Data Management Infrastructure

I hate coining buzzwords.  And maybe I didn’t even coin this one.  But we need some phrase to describe the following problem:

Data now comes in two processable flavors: structured and unstructured.  And stacks now exist for processing either flavor.  But each world is undergoing transformation.  And how the two worlds will be combined is up in the air.

Unstructured data: whether or not the Hadoop ecosystem is The Answer, there is vigorous experimentation with how to work with massive amounts and velocities of unstructured data, and there are some emerging norms.  Other NoSQL approaches remain as alternatives, and there will probably be use cases for almost all of them.  We are not even near the end of the beginning when it comes to defining how unstructured data systems will interface with applications (where is the NoSQL SQL?), and we are IMHO still at the very beginning of understanding what storage systems are optimized for these workloads.

Structured data: with NewSQL databases, it is clear how to interface them with applications but far from clear how they work with storage systems, particularly SSD-based storage systems.  Jury is out as well on how to multiplex the different databases in a use case.

I call of this “data management infrastructure”, and it seems to me like an emerging big design problem.

Thoughts?  Who’s working on this?  Where should we invest?

What is a Big Wind?

I blogged here last week about my “big wind/capable team” heuristic for picking startup investments.  Today I’d like to go into a bit more detail about the “big wind” part.

A Big Wind is a transformation of a large customer spend from one category to another.  It’s almost the same thing as Geoffrey Moore’s idea of a “tornado”, which is a phase in a market when the (relatively) vast Early Majority group decides to adopt a new approach.  Moving suddenly from “show me” to “must have”, the Early Majority creates a “tornado” of demand.  Examples: the PC “revolution”, networked storage systems, scale-out servers.

Smartphones are in the midst of a tornado-like Big Wind.  Sweeping away RIM and Nokia and perhaps Microsoft.  Sweeping in Apple and Google.

Maybe tornadoes are the only kind of Big Wind that exists, but in any case, most of the examples I can think of are tornadoes.  Maybe the ERP implementation at the end of the last century was a Big Wind without being a tornado.  It was driven in large part by Y2K FUD, not by a transformation of the Early Majority to a new approach.

Great example of not a Big Wind: healthcare IT so far.  Big market, but no transformation of spend (yet).  Right now only early adopters are embracing end-to-end IT for their health businesses.  The early majority is hanging back, saying “show me.”  It may tip.

Your thoughts?

The Euro Crisis and VC strategy

Watching the Euro community grapple with and fail (or so it seems) under the onslaught of treaty country performance, I’m minded of how I think about VC strategy, and what it means for what they’re doing.

I rate investments by two criteria: Is it a Big Wind, and is it a Capable Team.  There is a third as well, but it’s irrelevant to this discussion.

A Big Wind is a market opportunity where lots of money will change hands.  And Capable Team is a team that can seize the opportunity and ride it out despite all twists and turns (and there are always twists and turns).

So we invest, and things don’t go well.  Like Greece, the company’s income is less than its expenses.  Over time.  Repeatedly.  Without apparent hope of things turning around on their own.

I’ve developed a 2×2 grid for asking what to do with a company based on evolving circumstances.  It looks like this:

Download this file

If the wind is big and the team is capable, you don’t need to do anything: you let ’em rip.

If the wind is big but the team isn’t capable, you topgrade the team until they become capable.

If the wind is small but the team is capable, you pivot.  You find a new wind, or a better take on the existing wind.

If the wind is small and the team is incapable, you let it go.  You sell it, you shut it down.  You dispose of it.

The Euro community is foundering.  I don’t know enough to know if the wind is small or the team incapable, or both, but someone should make this analysis if they want the community to survive.

What the community is doing instead, and what a lot of VC firms do instead, is probably the worst of all possible worlds: put the incapable team on a starvation budget so they can’t possibly take advantage of a big wind, but at least they won’t go under right away.

It’s neither fish nor flesh nor fowl.  And it won’t serve the Euro community any better than the same strategy serves us in trying to get our startups to thrive.

Your thoughts?

The Maker Movement and Me

I got an Arduino for Fathers’ Day from my son, and an Arduino Cookbook from my wife.

The Maker Movement just went personal for me.

In case you don’t know what the Maker Movement is, it’s a bottom-up movement of hardware tinkerers, largely hobbyists today, which I believe will revolutionize manufacturing and basically end the Industrial Revolution (or cause a new one, depending on how you look at it).

In case you don’t know what an Arduino is, it’s a popular open-source hardware module with a programmable controller and some I/O junk that sells for ~$20 and is pretty easy to program.

(That’s pretty easy for me to see; I haven’t programmed it yet.  Catch up with me a few weeks and see how easy I think it is.)

First application?  My son suggested I make a gizmo to automatically power-cycle my cable modem.  Not a bad idea, although my ISP says you should disconnect the coax cable as part of the process, which means some way to interrupt the coax.

First step in any project of this source: scour what my son calls “the InterTubes” for prior art.  I’ll start on that this week.

Very cool stuff, at least for me.

What the PC Revolution of the ’80’s can teach us about the “consumerization of IT” in the Teens

The Cutter IT Journal just published an article of mine along these lines.  I’m a big fan of analogies, and the ’80’s presents a lot of analogies to today’s world: new disruptive platform, specialized apps as the key, ultimate disruption of an entire IT paradigm.

Here’s a link to the Cutter IT Journal issue.  For some legacy-ish reason, Cutter Consortium will let you download a pdf of the journal but won’t put it online.

As always, let me know your thoughts.

United/Continental Seat Snafu 2: The Wrath of Khan

I finally received a reply to my email to United/Continental re the Seat Snafu on May 18.

As one might expect, it was empty, impersonal, and non-committal:

11:46 PM (9 hours ago)


to me

Dear Mr. Gordon:

Thank you for contacting United Airlines. 

I am sorry we were unable to respond to your request sooner.  The merger
of United and Continental Airlines has been a successful one, but there
have certainly been challenges.

  An airline merger of this size has never been accomplished before now. 
Some facets of our airline may be different, but our fundamental
commitment to our valued customers has not wavered.  As a MileagePlus
member, your business has never been more important.  Please be assured
we do understand your concerns, and they have been documented for review
and appropriate internal action. 

Please visit us online at as additional travel needs
arise.  Many inquiries about our MileagePlus program can be answered
through our online FAQ. 

While my reply is brief and not as detailed as I would like, I want you
to know I very much appreciate your business.  To thank you for your
patience and loyalty, I have added bonus MileagePlus miles to your
account.  You will see the miles added to your account within three
business days. 

We are building an airline that will earn your confidence and approval,
and we look forward to welcoming you on board your next United Airlines


 Dan Thompson
Senior Manager

What is United thinking will happen as a result of a message like this?  What does any corporation hope for?  The only audience who would be pleased with a message like this is the Office of the General Counsel at United/Continental.  It certainly shows no cognizance of the much-touted effect of authenticity, intimacy, and a real conversation that is one of the best things to come out of Web 2.0

Your thoughts?

Ecommerce customers as “audience”, merchandising as “content”

We’ve started to think of our various digital media investments as ways of aggregating and/or monetizing a digital audience.  Obvious, perhaps, but leads us to think about things in new ways.

Consider ecommerce, for example.  From the standpoint of audience, ecommerce is a way of attracting an audience by interesting them in buying something.  An ecommerce site has, in this way, a lot in common with a content site:

  • The research for the purchase is of course, great content in and of itself.
  • The merchandising of the purchase, aside from not derailing the transaction, can please and delight the customer as well
  • The cross-sell and up-sell options are opportunities to start new cycles of research, merchandising, and purchase

It’s a bit like playing cards for money as opposed to playing cards for fun.  The cards pull in the audience, and playing for money makes the cards more exciting.

Services Innovation

It’s a truism, at least in VC circles, that “services busineses don’t scale”.  In the very basic sense that services are provided by a human (I’m talking about real services businesses here and not XaaS businesses) and that more services requires more humans, this is, of course, true.

But there’s motion in the ocean in services, and it’s worth paying some attention.

What’s new?

  • Crowdsourcing.  By chopping up a services function into minimal parts and farming the minimal parts out to a large crowd of temporary workers, one can do hefty projects cost-effectively, use “time slices” of otherwise-expensive experts, or even achieve novel insights (as with prediction markets).  Crowdsourcing is an extremely interesting area for Valhalla.

(Crowdsourcing actually reminds me of RISC (“reduced instruction set computing”), where a processor is designed to use a small instruction set at high speed and efficiency instead of a CISC (“complex instruction set computing”) architecture where each instruction is specialized but requires variable cycles to execute, hurting predictability.  Crowdsourcing = RISC, “conventional” services = CISC.  Well, CISC did win out more or less (although RISC is making a comeback in GPUs.  But we digress).)

  • DoGooders.  A significant number of youths want to do good, and are actually doing good.  Teach For America is sort of the paradigm for this kind of thing.  I don’t know if it’s significant compared to my youth, but as these new service workers come into NoGooder jobs later on they’re going to want those jobs to have some kind of significance beyond the paycheck.

  • Death of the professions.  I know the most about medicine here, which is transforming from a profession/craft business into an industry.  Since the drivers are the same, I expect we are seeing the same thing in accounting, law, science, engineering, and teaching.  What does it mean to have “professionals” without a professional esprit de corps or, more appropriately, a professional code of ethics?  I don’t know, but we’re going to find out.

Enough for one day.  Your thoughts?


    Hospitals: Still in the Middle Ages

    I had occasion a year ago to blog about my experiences as a customer of the hospital system.  With a relative this week undergoing elective surgery, I am reminded forcefully of how antiquated, creaky, and dysfunctional the healthcare business is, as a business.

    My relative’s surgeon, his PA, the nurses on the ward that he oversaw, and all of the staff, were individually superb.  It would be hard to find a group more knowledgeable, capable, and caring.

    As an organization, however, the hospital was really really creaky.  When we showed up for admission, we were directed to a waiting room for “same-day surgery”, where the three staffers promptly disappeared for something on the order of an hour, turning out the lights in their office and leaving us with no clue where we stood or when things would change.  The hour of our surgery approached, when our surgeon happened by in the corridor.

    “Why are you still in street clothes?” he said to us.  “I’ll fix that.” And he went backstage for a few minutes and produced one of the missing nurses.

    With the surgery in progress, we were told to go downstairs to another waiting room where we could get a paging device that would let us know when we could visit in the recovery room.

    Downstairs, the same thing: nobody in charge, a bunch of patients sitting around.  The staffer in charge had gone away some time ago.  No idea when she would be back.  Maybe 20 minutes later she sauntered back in.  Gave us a pager.

    After two hourse had passed (the length of time we had been told the surgery would last) another staffer came around and collected our pagers.  They were done with them for the day.  No problem: security would tell us when the surgery was over.

    Security, it turned out, was the lynchpin of this hospital’s operation.  Security knew everything; for the first time that stressful day, we felt someone was in charge.  Someone was acting as if a regular process might help patients and their families.

    All’s well that ends well.  The surgery went well and we are recovering.  And, once again, most of the individuals we dealt with in the hospital were superb professionals.  But somehow the hospital doesn’t work as an overall organization.

    Why not?  Thoughts?