I started the Internet Explorer team at Microsoft in October, 1994. Without benefit of Google or Amazon.com, I found this book, which answered many strategic questions for me about how Microsoft should approach the Internet. The case studies and analyses presented by author Jim Utterback 22 years ago are still vibrant and relevant today.
Please read on for my March, 1995, book review and my thoughts on the implications for Microsoft (the “To” list were all senior product leaders).
[Note: Two months later, I expanded on these thoughts, predicting the rise of the Internet and the decline of Windows in my May, 1995, memo The Web is the Next Platform.]
Sent: Saturday, March 11, 1995 1:41 PM
To: John Ludwig; Brad Silverberg; Paul Maritz; Steven Sinofsky; Peter Pathe; Anthony Bay; Nathan Myhrvold; Chris Peters; Ed Fries
Subject: Book review: Mastering the Dynamics of Innovation
I heartily recommend this book — it gave me all sorts of insight into how to think about the challenges MS faces as we move into the online world.
“Mastering the Dynamics of Innovation: How companies can
seize opportunities in the face of technological change”
(c) 1994, Harvard Business School Press
James M. Utterback
Prof. Utterback is a biz school prof at MIT, and it took him 20 years to get around to publishing this puppy. He looked at lots of different industries:
o harvested ice -> man-made ice -> electromechanical refrigeration;
o gas lighting -> electric lighting -> fluorescent lighting;
o mechanical typewriter -> electric typewriter -> dedicated word processor -> PC;
o multi-step plate glass manufacturing -> Pilkington float glass process;
o daguerreotype -> wet film -> instant film (these are just a few) that went through technology change, and these are some of the observations:
o He studied who were market leaders before and after the innovation, where the innovation came from, and how the superseded technology was enhanced in an attempt to compete with the innovation.
o He introduces (at least for me) the idea of a “dominant design”, and charts how the energies of the market focus on “product innovation” until a dominant design appears, and then transitions into a “process innovation” focus.
o Innovation rarely comes from the market leader, because the leader has a vested interest in the status quo *and* is unlikely to have the entrepeneurial mindset/skills necessary to pursue an innovation that is a threat to existing products/revenue. (look no further than mainframe to mini to PC transition, and companies like IBM and DEC).
o By contrast, incremental improvement usually does come from the market leader, because they have the most resources and arguably the best understanding of their customers. (Apple Computer vs. IBM).
o When an innovation first appears, it is typically not as good as the existing market leading product. However, the innovation (a priori) has much more upside potential, and will eventually blow by the prior technology. (Harvested ice was cheapest in New England, since shipping costs were low; man-made ice (electro/mechanical) was initially cost effective only in the southern US and South America; but eventually harvested ice lost out even in N.E.).
o The appearance of an innovation may spur the market leader to improve the current technology, but it still isn’t enough to ward off the eventual success of the new innovation (when Edison introduced electric lighting, the gas companies came up with a modified gas light that was *5 times* as efficient (the gas fire heated a piece of metal which emitted much more light) as the old gas light, but it still couldn’t compete w/Electricity).
o Just because you understand and keep your existing customers happy, that doesn’t mean you’re not susceptible to getting blown away by some new technology — an innovation can appear may not be quite as good as the existing product in some ways, but may provide other, offsetting benefits.
(IBM was the leader in disk drive technology, but other companies introduced smaller diameter drives (5.25″, 3.5″, 2.5″, …) that, while initially slower, eventually were cheaper and faster than the old technologies, so that even mainframe computers now use the small drives).
o The one failing of this book is that, while he covers cases where companies made major bets on an innovation and were successful, he doesn’t cite companies/individuals who bet big and failed.
I think we are doing many of the right things to help us transition to the new world:
1) We have a culture that embraces and encourages change.
2) We continue to try to supersede and render obsolete our existing products, since if we don’t others will.
3) We focus on solving our current customer’s problems, but are constantly searching for new markets and different ways to factor existing markets to come up with better solutions.
But, the bottom line is surfing the waves of innovation is a very tricky business, and most companies fail to adapt.
An Application of This Thinking:
I think about Microsoft Office as being in the incremental improvement phase now. Word is a very, very good word processor, but we haven’t made major innovations for many years. [NOTE: I’m not trying to pick on word or any other piece of Office. I think those teams are doing wonderful work. But I think it is an interesting thought exercise, so please stay with me here.] So Office is a mature product, and ripe to be picked off by an innovation.
Let’s suppose (just for the sake of argument) that the Internet is that innovation.
Sure, right now, HTML documents are not as rich as Word documents, but HTML is improving on a very fast path (tables in NetScape v1.1, for example).
In addition, the world wide web permits features (like stock quotes, interactive games, multi-player games, real-time news, Internet phone, etc.) that Word doesn’t begin to approach. And, interestingly, Word is adding some of the Internet features (like links) in response, making Word a better product.
But is not impossible to see a future where someone comes out with a very flexible, powerful HTML editor that, due to its extensible nature, permits it to be more powerful than Word. Pick a Word feature, say spell check, and ask how could you do this on the Web? Simple, a company provides a spell checking web server. The web viewer submits the HTML document to the Server (possibly permuted to remove any meaning?), and it performs the check and returns another HTML document that is tagged to indicate what corrections may need to be made. Sounds clunky? Or slow?
Sure, it is right now. But as bandwidth increases, services like these (plus services we can’t even imagine right now) will be possible. And HTML, as a general purpose way to represent data if you don’t understand a tag, just ignore it), permits clients and servers to operate on data even though they may not completely understand its format.
Maybe HTML never gets to the quality of Word today for text/page layout, but it may not matter, HTML may be good enough. Think about the offset press industry. At it’s limit, it can produce 3000 DPI output, which beats the heck out of dot-matrix printers. But the cost & turn-around time of computer printers led to wide spread adoption, and with the introduction and refinement of laser printers (color, even), I wouldn’t be surprised to find that offset presses eventually are relegated to strictly high-volume printing like books. Everyone else will use laser printers.
So, just some food for thought.