Since everyone seems to have an opinion of what tech trends we should look out for in 2013, I challenged Bob to point out five trends we should NOT follow blindly in 2013.
Here's Bob's list in response:
Trend #1 to Avoid: Trendspotting. Let’s get this straight between us: In
business, nothing is worth doing because it’s trendy. It either makes actual
sense or it doesn’t. Just think how many companies wasted how much time and
money recreating their websites on Facebook, only to discover that while the
Supreme Court might think corporations are people too, not very many actual
people want to be their Friends.
Trendspotting doesn’t pass the Mom Test. Remember it? When
Mom asked you, “If all of your friends were jumping off a cliff, would you jump
too?” Following business trends is like that.
Trend #2 to Avoid: Bypassing the IT department: Not only is it
fashionable to bypass IT these days, the big research firms have enshrined it
as a trend, and it serves IT right because they make everything so hard.
Go right ahead. Bypass IT. If and only if, that is, you (1)
are willing to live with a stand-alone “stovepipe” solution that only
integrates with the rest of your company’s systems through the magic of your
re-typing everything; (2) the application will live in “the cloud,” not your
data center; and (3) the only support you’ll need will come from the vendor
you’re using.
Otherwise, you’d better call IT first. There’s a reason IT
projects are bigger than you’d like, and it isn’t because they’re all idiots,
propeller-heads, or budding bureaucrats. It’s because integrating new
applications into everything you have is hard, complicated work.
Remember why God was able to create the world in only six
days: He didn’t have an installed base.
Trend #3 to Avoid: "Big" Data. There are companies and situations where
the amount of data flowing in is truly massive, and being able to gain some
knowledge from it will help you make smarter decisions. If you know the
questions you want to ask and know how you’ll analyze your data to ask them, go
for it.
But for every situation like this there are dozens where all
the “data scientist” (what trendy people like to call “statisticians” these
days) does is sort through everything, searching for correlations. But as
everyone knows, correlation doesn’t imply causation, so what do you think
you’ve actually learned? Worse, not all correlations mean anything – many are
spurious, for example, the statistically significant correlation between IQ and
length of the big toe. Size really doesn’t matter here. Neither will many of
the correlations you’ll find mining whatever mass of bits your company has
accumulated over the years.
Trend #4 to Avoid: Blaming the consultants. There are times I think
companies only bring in consultants for two reasons: To read a script (“Please
tell our board of directors we need to make massive investments in Big Data
technology. They’ll listen to you!”). Or to take the blame (“Our consultants
told us Big Data was a sure thing. So we collected it and analyzed it, and it
predicted exponential growth in sheep liver futures. What were we supposed to
do? They’re the experts!”)
If you know the answer you want, don’t hire a consultant. Hire an actor. They’re cheaper, there are more of them and they’re more convincing. And if you need a fall guy, call me. I have a whole rate schedule for blame-taking.
But really, don’t. Bring in consultants when you need a
fresh pair of eyes on a problem, and a thought process that will give you a
different perspective than the one you’re accustomed to. Listen to what the
consultant recommends. Understand why he or she recommends it. Then, visualize
the result in action in your actual company and decide accordingly.
Consultants know things you don’t, but on the other hand,
you know things they don’t. Together, you ought to be smarter than either of
you are on your own. Don’t insist on making it the opposite.
Trend #5 to Avoid: Ludditism: “I just want a cell phone that makes phone
calls!” is the new “I don’t know how to balance my checkbook!” If that’s all
you want, fine and dandy but don’t brag about it. All those other people who
like what they can do on a computer enough to want to be able to do it when
they aren’t in front of a computer too? Maybe they know something you don’t.
Or maybe they just have different tastes than you do. So if
all you want is a cell phone that makes calls, buy one. If you have no need for
an iPad, don’t get one.
Just don’t bend my ear about how smart you are for not
wanting them, and I promise, I won’t bend your ear about how smart I am for
using them.
Follow Bob's thoughts and musings on his blog.
2 comments:
Loved your post, Bob. I am constantly explaining the difference between causation and correlation so much so that I put it in an appendix in my own book! Plus, with Nate Silver's success spawning a big data fad, I find myself explaining that the problem he solved was incredibly simple compared to most real world problems: one of only two possible outcomes and the outcomes were already known - voters had already decided well before the election. Now, using his model to predict the 2016 election right now is much more like the problems I encounter.
Hah! Karen, I actually just sent Bob a copy of your book!
Post a Comment