Monthly Archive: January 2012

orchids and weeds. not your normal garden.

A number of years ago (7?), a colleague of mine at Stanford – Carlos Seligo – introduced the concept of “Orchids and Weeds.”  He meant it jokingly at the time, but it has stuck with me ever since.

Basically, there are two ends of the spectrum when it comes to technology, support, and adoption.  There are orchids, which are elegant, beautiful and perfectly aligned with needs but require an incredible amount of attention and resources to get running and properly supported.  It is unlikely that support will become more efficient over time.  Weeds, on the other hand, spring up organically and often as ancillary to something else.  They germinate quickly and soon adoption is very high and usage has proliferated.  Weeds might even be used in ways never originally intended.

An example of an orchid might be the Sakai Learning Management System, which has been developed by universities, for universities, but is quite a beast to implement and manage.  Perhaps even some of the major Microsoft products – Sharepoint comes to mind – would be orchids, too.  They can be ridiculously powerful if done right, but needs a lot of resources to get there.  The two “big” weeds would be e-mail and SMS text messaging. SMS is in particular is a great example.  Originally designed so that cell company employees could report coverage outages and other problems, it has become the primary way for many people to communicate (I send about 2000 texts a month and yet use fewer than 400 minutes).

It is important to note that there is nothing inherently bad about an orchid.  By definition, it is beautiful, elegant, and a perfect solution for the job.  It offers very high value and probably commands a high price (the amount of effort users are willing to expend to access the service).  But the cost is also extremely high.  So the overall value-cost gap (or V-C “wedge” as we’d say in my Capstone MBA course) is smaller than, say, text messaging.

But if the fundamental basis for maintaining a service portfolio is value, orchids are just fine.  You just have to pick and choose the right ones.  And that means a strong rubric for getting rid of the useless orchids (just pretty, but not the right fit) from the truly wonderful ones.

Basing services on value provision is pretty simple.  Well, actual “value” is hard to quantify, even in the business world where there is a number for everything.  But if you start with price – again, the expense of time/effort/distance traveled/hoops jumped through/etc that a user is willing to take on in order to use the service – then you can get an idea of value.  Value is always the same or higher than price (unless you have one seriously messed up measurement system).  If you put something somewhere (physical location of a high-end lab, a web service behind so many clicks of the mouse, etc) and people are still rushing in droves to use it, then the price is probably quite a bit lower than value.  Whether the price is too low is unclear, but it’s not like you’re going to disassemble that lab and move it farther away from the dorms or put even more clicks in front of the web service.  Bottom line – value is high, and it can be gauged if not measured exactly.

So once you have that, you just roll out services that are high value or high V-C.  The value could be high in an absolute sense, in which case cost is not an issue (striking a deal with a cloud-based backup company where the student pays and you get all the kudos but none of the liability).  Or it could be high in a relative sense (buying a bigger SAN for $75000 to do multiple redundant backups of faculty and staff data, on-site, in exchange for peace of mind to those users.  cost and value are high).  But if something is low in value, just don’t bother with it.  Because if it’s not of value to anyone, why are you wasting resources on it?  Worst case scenario, cost is HIGHER than value, and you’re just burning resources for nothing.

Ideally, you go low cast and high value – look for weeds.  Get as many of these as you can because they require low overhead but adoption will be high.  I’m not sure people will “value” it the same way they would with other things but they will surely use the service.  Align your staff to take advantage of these.  Have fewer staff managing more weeds – the ratio will be different.

Then go looking for orchids.  You will probably need project management staff just to test them, then a much lower staff:service ratio to maintain them.  But if you figure out the right potential orchids during testing, then deploy the best 2-5 or so and it’ll be worth it.