinventing that wheel. BUT–and this is a big but–JMP’s data tables and their sub-objects (columns, rows, etc.) come with a heavy overhead cost, and this is where I think it’s sometimes better to avoid using JMP’s object.

It all depends, of course. If you just need a quick result, by all means, let JMP do the work for you.

But if you need to do something computationally intensive, you’re far better off grabbing just what you need in a JSL data structure–a vector, a matrix, a list, an associative array, whatever is appropriate–and doing that computation outside the data table. You’ll speed up your code, reduce the memory consumption, and even avoid difficult-to-chase-down crashes.

Don’t forget that most of JMP’s formula operators can work on things other than data columns, too. I recently worked on a client’s script that was doing its own calculations of means on lists of numbers by adding up the values, counting the number of values, and dividing the sum by that. It was correct, of course, but it was SLOW. I got the script to run about twice as fast by using JMP’s column-wise mean operator, Col Mean(), on data columns instead, and then I got it to run five times faster still by using JMP’s regular Mean() operator on vectors instead of data columns.

Why did this change give me a tenfold speed improvement? Two reasons.

1. JMP’s internal calculation of means is more efficient than anything we can build in JSL—JMP’s developers have always optimized their code so that JMP’s numerical computations are done lickedy-split.
2. Those calculations happen much faster when you don’t run them through the data table, which is a complex object with a lot of internal and external dependencies—and a big, fancy window that needs to be repainted whenever the data change. Window-painting alone can cause significant script slowdowns.

Another more important reason to use JMP’s operators is that JMP’s developers have already thought through pesky details like the proper handling of missing values, underflow and overflow errors, and other arcana of numerical computation that can cause reasonable-looking calculations to get wrong results.

I’m not bashing data tables, by the way. It’s amazing what JMP’s data tables can do for us, with all their table and column properties, row states, formulas, dynamic links to graphs, and so on. All that power comes at a cost, though, and basic numerical computations will go faster when they’re not manipulating complex data objects.

Bottom line: if you’re taking advantage of the rich features in JMP’s data tables, use data tables, but if you’re just doing some calculations, and speed is an issue, then do the calculations outside the data table. In either case, take advantage of JMP’s built-operators as much as possible.

Previously I wrote a response to Adriel Hampton’s thought-provoking blog post entitled “Five Reasons to Kill ‘The Meeting'” in which I argued why I think live meetings, preferably in person, are valuable, even though many of us hate a lot of them. Now I’m going to share some tips on how to make your meetings better.

I’m writing this primarily for people who run meetings, but most of these ideas can be used to good effect by mere “powerless” attendees. These are all classic facilitation concepts, and while a designated, trained facilitator will have advantages that attendees don’t, attendees can often speak up and “facilitate from their chair” with astonishing effectiveness, and in some groups, a peer will be far more effective than any authority figure.

What people hate most about meetings is feeling powerless.

Or ignored.

But it’s usually the same thing.

• We all hate going to meetings where we’re talked at and nobody notices or cares if we fall asleep.
• We all hate meetings where the decision has already been made, but nobody’s being up-front about that.
• We all hate meetings where the people who need to hear the discussion aren’t in the room, or aren’t listening, or just don’t get it.
• We all hate meetings where we know we’re going to have the same old fights and end up in the same old impasse, and nobody’s going to make a decision (or realize that their plan hasn’t been working and isn’t likely to).
• We all hate meetings where only one point of view is important. I don’t really care if the CEO thinks this is the only way to save the company, if I know it can’t be done in the time and budget allowed, or if I know that the customers hate it when we do that, or if I know that’s the right thing to do but key stakeholders are too proud to accept a change in plans, or, or, or, or…

Meetings slow things down, and that’s good. (Sometimes.)

Central to many arguments about meetings is a premise that meetings slow things down. Certainly it’s true that many meetings are a waste of time for at least some if not all of the participants, and it’s not uncommon for people to have so many regularly-scheduled meetings that they effectively have only a one- or two-day work week. (More on that below.)

However, I question the premise that speeding things up is a good thing. The more important an outcome is, the more important I think it is to slow down and make sure it’s the right outcome.

“Go slow to go fast” is a facilitator’s mantra. It is far better to waste an hour in a meeting than to proceed with a plan that misses an important detail or a team that isn’t in full agreement.

A single team member who disagrees with the plan can sabotage an entire project. A good facilitator discovers who that person is and makes sure that person has a chance to voice their concerns. A good facilitator helps that person get the chance to explain what the others might not be considering.

Sometimes that person is just a nuisance. But even the troublemakers usually have useful points to make, even if you don’t like the way that they make their points.

When this person’s concerns are heard respectfully, and restated by others so that the person can be confident s/he was understood, then the group can weigh those concerns against the known constraints and competing concerns in a way that either incorporates those concerns or at least enables the person to go along with the plan. Even if the group reaches a different decision, if the dissenting concerns are acknowledged and weighed in a process that is transparent and is consistent with the group’s agreed-upon decision-making method, usually the dissentor(s) will be able to commit to the plan.

More on both of those ideas!

Transparent doesn’t mean public.

When I say that a group (or leader) needs to have a transparent process, that doesn’t necessarily mean that everybody is in on everything. It only means being clear and honest about how information will be explored and how decisions will be made. For example, a leader can say, “I want to get your feedback and ask for a show of hands on the different options today, and I will take that input to next week’s meeting with the directors, who will make a decision.” Most teams will accept that happily, but if they think they get to make the decision and then someone else does, they’ll be angry.

Transparency also requires following through on the stated process and being candid about any subsequent change of course.

Transparency and accountability means not pointing fingers at the team who tried to talk you out of it if it eventually turns out you were wrong. It might kill you to say it, but acknowledging that the team was right and you were wrong will buy you tremendous team loyalty—so much that I’d almost recommend doing that on purpose once. Almost!

Agree on (or at least announce) a decision-making method.

Decisions don’t have to be unanimous or even consensus or majority-rule. Many decision-making methods can work. The most important thing is to have one, and the next most important thing is to have group agreement or at least a candid announcement about what it is.

How do you decide how to decide? It depends on what’s at stake. Generally, the more say a team has in the decisions that affect them, and the more confident the team is that everyone on the team accepts the decisions, the more conscientious that team will be about executing on the decisions and being proactive about resolving issues that arise. The catch is that more say takes longer.

Here are some valid decision-making methods, from fastest and least engaging to slowest and most engaging:

1. Leader decides and announces.
2. Leader seeks input, then decides.
3. Majority rule (discuss and vote).
4. Consensus (keep at it until most people agree and those who disagree are satisfied that their concerns have been addressed or at least acknowledged).
5. Unanimous (keep at it until everybody can agree with the plan).

Having a fallback is helpful. For example, “We want to reach consensus, but if we cannot reach consensus by the end of the week, then we’ll take a vote on Monday.” Or, “If the team can reach a unanimous agreement, that will be the plan, but otherwise I’ll make a decision based on our discussion today.”

When is something important enough to justify a meeting?

What is the value of a good decision, or an effective plan, or a group that agrees enough with the plan to remain committed to it? What is the cost of not reaching these? What is the risk of proceeding without certainty that everyone is onboard with the plan? That is the value of the meeting. The cost of the meeting is the number of people in the room, times the number of hours, times the hourly wage, plus any other costs such as travel, room rental, web-meeting fees, etc. You might multiply the number of hours by the number of people by an average cost of $50 per staff member or manager and $100 per executive or hired consultant. If the value is higher than the cost, you should have a meeting.

It’s often hard to estimate value objectively, but here are some subjective criteria that are probably good enough. Ask yourself these questions about the outcome:

• Will more than a few people spend more than a few weeks working on it?
• Will a customer ever see it?
• Could a bad result lead to a lawsuit?
• Is there anyone affected by it who might be silently disagreeing?
• Is anyone’s influence out of proportion to his or her competence and credibility? (For example, a CEO who doesn’t understand crucial technical details, or a chief engineer who doesn’t understand business constraints, or a sales manager who is purely commission-driven?)
• Are you worried about what you don’t know, or what you might not realize you need to know?

If your answers to any of these questions is yes, then it’s worthwhile to have a meeting.

Minimize the intrusion of meetings on the work week.

Meetings burn time, and not just the duration of the meeting but also the time it takes to get to and from the meeting and time spent with meeting logistics like calendar management, preparation, follow-up, and rescheduling other commitments. Worse, meetings have an interruption cost. If my work requires focused concentration for several hours at a time, then a meeting that runs from 10 to 11 am pretty much destroys my 9 am to lunchtime shift. The most I’ll be able to get done from 9 to 10 and 11 to 12 is handle some email and maybe an expense report or travel reservation. There is no way I’ll be able to debug and fix some code, or write a proposal, or intervene in a staff problem, or persuade my manager about something. If I have another meeting from 2 to 3, then my afternoon is also shot, and my most important responsibilities—the things I’m paid to do—will be postponed another day, or I’ll be forced to put in some overtime that night.

Minimize how your meetings intrude on the work week. Some easy ways to start:

Have designated meeting days.

If you can get all of a team’s meetings out of the way on Tuesdays, that leaves the rest of the week free for focused work. Mondays and Fridays tend to suffer from higher numbers of absences because of people taking long weekends, so Tuesday, Wednesday, and Thursday are better. Ask yourself whether your team benefits from having a break from work mid-week (making Wednesday a good day to meet) or from having several days in a row available to focus (making Tuesday or Thursday better). More important than which day(s) you choose, though, is that you choose, and you enforce the no-meetings-days goal as much as possible.

Have the shortest effective meeting.

Go slow to go fast—make sure everybody’s voice is heard when it’s important—but don’t waste time with unnecessary agenda items. Don’t hesitate to adjourn early. Offer breaks when people get restless, start yawning, or are clearly needing to check their messages. Start on time no matter who’s late, and never go over the allotted time without group agreement and permission for some people to leave. Don’t go overtime if it would deprive people who need to leave of their chance to be heard.

State the agenda and goals in advance.

Nothing is more frustrating than having to go to a meeting whose purpose is unclear, or worse, where some people might have (or seem to have) a hidden agenda. Send out a written agenda, with time allotments for each topic if possible, and with clearly-stated goals. The goals should be measurable and believable, and they should be nouns—for example, “a list of issues needing further exploration” or “a decision about how to proceed” or “a preliminary schedule with major milestones and goal dates.” Ask yourself how the group will know if they met the goal(s) of the meeting.

At the beginning of the meeting, check with the room: “Are these the right goals? Can we do this today? Am I missing something important?”

At the end of the meeting, even if you think you know the answer, ask the room questions like, “Did we meet this goal? Do we need another meeting? Is there anything for next week’s agenda?”

Protect your team from the risks of vague agenda items.

Your agenda might be vague or contain a vague element. If so, take steps to promote confidence that those vague areas will be handled efficiently and nobody will be ambushed by surprises.

For example, if you need to go around the room to get status reports, have everybody remain standing so that nobody will drone on and on. Add a fun element, such as “What progress did you make, is there anything you’re stuck on, and what’s the next movie you want to see?” (If you do something like this, include the list of movies in your meeting notes.)

Sometimes issues arise that might feel like an ambush to some people in the room. Do what you can to make people comfortable raising those hot-button issues, because sweeping them under the rug is never better, but take steps to protect people from unpleasant surprises becoming nightmare scenarios. For example, you might ask, “Does anybody need some time to research this before we discuss what to do next? Is there anybody else that we’ll need to include in this discussion?” Often the best course will be to allow time right away for the basics to be laid out, let people ask any immediate questions, and then schedule further discussion after people have had some time to ponder and research.

If the circumstances demand an immediate decision, do your best to let people be heard, to record objections and unsettled questions, and then take responsibility for the way you proceed. If you must make an executive decision, be transparent about that. Be honest that you’re making a judgment call with incomplete information, and remain accountable for it in future. Do what you can to revisit the unsettled points when time allows. If possible, plan ways to revise the decision as better information becomes available. If your decision turns out badly, be candid about that, too, and acknowledge that some people did raise pertinent objections.

Follow up with brief meeting notes.

Brief is the key here. All you really need is a record of the decisions and agreements, a list of points needing followup, an acknowledgment of any important disagreements or what have you, and a list of open action items with names and goal dates. Some action items might be incomplete, with names, dates, or other details to be determined. If you are ready to include the next meeting’s agenda and logistical details, great.

Always provide a method for people to correct your mistakes and omissions. For example, “Please REPLY ALL with errata and addenda. Have I missed anything important?”

Avoid detailed summaries of who discussed what or disagreed why; you can only lose at this game. Just record what was agreed and decided, and if appropriate also record the points that were left unaddressed, or the objections that were raised, or the points needing further discussion, without commentary. Ask yourself whether anybody who was in the room will be surprised by your notes or would state anything differently. Ask yourself whether somebody who missed the meeting will learn what they need to know.

Sometimes it’s helpful to consult with the room about what should go in the notes, as a way of preventing misunderstanding later on, or even as a way to bring discussion back into focus. For example, after a lengthy discussion or an uneasy resolution, you might ask questions like, “How should I capture this for the notes? Can somebody restate that for me? Does anybody disagree with this proposal? Are there any action items to go with that?”

Overtime costs a lot more than time-and-a-half.

Be especially careful about scheduling meetings that will force people into working overtime. Even if it doesn’t bring a direct labor cost increase, it usually brings a psychological cost increase.

Speaking for myself, I don’t think twice about working overtime to make up for my own poor decisions, for example, or to solve a problem that I just can’t wrap my brain around during the day. But I resent being forced to work overtime because somebody else wasted my time or made a poor decision. If I have concert tickets or a family obligation or am not feeling well, I resent it even more.

I will forgive my colleagues and take one for the team occasionally, and I’ll gladly go the extra mile when it’s the difference between success and failure for something I believe in. (And now that I’m a self-employed consultant who bills by the hour, I am extremely flexible about when those hours need to happen.) But if any work situation (or a social situation, for that matter) creates a pattern of abusing my time, sooner or later I will resent it. And that resentment will cost the organization in terms of my reduced commitment, my less-than-stellar attitude, my frayed nerves, my depressed health, and eventually perhaps even my departure. I won’t sabotage a project—I’m just not wired that way— but you’d better believe it that if you push some people far enough, they will sabotage your project. Maybe not consciously, maybe not deliberately, but they will find ways to undermine even their own success to get back at someone who has done them wrong.

Do you feel the same way? Do your colleagues?

I have some beliefs because of my experiences. You have had different experiences and reached different conclusions. I would love to hear from you!

What am I missing?

What have I gotten wrong?

What do you see differently?

What did I say that surprised you? Do you think I might be right, at least for some people or situations?

What do you think would surprise me? Can you tell me a story from your experiences that would help me understand your point?

Adriel Hampton wrote an interesting blog post entitled “Five Reasons to Kill ‘The Meeting'” that I felt compelled to rebut in his comments thread, and I thought I would write a bit more here on why I believe meetings can be more valuable—and less horrible—than Mr. Hampton and other victims of bad meetings believe, and in my next post, some tips for making that happen.

This post is an elaboration on the comments I made to his post.

Mr. Hampton’s argument is that groups can be more productive by replacing live meetings with online meeting spaces. His reasons are valid, and for many situations I agree with him. Certain kinds of topics can be explored and debated much more effectively in this way. I, too, have a lot of experience with virtual collaboration (in my case, internationally-distributed teams) and have even found that in many ways I can work more productively from a remote office than a common office.

However, even for topics that lend themselves well to offline discussion (such as a team wiki or shared documents), for anything that is important, I have found that final decisions are best made in a live meeting—preferably face to face, if that’s practical, but at least a teleconference. While anything is better than nothing, the effectiveness of the group’s interaction degrades along with the resolution of the meeting method. The most engaging way to meet is to have everybody together in a room—preferably with some social lubricants like snacks, games, a celebratory meal. This is far “stickier” than everybody scattered around the globe, wearing headsets, and distracted by who knows what else. Ask yourself how many times you’ve surfed the web or answered email during phone or web meetings, and then figure that a good portion of your team is at least as distracted as you have been.

Not everybody learns and communicates the same way.

The main point I made in response to Mr. Hampton’s post was that while his suggestions can work very well for visual/verbal learners and communicators, not everybody learns and communicates best by written word and graphics. Many people—in my experience, more people—learn and communicate best by auditory and kinesthetic means.

Here’s the comment I left on his article:

I agree that most work groups hold too many useless meetings, and for people with a visual learning/communicating style, your suggestions will be helpful for many goals that are not being met well by meetings.

The problem is that other people have auditory or kinesthetic learning styles, and they don’t grasp or convey information as comfortably through the written word as you might. Learning and communication styles also break down along another axis, logical vs. social vs. solitary (clearly your style).

If all of your stakeholders are visual, solitary learners, then shared written methods like you’ve described will work very well for a lot of meeting areas. But most workgroups have a mix of learning styles, and in my experience, auditory/social learners are the majority. Your strategies will tend to minimize their contribution.

Another problem is that even among visual, solitary learners, many important topics are best explored with real-time back and forth in which all participants listen as carefully as they talk, seeking to understand as well as to be understood, with clearly understood goals and decision-making methods. If that doesn’t describe the meetings you’re used to attending, I’m not surprised, and no wonder you feel this way! Most of us have attended far more terrible meetings than good ones—myself included.

Most groups benefit from some guidance and ideally instruction from a skilled facilitator. I have experienced for myself many times the incredible difference that good leadership can make, and if the meeting is about something important, hiring an impartial professional facilitator is something you can’t afford not to do. I greatly improved my own effectiveness as a program manager by learning and adapting facilitation principles and techniques, and I went from being someone who dreaded even my own meetings to someone who eagerly looks forward to facilitating for other groups.

Let’s break these ideas down a bit. First, about visual/solitary learners (writers and readers) vs. those other people (talkers, drawers, builders, tryer-outers).

Have you ever had an experience like this?

If you’re reading this article, then we probably have a lot in common. You write. You read. You think. Alone. And you’re good at it. Me, too.

But here’s what happens—right?

You carefully write up a proposal and send it around by email. You take pains to write a thorough discussion, detailed enough but not too long, with supporting illustrations and even good summary bullet points. You put a clear question or call to action at the end. You leave it in Drafts overnight and come back the next morning to fix up a few details before sending it out.

And then nothing happens.

You send another email. No response, or just a few one-liners come back. You phone a key stakeholder or ask them about it when you run into them at the coffee machine, and they say, “Oh, right. I read that, but…” and then they ask questions or raise objections that make it obvious they didn’t understand a thing. You’re pretty sure they didn’t even read it.

It’s frustrating! You know it was all there in your beautifully-written email, and you know that you covered all the most important points. But they don’t get it!

Why not?!

Try not to jump to conclusions. You’ll never know for sure.

• Some people weren’t paying attention.
• Some people read and understood but forgot.
• Some people got behind on their email and are afraid to admit it.
• Some people disagree so violently they can’t even think about your points.
• Some people are too busy.

Story time!

I once had a boss who told me, “If you can’t get it down to one inch, keep it in Drafts until you can.”

Oh, my G-d.

I wanted to strangle her!

But eventually I learned. I found that the shorter my email, the better my chances that she’d sign off and support me later, or answer my question. The longer my email, the more likely I’d get a brusque response that made no sense, or no response at all.

At first I thought she just didn’t appreciate my attention to detail and the subtle nuances of the  situation. Eventually I realized that she appreciated all that and trusted my judgment but didn’t have time to get bogged down in all the grey areas. That was my job, and as long as I kept her informed, she’d support me to the end.

I (eventually) figured out that the thing to do was get her on the phone and tell her I had a plan but that I wanted her ideas on this or that aspect of my plan. She was great at brainstorming solutions and seeing when my thought-framework was off.

She learned, too. She figured out that where she was good at plotting strategy, I was good at anticipating risks. Where she was good at selling ideas, I was good at making sure her plans were bullet-proof. And together, we were better at collaborating over the phone or over lunch, even though sometimes I needed to write an email to myself to figure out what I thought, and sometimes she needed to enjoy a cocktail and ignore my babble while I worked something through.

So what do we writers do about all that?

No matter how well you write, you have to face the fact some people just don’t take in written information. Some people need to:

• talk things out
• touch things
• draw pictures together
• make physical models
• conduct experiments
• listen to descriptions
• see people’s faces
• think “out loud” and ask “dumb” questions
• spell out the details of who, when, what, how

If you’re a good writer and you like working on things alone, in your own time, you might find this frustrating—I sure do!—but remember, other people find it frustrating having to read and work alone.

You’ll come out ahead if you take a variety of approaches.

I wrote more on the topic of written vs. phone and other communication methods in a Point/Counterpoint column with Tina Wuelfing Cargile.

Embrace diversity!

Rather than dwelling on your frustrations, take advantage of people’s differing skills and preferences.

• The people who prefer talking things out are also often good at enrolling others in the decision and will enjoy presenting the plan to other groups (whereas many excellent writers would rather have a root canal than give a presentation).
• The people who like to draw diagrams together often bring new insights because of their superior spacial reasoning abilities.
• The people who like to build prototypes or conduct experiments will help you find the gaps in your plan, and often they’ll come up with improvements on your idea that you’ll wish you thought of. (Ask me how I know.)
• The people who just don’t pay attention to their email are likely to pay closer attention and ask good questions when you talk to them.

But not just any meeting! A good meeting!

And how to have good meetings instead of crappy meetings will be the subject of my next post: How to hold better meetings.

What do you think?

What have I missed?

Which meetings should be killed and which should be resuscitated?

What are the tools you’ve found that work best at replacing meetings?