Business Partnerships for Engineers
As a software engineer, I sometimes have to remember that the best way to explain something is to break it down into individual logical steps, so that there's no misunderstanding what's being said. Computers, after all, are just really fast tools that need things described to them very specifically and very exactly so they know what to do.
The single most powerful construct in software languages is the conditional. If/then is the centerpoint of every program. After all, without it, we couldn't even end the program, because we wouldn't know that we'd reached the end of what we were doing. So it's reasonable, I think, to expect software engineers to understand how to compare conditional statements, and values that can result from them.
What's also necessary is to understand initial run conditions. In logical proofs, these are called "premises." So let's start by defining some initial assumptions about our theoretical program...
CONST_0) We Lotus professionals would like to be more successful.
CONST_1) CONST_0 is enabled by a vibrant market in Lotus products. That is, lots of customer demand and profitable implementations.
CONST_2) CONST_1 usually (but not always) comes from having products that offer greater value to customers. This could be achieved by having a better product, or by making it available for less cost -- both achieving a higher ROI.
CONST_3) Delivering CONST_2 doesn't do any good unless the customers know that there's greater value to be had, so they must be informed.
CONST_4) IBM controls in large part how CONST_2 and CONST_3 happen.
(continued...)
Given these assumptions (and readers are invited to challenge them!) it seems a logical conclusion that Lotus professionals will attempt to affect CONST_4. That is, we will attempt to influence how IBM evolves the product, the pricing and the marketing of the Lotus products, in order to create the vibrant market needed to improve chances of success. Controlling CONST_4 leads to changes in CONST_2 and _3, which yields a higher chance of CONST_1, and therefore CONST_0.
So how do Lotus professionals control CONST_4? How do Lotus professionals influence IBM's decisions on how to achieve a more vibrant market? I can think of 2 ways: private communications and public communications. (Note: they might also decide to not communicate at all and simply accept whatever IBM does -- there are plenty of those in the marketplace who do this. They might also think that the achievement of their own success demands that they no longer be LOTUS professionals, and move on to a different set of technologies. We are not concerned in this case with either of those two conditions -- though both of them definitely affect IBM's course of action as well, as they happen en masse.)
Let's say that private communications cause IBM to behave in fashion X, and public communications cause IBM to behave in fashion Y. The choice, then, of public vs. private communication should be made because, in the eyes of the communicating professional, the value of Y exceeds the value of X. That is: public communication of a point achieves greater impact than private communication. (It's also possible that public communication is simply easier than private communication, and therefore the cost to the professional is lower, even if the outcome is equal. But I have trouble believing that leaving a comment on a blog site is more expensive than emailing Ed Brill.)
Of course, public communications has the side effect of being, well, public. Therefore it is communication with more than just IBM -- it is communication with the entire world. Indeed, this is the basis of the belief that Y is greater than X. If you tell someone to change their behavior behind closed doors, it's much easier for them to ignore you than if you do it in a room full of people.
But that communication then also has another outcome. It creates an effect on observers to the interaction. Let's call that effect Z.
When we take that effect into account, then the difference between public and private communication is no longer Y-X. It's now Z+(Y-X).
This difference is the essence of what I'm getting at in my prior observation about real impact on the market. If Z is a negative, then it counteracts then marginal positive of Y-X. Remember, you could always have effect X by communicating privately. So the real assumption with any public communication is that (Y-X) > Z.
The interesting aspect of this is that there is only one instance of the value of (Y-X). There's one target audience for that: IBM. (Although admittedly there are thousands upon thousands of individuals affected by any communication, there's still only one resulting behavior that can change.) While for Z, there is an effect on thousands upon thousands of behaviors. For every prospective customer that observes a public communication, there is a single instance of Z. It might be very small. But the real effect on the marketplace is (individual Z) * (the number of customers who observed it.)
So if Z is a negative, then the marginal value of (Y-X) needs to be very high in order for the net effect to be positive, doesn't it?
Make no mistake: the internet amplifies the hell out of Z. The visibility of negative communication is enormous and almost unmissable in the modern world of Google and the Blogosphere.
Remember, it's our assumption that the net effect being positive is good for the Lotus professional doing the communicating. That's a challengable premise. There's another effect we should consider as well. Observing the communication between IBM and a given Lotus professional also creates an impression about that professional in the minds of the observer. Let's call this Z1. It's possible that Z1 is very different than Z. In some cases, we can imagine that the direction of influence of Z1 might be different than Z. That is: the observer might form a more favorable impression of the communicating professional, while forming a less favorable impression of the Lotus product line.
So the decision process for the communicator would be (Y-X)+Z1 > (Z)
This is the "tough love" or "keeping IBM honest" strategy. It's a difficult metric to gauge, because it really only applies to one Lotus professional's relationship to their customers. If, in order to achieve a more positive Z1, a professional encourages a more negative Z, then they are drawing in greater personal success at the sacrifice of the rest of the market. It might be the case that for the market as a whole, (Y-X) < Z, but when you factor in Z1, then the decision of the individual becomes to act publicly. Remember, Z1 only has to be greater than the marginal effect of Z for that particular professional.
The entire point I've been trying to make in the last week is that Z exists. There is an effect on the observer, and when the message is negative, the effect is negative. So the only rational strategies involving public negative messages are...
1) because you think that the marginal benefit (Y-X) on IBM by having to address the matter publicly is greater than the cost (Z) of the negative influences on observers (the "customers understand tough love" strategy)
or...
2) because the positive observational effect (Z1) to the individual outweigh the negative observational effects (Z/market share of partner) to that individual as part of the collection market place (the "my customers admire my honesty" strategy)
Any other motive is in error, whether of logic or of knowledge.
I don't really buy that strategy 1 is very smart -- but there's at least some ground to stand on. If you're going to create the Z effect for your fellow Lotus professionals, I would encourage you to consider carefully whether it can be avoided. This isn't the IBM of 2002, and the community of Lotus professionals is highly valued by every IBM exec I deal with.
Strategy 2 is definitely rational and self-interested. Of course, it's also greedy and short-sighted. But there are plenty of people who are greedy and short-sighted in the world. In my experience, they are not good choices for friends.
The values of X, Y, Z and Z1 can all vary greatly depending on stuff like subject matter and location, of course. For instance, it might be the case that the values for these things are very different when talking about engineering topics vs. marketing topics. In my own experience, communication with IBM about product branding has very little impact, whether public or private. Y-X is very small because both values are very small, and therefore any Z effect at all is likely to create a net loss. On engineering matters, IBM is usually much more responsive, so Y-X could be quite large. In specific areas, such as security, the value of Y can be huge, but then X is usually pretty large also. The Z effect, of course, can be truly devastating. Which is why Lotus professionals are, by and large, cautious about going public with security concerns unless there is some major problem with private communication.
I hope this discussion provides some insights for my fellow engineers on how to observe the cause and effect of conversations in the Yellowverse.
P.S: Let's be 100% clear that I'm not singling out the behavior of any individual. If you see yourself or someone you know in any descriptions here, that's your choice, not mine.
Comments
We have a name for this public communication with the intent of creating negative impact in the market. It's called FUD. It doesn't really matter whether FUD creates a change in the behavior of IBM -- it matters whether it creates a change in the perception of the market.
Of course, FUD has a Z prime as well. If there is enough observation of FUD communication in the market, the communication itself becomes suspect. People form biases based on the source of the communication. There's a reason that MSFT's motives are so frequently questioned.
Note that FUD isn't achieved by truth or falsehood, but by the purpose of the communication. It's quite easy to make communication ambiguous enough that it's truth value is assured, while still creating a negative Z effect. eg: the statement "Notes is dead."
That's why it's interesting to watch communications from a competitor. Since you know that the value to the competitor from Y is either zero or negative, then the only purpose of that communication is the effect on the observer. It's ALL Z effect. Which is why it's so monumentally silly to rely on a vendor to tell you the plans or strategies of their competitors.
Posted by Nathan T. Freeman At 04:10:44 PM On 09/14/2008 |
There's also a school of thought that the Z effect of negative communications is positive. I'd call this the "fair and balanced" strategy. The assumption is that the observer values the general availability of both positive and negative communication to outweigh the negative impact. This is based on the concept that openness has an inherent value (and usually that the value is substantial.)
Personally, I find that the "fair and balanced" strategy only works with highly educated and involved observers. Insiders see the inherent value. Outsiders usually do not make that overall connection. I believe this effect is observable in all walks of life, whether from stories about Debian's SSL to Sarah Palin's term as Mayor to whether male circumcision is a good idea. The effect on the observer of negative communication is inversely proportional to the knowledge and involvement level that the observer already possesses.
Engineers like to understand the fullness of a problem. Most people do not have an engineering mentality, and would prefer to use other people's knowledge as a filtering mechanism.
Posted by Nathan T. Freeman At 04:37:59 PM On 09/14/2008 |
The page I linked to above has an intriguing description of the intended purpose of those initially added to the discussion when private communication fails:
"1. discerning whether the accused party really has done wrong,
2. witnessing what is said, and
3. adding more “peer pressure” to repent."
Here's where I think the analogy is apt: there have been times where I've excoriated IBM about some perceived flaw in Domino, only to be informed by one or more members of the community that IBM hadn't "sinned" to begin with. By the same token, when we've asked privately for something and waited years with no discernible response, it's interesting to see how IBM responds to recurring themes on blogs and heavily promoted ideas on ideaJam.
Beyond this stage, however, that analogy rapidly breaks down, because there's no official community leadership to escalate to, and we can't exactly ex-communicate IBM. After all, it's their product line... so the closest equivalent would be mass migration to competing technologies.
On the other hand... it's possible the analogy is even more apt when applied to each of us instead of to IBM. While I doubt that you chose your variables with this in mind, it's fitting that your Z refers to the impact of each public communication on the market, because (although this is the opposite of actual CSS rules) it's the negative Z's that tend to be the most visible. I try to keep my praise and criticism of Lotus products in perspective (as opposed to in balance, which would require me to hunt for something to gripe about every time I find something that I like), but if I were to shift into a mode of constant negativity, what would it look like if someone applied Matthew 18 to that behavior?
1. They'd confront me privately, suggesting that my behavior is damaging: to myself, to others, or both.
2. If I blew them off, they might involve a few others in the discussion to see if they were just being too hard on me or if the criticism was actually valid, and if so, to lend additional weight to that criticism.
3. Although there are no official community leaders, there are many members of the community that are pretty universally respected, looked to for unofficial leadership, etc. Escalate it to some of them in the hopes that maybe I'll tone it down a bit if they think I'm hurting (or, at a minimum, distracting) the community with my negativity.
4. If the behavior continues, ex-communicate me. While I might think I'm providing a valuable service by "holding IBM accountable" or serving as a counter-balance to "fanboy mentality", etc., am I really contributing to the community anymore (if I ever was)? Am I even a part of it to begin with, or just standing on the sidelines lobbing rocks at people's heads?
I'm a firm believer in free speech. I have the right to say whatever I want on my blog. But having experience with these products (whether as a user, administrator, or developer) or even just an opinion about them doesn't make me a member of the community... my interaction with others based on a shared interest is what defines me within the community. If that interaction is a net negative, kick me out. Take me off of PlanetLotus; drop me from your feed reader; ignore me. Maybe I won't go away, but at least my obnoxiousness is no longer a distraction from what the community would otherwise be talking about: you know, iPhones.
Posted by Tim Tripcony At 05:04:42 PM On 09/14/2008 |
For the sake of focusing on the process rather than the specific example, I've like to point out that the same cycle is used for voting out politicians, firing employees, or discrediting scientific theories. Effective conversations start between private individuals and escalate as-needed and using peer review as a guidance tool.
The internet has a tendency to turn every issue into a soap box.
Posted by Nathan T. Freeman At 05:11:28 PM On 09/14/2008 |
because there's no official community leadership to escalate to
Of course there is! There's Mary Beth Raven and Ed Brill, for one. There's the hundreds of IBMers on the Bleedyellow IM server for another. And in your case, there's the ability to communicate directly with IBM leadership via participation in the Design Partner program. (Which is so inclusive these days as to be straining under it's own weight.)
Posted by Nathan T. Freeman At 05:14:22 PM On 09/14/2008 |