In all humility, I’m the best competitive analyst I’ve ever met. I’m sure there are people who are better than me, but I’ve never met them, so . . . . For example, despite some wobbling, I called ZFS on Mac months before anyone else I’m aware of did. But this isn’t about me.

It is total war and your country needs you
The man who inspired me to get serious about competitive analysis was an English physicist named Reginald V. Jones. Jones led a British Scientific Intelligence group during World War II. He was in his 20’s when the start of the war, and his friendship with Henry Tizard and Frederick Lindemann, thrust him into the scientific assessment of German technology.

It is a great story, well told in his autobiography Most Secret War. I think any reader of StorageMojo would enjoy it.

Jones’ Principles
Jones wrote a plan for scientific intelligence that outlined the key issues. Issues that are not terribly different when performing commercial competitive analysis today.

Jones defined his mission as “. . . early warning of the adoption of new weapons and methods by . . . enemies.” He looked at the stages of weapons systems development cycles, which look a lot like high-tech product development.

  1. General scientific research
  2. Some one thinks of a possible military application
  3. Ad hoc research and small-scale trials
  4. If successful, large-scale trials
  5. Finally, product roll-out

Stages 3 and 4 are the most difficult to observe, especially during wartime conditions, but there are ways. Jones lists how information leaks out:

  1. Accidental indiscretions which, if pieced together, valuable info derived.
  2. “Indiscretions encouraged by alcohol and/or mistresses.” I’ve never had the budget to employ mistresses, but the increasing number of female execs calls the value of the technique into question. Alchohol still works, though.
  3. Information that can’t be kept secret, such as radar transmissions or, today, the blogging schedules of key engineers.
  4. Agents in the enemy’s research teams – difficult then, impossible now for legal reasons.
  5. Information from disaffected nationals – or angry ex-employees.

Jones was a crafty one, and two goals for the organization he proposed reflected this:

  • Mislead enemies about our own weapons
  • Mislead the enemy about the success of his own weapons

Most of that would be wasted effort in today’s storage industry, since folks seem to pay little attention to competitive strategy, while paying a great deal of attention to day-to-day tactical strategy.

The big problem of WWII: RADAR
Popping a pizza in the microwave is the closest we get to radar most days, yet it was the technical challenge of WWII. While the A-bomb had a bigger impact on post-war national security calculations, radar has changed more lives than any other WWII technology. Arguably, computers are also a WWII technology, yet the most important developments, such as Project Whirlwind, came in the 1950s. And the storage problems weren’t figured out until the 50s either.

In 1939, at the beginning of the war, aircraft and ships were essentially blind at night. British scientists developed the resonant-cavity magnetron, a low-cost device for creating powerful microwaves, whose short wavelengths made it possible to distinguish much small objects, such as aircraft, by capturing the radio reflections.

The “Battle of the Beams” was fought on several fronts, including airborne radar for nightfighters, high-frequency navigation beams, early detection of air raids and the guidance of the jet-propelled V1 and the V3 rockets. It is a fascinating story of compressed technological development in the furnace of war.

Technical excellence and operational flexibility
Jones’s developed other ideas that have influenced me. For example, here he contrasts German and British approaches to technical issues:

I had come to have a ‘feel’ for the way the Germans did things. They would take simple ideas, and put them straight into practice no matter what technical effort was involved, because they had a far greater command of precision engineering than we had . . . . [W]e would take the simple idea, look for the technical snags in the way of its realization, and think of ways of getting round them without . . . great precision of design or workmanship. In the end, I suspect that we often took as much trouble in avoiding the difficulties as the Germans did in overcoming them by good workmanship; as it turned out . . . the advantage . . . lay with us, because while the German equipment was technically very good, it was also less adaptable, and we could more easily change ours to meet a new situation.

In Silicon Valley this is like the “ship early and often” philosophy of many software companies.

Understaff for success
Staffing is another of Jones’ ideas that resonated for me. As he wrote in a memo during the war:

[O]ur policy [is] to keep the staff to its smallest limits consistent with safety, because the larger the field any one man can cover, the more chance there is of those furtunate correlations which only occur when one brain and one memory can cennect two more more remotely gathered facts. Moreover, a large staff . . . requires so much administration that its head has little chance of real work himself, and he cannot therefore speak with that certainty which arises only from intimate contact with the facts.

I’ve had dire results the few times I’ve ignored this wise principle. It is far better to fire people and keep the remainder busy than it is to keep people on with too little to do. A hard truth, but there it is.

The StorageMojo take
The industry is at a major threshold today, similar to where it was in 1992, when the first storage arrays started coming to market. Today the technology is the storage cluster.

Just as in 1992 there are incumbents making great margins on the mainstream product set – then, software mirroring; now, storage arrays – who are taking different views of of the coming change. Some are in denial, others are getting their toes wet, others, mostly challengers, are diving in as fast as they can. Remember, in 1992 EMC was a little company whose business was mostly making add-on memory boards for the dying minicomputer firms of Route 128. And look at them now!

In times of great change, competitive strategy is a critical success factor. Competitive strategy has atrophied over the last few years of relative competitive stability among storage vendors. A major problem for the incumbents will developing and executing a competitive strategy for storage clusters. They not only have the competitors to watch, but they have to watch the ravenous Wall Street hordes at their backs. Ultimately though, every big array maker will have to transition their focus to storage clusters over the next five years or get rolled over.

Here’s another prediction
Based only on a hunch, I believe that two of the lead engineers on Sun’s ZFS, architect Jeff Bonwick and Bill Moore, have been hard at work in Cupertino getting Apple’s Leopard implementation of ZFS ready for prime time. They haven’t said word one to me about any such thing, but their blogging has slowed way-y-y down, and I never see them on the zfs-discuss thread. Two smart guys like that must be doing something and I think that’s it. Now, if I only had spies in Apple’s cafeteria. . . .

Comments welcome, as always. Have you done any cool competitive analysis you can brag on? I’d like to hear it.