A View from Beneath the Dancing Elephant – another book

Now this book is different from Cringely’s. It was written by Peter E. Greulich, an insider. He tells the story of the Watsons in IBM and the impact they had on a positive corporate culture. Something inexistent today, as the author believes.
While Cringely relies on information he can gather from the outside, Peter E. Greulich has worked for IBM for about 30 years in various positions, including management. He has many interesting stories to tell, but the essence of the book is his grief with IBM’s current situation and how it came to it. For him it started with the reign of the white knight in the person of Louis V. Gerstner, Jr. who at first was his hero. Later, especially after changing the retirement schema, he started to feel, that IBM was going in the wrong direction. With Sam Palmisano, it got worse. Management started to use rules and tools, that were either incomplete or not understood. As an example he recalls Tivoli Configuration Manager. At first a game changing product, but for the lack of continuous developement it became obsolete.

The development director said, “TCM is a cash cow. Why should I spend money on a product that is nothing but pure profit?” He believed that products in this market were commodities with low margins, and should be left to others to waste their profits on. We never convinced him otherwise.

The cow in „cash cow“ is quite a good example for this. When you got a cow you have still to feed it and you have to look for a replacement – its offsprings normally – while the profit you make with the milk, pays also for growing (cash-)cow. But until the old cash cow becomes a poor dog (meat loaf), it is the duty of the product manager to push the poor dog as far in the future as possible. In the software business that means constant upgrades and enhancements and at least some marketing, until nobody has any new ideas (good ideas that is) to keep the milk flowing. By then, the rising star should be ready to take over and eventually become the next cash cow. That’s the ideal world and it does not always happen but still, Just abandoning a product, just because one thinks it good enough right now, without having the next thing in the drawer, is rather short-sighted.

The other story was with „Lean“, even so he does not mention the word. IBM used Toyota’s „Lean“ initiative in a perverted kind of way. The real „Lean“ wants to free resources by using everyone’s knowledge to eliminate inefficiency AND augment quality. The idea is not to find ways to fire people but to free resources to use them better. That’s one of the main principles of Lean, not to fire people. Greulich recounts how he lived through that time, when more and more of his friends disappeared and how that hurt his and others work. Key people were „resource actioned“ (fired), because that department just had a quota to fill. It works exactly like Microsoft’s performance management. If you have a bad quarter, you are a goner. Does not matter, if your family just died in a train crash. Or in other words, if you had Einstein, Newton, Curie and Pauli on your team, Curie probably would have let go, due to her morning sickness and the resulting slightly lower performance at 7:30 AM. With that IBM became even more inefficient and the quality dropped. To make up for the loss, IBM bought more and more other companies for a lot of money, only to bluewash and crush them. Small example: Nitix. Developing their own products ahead of time with freed resources from a proper Lean program, could have saved and made IBM billions (in the case of Nitix, IBM once had a lot more Linux developers). But that would have cost shareholder-value. Now that’s bad, if the friends at Wallstreet don’t like you anymore because you don’t fill their pockets enough. It might be a wild guess, but if you look a the prices IBM paid for some companies and the number of products that after a few years are still alive, one gets some doubts about that strategy.
This kind of senseless loss of manpower (today: knowledge drain), did not happen 20 years before, when Greulich had a hard time himself as a single parent with three small kids. His friends at IBM and his manager just helped him through this time and he is forever grateful. First, it did not hurt IBM, because the work got done anyway and second, Greulich committed himself even more. If that isn’t a win-win situation? And don’t tell me that isn’t possible today. Many companies all over the world show that you can prosper and accepting a social responsibility.*
The numbers are also different to Cringeley’s. According to Greulich, for every old IBMer (in the US, Europe and so on) let go, three to five Indians or Chinese developers were hired and still they can not match the quality and therefore the speed of the old team. Not even financially that makes sense. As a side note, many big european companies in-source again. Even Apple starts to build computers in the US. Do they probably know something IBM does not? None of all those companies do that because suddenly labor costs dropped below China’s, oh no, they do it because it makes sense financially. More money in the long run!

Worth a read? Definitely. It does give you a point of view of somebody inside. Cringely looks for sensation (that’s his job) and Greulich is just sad. Somewhere in the middle lies probably the truth, but all in all, the pictures match.

A View from Beneath the Dancing Elephant
Rediscovering IBM’s Corporate Constitution

by Peter E. Greulich

* One of the best example came from the union representative at Porsche during an interview at Le Mans. You can be nice to your workforce and demand that extra effort if somebody wants to work for Porsche. That’s the view of the „Betriebsrat„! In early days the enemy inside, today an important asset to the company. No wonder VW wanted one for their american workforce, because it pays off.

Snow and project management

One week of skiing. Weather was nice, but way too warm. Snow was too wet. I hate skiing in spring, really, I don’t mind – 15 °C but + 5 °C at 2500 m is silly. Days were packed but I still had time to read a book of Clarence (Kelly) Johnson. His work always interested me, because he is one of the most successful aircraft designers. Probably you don’t know his name, but Lockheed’s Skunk Works should be known to everybody. Also the planes he and his team designed:
P38 Lightning
P80 Shooting Star
F104 Star Fighter
A12 – SR71 Blackbird
… to name a few
Apart from the technical stuff he describes, his way of running the shop and getting projects done in time and below budget is remarkable. It is almost unimaginable today, that money is payed back to the customer and he finished most projects, even very difficult ones, before dead line.
He had 14 points of management. Not all of them are suitable to IT projects and corporate customers, but some certainly are.

  1. The Skunk Works manager must be delegated practically complete control of his program in all aspects. He should report to a division president or higher.
    Definitely! IT Project managers must have the absolute power.
  2. Strong but small project offices must be provided both by the military and industry.
    SMALL! You can easily replace the military with a corporate customer. One reason why projects go awry, is the long chain of command. Many managers are more interested in covering their ass, than taking decisions.
  3. The number of people having any connection with the project must be restricted in an almost vicious manner. Use a small number of good people (10% to 25% compared to the so-called normal systems).
    Logic, no?
  4. A very simple drawing and drawing release system with great flexibility for making changes must be provided.
    That works for other documents, too.
  5. There must be a minimum number of reports required, but important work must be recorded thoroughly.
    Doing reports is important but unproductive and most of the time nobody reads them anyway. Therefore keeping it to a minimum makes sense.

  6. There must be a monthly cost review covering not only what has been spent and committed but also projected costs to the conclusion of the program. Don’t have the books 90 days late, and don’t surprise the customer with sudden overruns.
    No comment necessary.
  7. The contractor must be delegated and must assume more than normal responsibility to get good vendor bids for subcontract on the project. Commercial bid procedures are very often better than military ones.
  8. The inspection system as currently used by the Skunk Works, which has been approved by both the Air Force and Navy, meets the intent of existing military requirements and should be used on new projects. Push more basic inspection responsibility back to subcontractors and vendors. Don’t duplicate so much inspection.
    In some cases four eyes see more, but 16 is overdoing it.
  9. The contractor must be delegated the authority to test his final product in flight. He can and must test it in the initial stages. If he doesn’t, he rapidly loses his competency to design other vehicles.
    OK, that is a bit difficult, since we often don’t need that functionality, but „eat your own dog food“ goes in that direction.
  10. The specifications applying to the hardware must be agreed to well in advance of contracting. The Skunk Works practice of having a specification section stating clearly which important military specification items will not knowingly be complied with and reasons therefore is highly recommended.
    Very, very important. Many projects fail, because the desired outcom wasn’t clear at all. Customer and contractor had different ideas. Measure twice, cut once.
  11. Funding a program must be timely so that the contractor doesn’t have to keep running to the bank to support government projects.
    That’s only fair.
  12. There must be mutual trust between the military project organization and the contractor with very close cooperation and liaison on a day-to-day basis. This cuts down misunderstanding and correspondence to an absolute minimum.
    No cc, no bcc, one email a day or whatever works. I strongly suggest a team room and no email conversations at all.
  13. Access by outsiders to the project and its personnel must be strictly controlled by appropriate security measures.
    That could prove useful, if too many want to influence.
  14. Because only a few people will be used in engineering and most other areas, ways must be provided to reward good performance by pay not based on the number of personnel supervised.
    The guys who do the work are important. Not the ones pushing excell sheets and reports around.

    He made a few other points in his book.
    In my own words:

    Don’t work overtime. From the start of the project, work on it with constant pressure. Don’t be lazy at the start and get in a hurry in the end. Otherwise peoples performance will fall towards the end, when pressure rises.

    We owe our employees good working conditions and some kind of job security.
    The „you are below average therefore you should look for another job“ mentality that is more and more common today is just plain stupid and shows that the Cxx-level has not understood statistics or are just using this argument to get rid of people. By far most people are just average, including the Cxx-level.

    Keep everybody close together. In the Skunk Works, engineers worked in the same building or even rooms alongside the mechanics, where the parts and the planes were build. It’s not a bad thing, if the engineers get their hands dirty.
    In our world of home office and working everywhere, this sounds odd, but it still holds true. People who work close together and are in continuous communication get better results.

    Let people make mistakes and let them fix them. If people are afraid to make mistakes, they are more concerned about not making mistakes than about the projects. Rather look for the reason why mistakes happen, than for the culprit.

    A book worth to read.

The Starship Diaries

Nope, that’s not about science fiction. I just read the book from Dallas Kachan about his voyage around the world. He was one of the lucky ones, who sold his internet company at the right time.
It took him two years flying around the globe in the coolest – IMHO – GA plane ever built, the Beech Starship I.
The book is not only a list of places he visited, but also about his reflections about the world in general. The people he met in places the common tourist never see, changed his view of the world and especially about the western world.
It took him only about flight 200 hours to fly the 45’403 miles. Something he could have done easily in a few weeks, but he took his time. Stayed sometimes month in one place. Sometimes he had to, because the plane had to be serviced – due to bullet holes – or fuel wasn’t available or he got sick on a remote island in the pacific. He also talks openly about the mistakes he made as a pilot. He flew a 16’000 lb aircraft single pilot, which is pretty stressful in itself. He dozed of over the jungle in Africa and other things. You never get the impression that he thinks that he is one hell of pilot, which nobody would believe anyway with only a couple of hundred hours under his belt when he took of from California. I wonder if anybody from the FAA ever read his book and had a friendly chat with him, for example about the extension cord for his headset. He wanted to be able to use the potty in the back of the plane while flying.

There are old pilots and bold pilots, but now old and bold pilots.

Unfortunately the Starship is history. Developed by Burt Rutan for Beechcraft, it was way ahead of its time. Just when the public began to accept it, Beech pulled the plug on it after only 53 produced and bought them back. Only about five of them are still flying (there are rumors about a sixth being restored to flying status). My chances to fly it one day are pretty slim.

Worth a read, if you are an aviation buff.