Archive for the ‘Strategy’ Category

Optimization creates inflexibility

Friday, June 27th, 2008

Programmers know this, but I’m not talking about code.

In any field, it seems, optimization creates inflexibility. This is because optimization (maximization of performance) generally requires assumptions about the continuity of conditions.

In computer programming, you gain great speed benefits by writing graphics software in assembly language, rather than a compiled language such as C. But assembly is purpose-designed for a single microprocessor family. Change families, and all your work must be thrown away. So this optimization require one big assumption about future use of your software.

In investing, you can gain higher returns with leverage. But the greater the leverage, the more central your assumption of price stability or continuity, the greater your reliance on avoiding one really bad day, and its associated margin call.

In advertising, you can maximize conversion rates by having each CPC ad click through to a specific landing page designed for that ad. But the amount of work required to make a campaign-wide change increases by an order of magnitude.

Maximizing international trade optimizes economic output but greatly increases exposure to common-factor economic collapse, such as a discontinuity in the oil supply.

Thus optimization increases performance (narrowly measured), but also exposure to Black Swans.

How to destroy Google

Wednesday, May 21st, 2008

I like Google.  They are brilliant, non-evil, I use their various services daily, my best friend from college works there, etc.  No complaints.  The following is merely an academic exercise in business strategy.  

If you were a loser in pay-per-click search advertising (e.g. Microsoft), how would you fight back?  Taking Google’s customers is too hard, because there’s no real reason to switch.  Instead, why not just destroy the entire PPC ad market by open-sourcing the web search industry?

Imagine that MSFT and other also-rans funded a truly independent, nonprofit, open-source-based search site, which ran no ads. Computer makers could easily be convinced to use it as their default search engine, because it would reduce their dependence upon Google, without increasing their dependence on Microsoft.

This strategy would steal page views from Google.  It is not necessary to kill Google to end its dominance:  a 30-point loss in market share could drive Google into the red, throwing its virtuous cycle (earnings, stock options, recruiting) into reverse:  share collapse, layoffs, reduced retention, etc.

There are no anti-trust concerns — this would be a truly independent entity, jointly funded by a consortium, rather like Mozilla.

The technical hurdles are low.  Search engines long ago reached the “good enough” stage.  The insights of Page and Brin in the mid-1990s were so great that little has changed in the past several years.  There already exist reasonably good open-source modules for web search.  

The main hurdle is the capital expense of a big server farm.  Assuming a consortium pays for this, all that’s needed is to plug some GPL modules together, scale them up, and set this as the default search engine on all new HP and Dell computers.

Such a service would not work as well as Google.  But it would be good enough, and would be the default on every new computer.  By thus sucking revenue out of pay-per-click, MSFT could sustain the Empire for a few more years.

Why would I write this, if I love Google?  First, because it’s the truth, and second, because there is no risk that paralyzed, rudderless Microsoft would actually do it.



Traditional Strategic Advantages of Google

Tuesday, February 26th, 2008

None other than Hal Varian, luminary technology strategy guru, argues that Google has no traditional sources of strategic advantage, only a temporary knowledge curve advantage.

This seems overstated. True, Google ads do not enjoy the network effect often ascribed to them, but the company has at least the following traditional strategic advantages.

Web searches rely upon giant data centers. Google’s massively parallel data center architecture facilitates the fastest web searches, most frequent index rebuilds, etc. It also facilitates the cheapest hosted services, such as email, photos et al. They have way more computing power than anyone else. As Varian might point out, this architectural advantage is temporary. However, when everyone else inevitably copies this architecture, data centers revert to a pure scale contest. Whoever buys the most boxes wins. Google can currently outspend everyone but Microsoft; by the time Microsoft builds a similar architecture (probably years), Google will be able to outspend them, too.

For 10 years, Google has defined web service innovation. They are increasingly trusted outside the narrow field of search (email hosting, for example). As a result, increasingly, when Google releases a new web service, other things equal, they will automatically win.

The anti-spam system in Google’s hosted email enjoys network effects, because it relies upon group intelligence to decide what is spam, then rolls out the answer to all users simultaneously. The effectiveness of this system improves dramatically with scale. The largest network has the best spam control. Yes, anyone can copy the architecture, but again, by the time they do, Google will already be the biggest. They are probably #2 or #3 already, and their most logical competitors — Yahoo and Microsoft/Hotmail — are asleep at the wheel. By the time they figure it out, Google will be #1, and able to run away with this network effect.

No suggestions being made here. Just observing.

Embracing the inner geek

Tuesday, November 13th, 2007

After fighting nature for decades, the transition begins with this 20-year-old bicycle, bought in September:

In keeping with my newly overt nerdiness, I will gladly bore you for hours with technical explanations for why this is the fastest non-recumbent bicycle you can buy. But the short answer is rigid wheels, 110psi tires, full suspension, and rigid space frame.

Hard to believe? This exact model holds the world speed record for non-recumbent bicycles, at over 51mph with a full fairing.

This bicycle is a strategic decision, in that it gains a persistent edge by a path that competitors cannot or will not follow. Practically no one will ride this bike, no matter how fast, because it looks weird.

I consistently pass serious athletes in Lycra finery and $4,000 carbon bikes. They stare at the rusty bike, the blue jeans and flip flops, and pump furiously, but can’t catch up. Yet, even after seeing it with their own eyes, they will never make the obvious competitive move, which is to buy this bike. So I keep passing them. That’s strategy.

How Steve does it

Thursday, May 24th, 2007

Why is Steve Jobs so good at making products people want? An Apple engineer I know had this to say on September 25, 2005:

“Say what you will about SJ, but he (and/or the people he surrounds himself with) are genius. I feel part of the secret is simply having built a culture where everyone cares about things down to the smallest detail — we sure do that with some of the stuff I’m
working on now — but there’s also the quality of what I’ve always thought of as ‘yummy.’ You make a yummy little morsel of a couple of different kinds of already-existing technology and mash them together, make sure it is (deceptively) simple, and make sure the fit+finish is perfect, and use ultra-clean modern design. Still, it’s that ‘yummy’ quality that’s paramount.”

Can this be copied? Possibly not. Why?

Aesthetics and integration are organizationally thankless. They require heavy up-front investment in things that are not identifiable product features, and thus carry benefits that are hard to communicate. Integration is particularly thankless, because it requires iterative design, causing unpredictable delays and blown budgets, again in return for no specific feature. Middle managers are typically punished for such things, and learn to avoid them.

Most companies are unable to measure the benefit of aesthetics and integration. So even when these ideas can be communicated, their benefit (price premium, increased sales, increased customer satisfaction) cannot be assessed internally before release, and so there is no way to decide how much to invest. So no one else does it.

Simplicity is worth more, but appears less.Ease of use often takes the form of superficial simplicity, creating the appearance of fewer features, and thus is potentially risky to an organization unable to assess its true market impact.

Apple has invested decades in a corporate culture that values aesthetics.This is singular. Apple has been indoctrinating new hires with its aesthetic “rules” for at least twenty years, giving them a big advantage at producing products where such things are valued. Since everyone in a software company imagines himself a creative genius, it’s a huge advantage for Apple to get everyone pointed in the same direction from the very beginning. That said, it’s probably safe to say Apple’s vision is all Steve, and wouldn’t survive his departure. Similar to Akio Morita at Sony: when he passed away in 1999, the company ran down like an old battery.

To the extent Jobs’ Apple enforces a unified, simple, aesthetic design sense, it’s unreasonable to expect any other company to copy. Thus it constitutes a sustainable competitive advantage.

Strategy begins with tactical asymmetry

Friday, March 30th, 2007

What happens when every MBA has read Porter? or when every general has read Clausewitz and Sun Tzu? What happens to strategy when everyone knows about it?

What’s left is the exploitation of tactical asymmetry — usually in the form of a competitor’s inflexibility.

To see why, try a thought experiment: what if a natural monopoly opportunity is revealed simultaneously to two companies of equivalent resources and competitive position?

If firms were rational agents with equivalent information, then each would invest one half of the present value of the monopoly over its lifetime, and NPV across that entire industry would be near zero, no matter who won.

But real-world agents are not rational, and the bias is toward overinvestment. In particular, individuals within firms have personal incentives to overestimate market size and odds of success, and to abandon failed efforts only reluctantly. As a result, in the thought experiment, we would tend to get negative NPV in real-world cases.

This negative NPV results from the symmetry of the two positions: even though the industry may be a highly profitable natural monopoly in the steady state, no firm can reach a dominant position. Yet it continues to invest as if it can.

Historical examples abound. English canals and American railroads, obviously natural monopolies, yielded many investment failures in part because of such overinvestment.

The symmetry is the problem. So, to have a reasonable expectation of positive NPV in chasing a monopoly, you must begin with a significant tactical asymmetry over competitors — resources (money, incumbency), information (insight, access, surprise), or preferably all of the above.

For a venture capitalist, this would be the second half of an investment evaluation. The first half would be “is the market likely to be a monopoly (i.e., to yield supernormal profits to the winners) at all?” If so, then to have fair odds of success, there must be an tactical advantage that may be exploited as a strategy.

Honest Search Optimization

Thursday, May 25th, 2006


There are two approaches to promoting websites, one adversarial and one
cooperative. You can either fool search engines into thinking you are
useful, or you can actually become useful, attracting search engines and
users. Both can work. But in the long run, the latter is more effective
with less effort.

My qualifications

I’m not a recognized SEO expert. I don’t consult or sell books. This is
the only thing I’ve written on the subject. Yet in a few years, in my spare
time, with a budget of zero, I have sent several websites to #1 at major
search engines, generating 50,000 to 100,000 page views a month.

Since it cost me nothing to learn this, I will now tell you, for free, how to
do it.

Do this for each site

  1. Choose a domain name that describes the content.
  2. Put interesting and useful content there.
  3. Get listed at
  4. Describe each page clearly and succinctly in the page title and body.
  5. Get links from established, reputable sites with similar content.
  6. Wait several months.

That’s it. You don’t need to hire anyone, nor buy any tools or books.  This method works for nearly every site, nearly every time, for years, with almost no maintenance.

Why it works better

This approach arises from a simple guiding principle:

Understand how search engines work, and what they want to do — and cooperate.

Traditional promoters want to use Google like a megaphone, because that’s how news media works these days. But Google is not a megaphone. It is a listening device.

To really understand that, and to appreciate the cooperative approach to SEO, you need to be in right frame of mind, which by itself can be difficult for the typical person reading an article like this. You very likely work in a frantically competitive job, involving long hours and the threat of layoff or reassignment. This tends to create a short-term, shortcut, screw-you mindset. So first, a brief exercise to distance us from that world.

Preparatory Exercise

As you’re sweating alone over a particularly tedious project at 11pm on a Saturday night, please stand up, look into your bloodshot eyes in the mirror, and read the following:

“Out, out, brief candle!
Life’s but a walking shadow, a poor player
That struts and frets his hour upon the stage
And then is heard no more: it is a tale
Told by an idiot, full of sound and fury,
Signifying nothing.”

If, at this point, you still feel highly motivated to tap away like a chimpanzee, then there’s nothing I can do for you. Try clicking some helpful consultant’s ads — they’ll probably sell you cool, colorful charts or something.

On the other hand, if you are now in tears, bemoaning your wasted youth, you are mentally prepared for cooperative SEO.

Choose efforts that will last

It’s not widely recognized that Macbeth was discussing search optimization philosophy. Do things that will still have value in a year or two, and you’ll feel less like a chimpanzee. Even a tiny effect is valuable, if it lasts for years with no further effort.

The goal in search optimization — the true goal — is to gain a position whose value (measured in dollars, readers, or whatever else is important to you) greatly exceeds the effort required to get there and remain there.

This means working with those features of search engines that are utterly stable, and likely to change slowly or never.

Stable features of search engines

Search engines use a huge and ever-increasing array of techniques to measure importance and relevance, and to filter out scams and spam. But the following have been stable for about 10 years.

  1. Your page’s importance is based on the number and importance of links coming in to your page from other pages. 
  2. Importance is calculated iteratively, and must begin from a set of initial conditions provided by a reference directory. For Google, the initial conditions are provided
    primarily by 
  3. Relevance is inferred from the similarity between a user query and the contents of a page and pages linking to it.

The reason this hasn’t changed in 10 years is that it works. It is a powerful set of insights into delivering what people need. So powerful, in fact, that it’s difficult to improve upon. And difficult to change.

Stated another way, the search engines have solved the search problem so well that they’re now beneficially unable to change their basic ranking approach. That inflexibility means that if you can gain high rank according to the above three criteria, you will tend to sustain that rank indefinitely, with no further effort. That’s the promise offered by cooperative search optimization. Now, let’s look at the alternative.

Gaming the system — beyond ethics, it’s a waste of time

Looking at the stable features above, you can see the potential adversarial strategies: link farms, link exchanges and so forth. Some of these probably do work — for a while.

Let’s say you had no ethical qualms about gaming the system. You’re only concerned about return on investment. Is this an effective way to spend your time or money? Probably not, because you are exploiting an unstable set of loopholes.

The search companies have huge teams of smart guys dedicated to prevent cheating. As a result, adversarial strategies are a constant arms race. You would need constantly to invent new strategies as they discover and shut down your old ones. All your old work becomes worthless within a few months. A sound
and fury, signifying nothing.

Spending your time becoming interesting

By contrast, providing end users with something uniquely useful is a great investment of time or money. To see why, look at the two main hurdles you need to jump in gaining a good search rank: DMOZ and inbound links.

People know DMOZ is critical to a good search ranking, so this creaky all-volunteer Web directory is utterly overwhelmed by requests from websites to get listed. Yes, there are probably ways to sneak in, but the most straightforward solution is to offer something unique, useful and authoritative. The DMOZ reviewer will immediately recognize this in your site, if you have chosen the right category and are truly providing unique utility.

Inbound links are easier to obtain than a DMOZ listing. But authoritative, closely related inbound links are not. You need to create something sufficiently distinguished to warrant attention from people with no vested interest in your success.

Your content is not automatically unique or useful

Note the previous section is not saying, “Spend all your time creating content.” If that were all you needed, every high school student’s blog would get more traffic than the Toyota home page. No, the previous section says to spend effort creating not just content, but useful content. And not just useful content, but uniquely useful content.

My blog (I do write one) is unique, but not useful. As a result, no one links to it, and so it’s mostly invisible on the Web.  Similarly, if you create a website full of recipes, it is likely useful, but absolutely not unique, and thus unlikely to gain a good rank.

You need to think about what people need, what you can provide, how you are unique, and how that uniqueness might be valuable. This is a completely different type of challenge from frantic chimpanzee tapping. It involves talking to people, asking questions, and just sitting in a quiet room, thinking.

Knowing the difference between adversarial and cooperative

Now you know the answer to success on the Web. In the abstract. But in practice, for someone not accustomed to thinking in this way, it may not be obvious which activities are adversarial and which are cooperative. Here are a few examples.

Adversarial (basic intent is to fool a search engine)

Link exchanges offered by mass email.

Buy a domain with existing high rank, intending to change the content.

Cooperative (basic intent is to deliver value to site visitor)

Buy a domain with existing high rank, slowly adding similar content.

Get a legitimate online newspaper to link to your site.

How would you rather spend your time?

In conclusion, your mom was right that cheaters never prosper. But beyond that, why would you want to live like that? Always looking over your shoulder, trying to stay a step ahead of the anti-spam cops. Anyone smart enough to do that successfully for a while is more than smart enough to generate actual content, of actual unique value, to benefit a lot of people.  

Why would you want to do it any differently?

Note: If you find this article useful, please link to it, thereby helping to prove its thesis. :-)


Handheld Game Software Business Strategy

Monday, September 1st, 2003

by Bill Mitchell

Originally published in Mobility magazine, September 2003.

When Tiger Woods PGA TOUR Golf for
was released in
1999, the deck was stacked heavily toward success.  Palm hardware sales were
doubling every 10 months, Tiger was the first major gaming brand on the Palm platform,
and there was almost no commercial-grade competition.  Best of all, for several
months, Tiger Palm was the only software cartridge available at the Handspring

How times have changed.  Today, with industry growth at zero
and more than a thousand games available, even brand names and commercial polish
are not a guarantee of success.  Most new titles simply vanish without a trace,
and neither great design nor promotion alone will change that.  If you are
serious about winning in this business, then what you need, more than anything
else, is a strategy.

What is “strategy?”  Most people think it’s synonymous with “plan.” 
But a great strategy means much more than just a plan.  At its heart, strategy
means this:  doing what customers value highly, but competitors can’t copy.

That’s a simple statement with a lot of meaning behind it. 
Let’s consider a couple of examples.

Say you intend to write the first PalmOS bicycle racing
game.  You are certain you can build a quality product;  moreover, since it’s a
popular genre not yet available on Palm, you figure you can probably get some
free publicity.

This constitutes a good plan, but not a strategy.  Why? 
Because anyone could copy you — and probably will, especially if your game is
initially successful.  A bad omen for the long run.

By contrast, a strategy, to use a whimsical example, would
be to publish your game while living in Nicaragua.  Rock-bottom living costs
would mean a decent return on your invested time and money, yet you could price
the game so low that no one in an industrialized country could ever compete. 
In general, developing a permanent cost advantage is a strategy, because it
increases your value to the customer in a way (lower price) that most
competitors cannot easily imitate.

Another strategy might be to obtain long-term rights to use
a branded theme, such as the Tour de France or Greg LeMonde.  The familiarity
of a known brand has great influence — it’s considered one of the few sure
things in the videogame industry.  But note the weakness in this strategy:  all
of your advantage results from something you are renting from another company. 
What the licensor giveth, he may eventually taketh away, or at least reprice to
your disadvantage.

To use a more subtle example, strategy could rely upon a permanent
distribution advantage.  For example, if you could obtain exclusive PalmOS
rights to the mailing list for a popular Macintosh-based cycling game, you
could simply contact every Mac user and ask them to download your new PalmOS
product.  Your product would instantly vault to the top of its category at
PalmGear and Handango, in a way that a competitor would then have great
difficulty copying.  Since most new visitors to PalmGear and Handango simply
download whatever is already popular, as shown by downloads, you gain a persistent
advantage on the basis of just a few plaintive emails.

Still more subtle is to “design” a distribution niche around
yourself, such that no one at all competes with you.  For example, if you set
up a table selling CD-ROMs to spectators at the finish line of the Tour de
France, you would have no competition at all.  Of course, you might not sell
many copies, either.  It’s quite difficult to come up with a distribution niche
that is both strategic and cost-effective, but if you can do it, the payoff is
great.  There’s more on this in the real-life examples below.


So how do you go out and develop your own strategy?  Coming
up with an answer to this puzzle is very hard, and rightly so:  it practically
guarantees success in any industry.  Start by thinking about unusual long-term
experience you may have, or unusual personal contacts, which might be of value
in developing or promoting your application.  If that doesn’t work, try
brainstorming unusual ways to develop or distribute;  ways to attract or
enforce customer loyalty;  or ways to gain benefit from the existing success of
someone else.  Crafting good strategy is a creative puzzle to match the best
problems in game design.  The key is to remember that it’s not enough just to
have something no one else has.  You must have something of value to customers
that no one else has today, and that no one else can easily copy,
even if they tried.

If you look closely at the sustained success stories in the
Palm software industry, you will see solid applied strategy over and over again
– whether on purpose or by sheer luck.  Here are just a few examples, all drawn
from the Palm economy, based on my observations of public information since

  • Astraware appears to use the mailing list from its initial
    megahit, Bejeweled, to promote its growing stable of products.  In effect, this
    list provides a permanent cost advantage in advertising any other product.

  • There are over two dozen golf
    scoring applications available, yet IntelliGolf consistently
    outsells most or all of them.  They did this in part by niche distribution:  airline
    magazines.  This was no doubt expensive, but it guaranteed that they had no
    competitors in their sales channel.  Since travelers are also often Palm owners
    and golfers, the tactic is a logical one.  It’s also a pretty good strategy,
    since it would now be very difficult for a second golf scorer to enter through
    the same channel alongside IntelliGolf.

  • PalmAid from Emcon Emsys
    makes money by virtue of a low-cost operation:  the company is based in India.  As
    long as the product they offer is of high quality, any Western competitor might
    as well just give up on direct competition.  To beat PalmAid, a new entrant
    would require some other advantage, such as cheaper access to advertising.

This is not to say product design doesn’t matter.  But the
Palm games business is so competitive that great design is simply a starting
point.  Many Palm games are released with adequate attention to design — and
sometimes promotion, too — but very few have a well-thought-out strategy.  By
taking that extra step, you can gain a jump not just on today’s competitors,
but on tomorrow’s as well.