Archive for the 'Strategy' Category

Google Chrome to end H.264 Support

Published
by
Victor Nicollet
on January 12, 2011
in Strategy
. 0 Comments

This has been announced yesterday on the chromium blog:

[...] we are changing Chrome’s HTML5 <video> support to make it consistent with the codecs already supported by the open Chromium project. Specifically, we are supporting the WebM (VP8) and Theora video codecs, and will consider adding support for other high-quality open codecs in the future. Though H.264 plays an important role in video, as our goal is to enable open innovation, support for the codec will be removed and our resources directed towards completely open codec technologies.

And of course, it was met with a decent and quite enjoyable wad of trolls and flames in the comments section.

H.264 is a video compression method (the complete name being H.264/MPEG-4 AVC) which happens to be the de facto standard for the up-and-coming HTML5 revolution—the next step in web technology spearheaded by Chrome, Firefox 4, Safari and maybe Internet Explorer 9, which lets the user view videos without requiring a Flash Player. The H.264 standard is supported by most modern browsers, with the notable exception of Firefox, by Adobe Flash, as well as the hardware of several mobile devices including the iPhone and iPad. Yep: there are special chips available for decoding H.264 while using less battery power.

In short, if you encode video as H.264, chances are that anyone on the planet will be able to view it.

How widespread is it? If you’ve watched YouTube recently, you’ve seen a H.264 encoded video.

By contrast, the alternative proposed above (WebM) does not have the same range of support: Firefox and Chrome do support it, Adobe Flash support should happen “any day now”, and no significant hardware implementations exist yet. This means that a WebM video can be viewed in HTML5 on Firefox and Chrome, through a Flash-based player in the other browsers, through a battery-guzzling software codec on Android phones, and through the power of your imagination on iOS devices.

So, WebM is the codec with inferior support. Why is Chrome moving away from it?

This is actually a gambit to force content producers away from H.264, because Google is uncomfortable with the fact that H.264 is patented.

Being patented means that if you write software or hardware that encodes H.264 video (such as a camera), you need to pay royalties to an organization known as MPEG-LA. The same happens if you create software or hardware that decodes H.264 video (such as including a codec in the HTML5 implementation of your browser). And even if you only distribute content (using tools provided by others), you still need to pay royalties — the one exception here is that if you distribute content for free, you will never have to pay royalties.

The amount paid is not really an issue: it’s about 2% of the price of any content you distribute that’s over 12 minutes in length, $0.15 per subscriber if you have more than 100,000 … basically, by the time it starts to hurt, you’re already bringing in a lot of cash to cover your losses. So, while a free format like WebM would let you save those royalties, it’s probably not worth losing iOS customers.The only exception here is when Firefox needs to implement the H.264 codec for its own HTML5 support, which falls in the “decoding video” section above : this would end up costing the foundation a whopping $5 million for its 270+ million users worldwide. This explains why there is no such support.

Could the rates increase over time? Read for yourself:

Interestingly, MPEG LA calls out that fees cannot increase by more than 10% per year, but the bump from 2008-2009 to 2010 is almost a 20% annual increase.

Another very real issue is that to do anything with H.264, you need a license. If you have a single video on your small web site and it’s encoded as H.264, you need to contact the MPEG-LA and ask for a license. You will pay zero royalties, but you still need the license. If you’re a corporation, this means you need your lawyer team to study the topic to determine whether said license creates any liabilities for you, which isn’t free. That, or you accept the risk of doing things without a license. Your funeral.

By comparison, WebM is a free, open standard. No paperwork, no royalties, no licensing required.

The largest pain for Google remains Apple — a move away from H.264 is possible for anyone who does not need to support iOS. This will likely end as a battle of titans between the market share of Chrome and that of iOS devices, until one of them caves in and implements the other’s format.

As for HTML5, the standard is still being built, but there are three evils one can pick from and the lesser is not easily found:

  • Making H.264 support mandatory would be an honest acknowledgement of the format’s current omnipresence, but it would spell doom for any platform that cannot pay for the decoder license — open source browsers would only be able to display H.264 through HTML5 if someone decided to pay for the license (either in a fashion similar to how Adobe pays for including H.264 in Flash, or by offering a special plugin for a fee) and the hairy tangle of open source implementation of patented algorithms is sorted out.
  • Requiring support of either H.264 or WebM (or both every browser decides) seems like a good compromise, but the cost of hosting and serving both H.264 and WebM video is steep, so every content provider will probably end up providing only one of the two and rely on Flash to display the videos on non-supporting browsers. Seems like a standardization failure to me.
  • Making WebM support mandatory is an interesting solution, but it’s a waste of a perfectly good standard (H.264 is a good standard).

I wonder if the waste and confusion caused by the patents on H.264 are worth it, especially since for many of these patents the royalies are a reward for being the first to patent the ideas, and not for working hard to find them. In a world when every idea emerges from the brains of dozens of computer scientists and engineers, what is the rationale behind patents?

Diagon Alley – When to Jump on the Bandwagon?

Published
by
Victor Nicollet
on January 10, 2011
in Strategy
. 3 Comments

In J. K. Rowling’s Harry Potter series, Diagon Alley is a fictional place in London filled to the brim with magic-related shops and institutions, hidden away from the eyes of non-magical humans. It makes sense, if you’re a wizard wishing to establish a new shop and seeking as large an audience as possible, to do so in Diagon Alley: not only would you benefit from the existing infrastructure that keeps Muggles away and allows easy access to wizards, but you would also have improved access to the customers of existing shops that happen to be in the area. More wizards walk through the alley in minutes than would walk through any other street in London over the course of an entire week.

In Paris, I enjoy the services of our very own Diagon Alley. It is actually called Rue Monsieur-Le-Prince, and it can be found stretching from the Luxembourg gardens near the Senate and north up to the Place de l’Odéon. Instead of magical shops, it is home to a massive number of Japanese sushi restaurants: Itadaki, Tokiotori, Kiotori, Yokorama, Top Sushi, Sushi Yaki and Sushi Royal among others. The number of such restaurants is surprising: this is a short, narrow one-way street, much smaller than the nearby Boulevard Saint-Michel and Boulevard Saint-Germain, which means that you literally cannot walk ten feet without seeing a sushi restaurant. The local market for sushi food is beyond saturated, and the competition between restaurants is fierce — a quick strategic analysis would determine that a new restaurant would enjoy far less competition if they were to establish themselves anywhere else.

This is not caused by a heavy immigrant population. In fact, the area is known for its high real estate rates and is out of reach of the average immigrant budget.

Nor is this a unique situation: Rue Saint-Anne is a sister street on the other side of the Seine where most of the Donburi and Okonomiyaki traditional Japanese restaurants can be found. All together, clustered on a single street.

This makes no sense. Why is this happening?

Back in 1991, Paul Krugman penned Increasing Returns and Economic Geography: as transportation costs decrease and economies of scale increase, Krugman argues, it becomes more efficient for a given industry to concentrate in a single location, because the transportation costs involved in having to move the products are paid for by the economies of scale involved. For instance, building cell phones in a single city and selling them in an entire country is cheaper than having a cell phone factory in every city.

Sushi restaurants have minor economies of scale as far as production goes: fresh fish is a bit cheaper when the delivery does not involve a one-hour detour (especially since small daily deliveries are involved to keep the fish fresh), and it’s easier to set up a new sushi restaurant in a building previously housed another that went bankrupt, because the fridges and tools and kitchen are already there. But does this explain everything?

Actually, there are economies of scale in marketing here: being an exotic food, a local market alone is not enough to support a sushi restaurant because most people don’t eat there daily. To survive, a restaurant must have a brand strong enough to cover a wider area. Few sushi restaurants have the power to create their own brand, but a dozen of them on a single short street is enough to make Rue Monsieur-Le-Prince the de facto location where sushi and yakitori can be eaten. This, in turn, creates an incentive for more sushi restaurants to appear there: customers are loyal to the street, not the individual restaurants, and a cleaner (by virtue of being newer) restaurant might attract more customers.

Of course, there’s still niche sub-markets within the street:

  • There’s at least one restaurant open for lunch or dinner, including holidays and week-ends.
  • There’s at least one restaurant open on afternoons.
  • The restaurant closest to the large Luxembourg underground station also happens to be the largest, in order to have as many customers as possible with average food quality.
  • There are two restaurants with quiet, clean private rooms for business customers.
  • There’s at least one restaurant with a nice second-floor view on the Boulevard Saint-Michel.
  • There’s at least one tightly packed chinese-friendly restaurant.

In conclusion, while there are advantages to finding your own market and having a monopoly there, there are also advantages to sticking close to existing competitors in the form of economies of scale in production and marketing: if you use mainstream tools, you will find more support for them ; if you provide a well-known kind of service, you will have an easier time convincing your customers that they need it.

The RunOrg Poll

Published
by
Victor Nicollet
on December 20, 2010
in Strategy
. 0 Comments

This week-end, we published on the RunOrg website a poll about the use and misuse of information technologies in associations. Why? Because we’re Start-Up founders, and working week-ends is part of the package. Ha ha, I mean, why this poll? Oh, right. I’ll get to that in a minute, but first, if you’re one of my french readers and happen to be a member of an association, please visit the poll page and respond. It won’t take long, and it might also answer that question.

So, why this poll? Well, there’s a distinct lack of statistical information on the topic — and although our market analysis answers the critical question of whether there’s money to be earned, it’s not nearly as precise or global as our curiosity would expect. There are many problems that information technology can solve in association, and until we take over the world in Q4 2011 we don’t expect RunOrg to solve more than a handful of these by itself, but we would still be happy to know what these other problems really are and how they are solved so far.

Gathering information about unexplored topics is also a great way to build a reputation of expertise and excellence. The results of the poll will be made available freely on the internet, in the same fashion as OkCupid’s How Races and Religions Match in Online Dating and the Mint infographics, which should improve our google ranking bring us more customers help people find the Association IT experts they have been looking for.

The results should be interesting, but the questions themselves are hopefully quite helpful: most of them are pretty obvious to natives of the information age, yet a complete surprise to everyone else. A significant number of associations from our initial sample still rely on a pen-and-paper system for handling everything, with a nice web site designed for them by a tech-savvy member that left two years ago (and that hasn’t been updated ever since). If something as simple as getting the web site right is so hard, what about handling a newsletter or mailing list? Letting members pay their subscription fee online? That our poll asks associations whether they do it hints at the possibility of doing it.

Oh, you can have people pay their subscription online? Asked one early participant (well, poll beta-tester, actually), who has since set up a PayPal subscription process.

If we’re in the business of building wondrous solutions to pervasive problems, I’d rather have everyone know that these problems can be solved. I strongly suspect one of the biggest issues with our product is that our potential clients believe no solution exists — anything we can do to alleviate this issue is welcome, even if it means those potential customers will also become more aware of our competitors’ products in the process.

Really, I’d rather have everyone using our competitors’ products than using pen and paper in the XXIst century.

Amazon-WikiLeaks sets a Scary Precedent

Published
by
Victor Nicollet
on December 2, 2010
in Strategy
. 1 Comment

For those of you living under a rock and not having an internet connection down there, here’s the story so far:

  • WikiLeaks, originally hosted in Sweden, announces that it will publish several hundred thousand U.S. classified documents.
  • A hacker runs a denial of service attack on WikiLeaks, bringing them down.
  • WikiLeaks uploads some of their data to Amazon’s S3 file hosting service, and goes live
  • Amazon pulls the plug on the WikiLeaks hosting within 48 hours.

I will not under any circumstances condemn or condone what either WikiLeaks or Amazon did there. That topic is too complex for me (and, I suspect, most people) to form an adequately justified opinion, and my biased unjustified opinions are best kept off the Internet.

On the other hand, what Amazon did was terrifying. After toiling for years to convince the general business public that moving to the Cloud does not imply accidental data loss or vicious hackers accessing your secrets, Amazon have reminded us of a basic, uncomfortable truth: they who handle your data can kill you on a whim.

«But WikiLeaks is not dead!»

I know. Keep in mind that WikiLeaks team has backup copies with strong encryption stored by a multitude of anonymous individuals, access to international hosting in a variety of safe havens, a dedicated team of sysadmins on call to move around the site and the data whenever something dies, and a willingness to fight for the availability of that information even if it entails going to jail. The reliability of their data storage exceeds that of almost any other entity on the planet, including Amazon S3. To them, having their hosting shut down is a minor inconvenience. To a normal business with their data to the Cloud, and all the bills, orders, paychecks, contracts and documents for the last year are lost: it’s an unmistakable death sentence.

How can Amazon S3 do this? Here’s the relevant part of the Amazon Web Services customer agreement:

3.4. Termination or Suspension by Us for Cause. We may suspend your right and license to use any individual Service or any set of Services, or terminate this Agreement in its entirety (and, accordingly, your right to use all Services), for cause effective as set forth below:

3.4.1. Immediately upon our notice to you in accordance with the notice provisions set forth in Section 15 below if:

[...]

(vii) we receive notice or we otherwise determine, in our sole discretion, that you may be using AWS Services for any illegal purpose or in a way that violates the law or violates, infringes, or misappropriates the rights of any third party;

This grants Amazon the right to terminate your service by snapping their fingers (and sending you an email) if there’s any hint of you doing something that might be construed as illegal.

«You’re another guy who stumbled upon a piece of legalese in a customer agreement, misunderstood it, and tells everyone how evil that corporation actually is…»

No, I’m not. I knew this termination clause had to be in there before I even looked, because it’s a fairly standard one and even my own business has it. Amazon needs this part to be able to eliminate child pornography or copyrighted books/songs/movies stored on its servers without waiting for a judge to determine that the content is actually illegal. There’s nothing evil about having that clause, and the reason we accept this situation is that we expect Amazon (and any other service) to use this power responsibly: as long as you don’t store any illegal files, you need not fear anything.

Keep in mind that while obtaining those leaked documents was illegal, distributing them has not yet been ruled illegal. It might happen in the near future on the grounds that it endangers individuals and/or governments, or it might end up under the protection of the First Amendment, and there seem to be fairly intelligent and reasonable people arguing for both sides.

You just moved your data/computations to Amazon to eliminate any data loss or denial of service risks. But now, there’s the risk of Amazon shutting down your account — what are you doing to make sure that isn’t going to happen? How do you intend to get back up once it happens? Is it really worth it?

Get Rid of your Brand

Published
by
Victor Nicollet
on November 29, 2010
in Strategy
. 1 Comment

Brands are universally recognized as being a good thing to have—a form of crystallized identity that lets your customers recognize that they’ve heard about you, or bought from you, before.

On the other hand, brands require certain sacrifices. First, for a brand to be meaningful, you need to support it by being reliably exceptional. The «exceptional» part isn’t the hardest, because unless you’re entering a market that’s already fairly saturated (say, creating a new take-away chinese food outlet) your very existence is exceptional. Or maybe it is the hardest, but I don’t want to discuss it right now so you’ll have to wait until later. On the other hand, the «reliable» part requires some pretty sexy logistics: if Coca Cola tasted differently every time you bought a new can, the brand wouldn’t stay around for long, which means most of the engineering firepower was applied to creating an easily reproduced (and transported/stored) beverage, instead of having a great but easily messed up taste.

Second, you simply cannot be an ass to your customers. In fact, if screwing your customers over is part of your business plan, you actually want to avoid having any kind of brand that would let customers recognize you. Acquiring a non-brand is fairly easy if you stick to online-only presence, generic logos and a content-free domain name. If a bittorrent download website proves utterly unhelpful and annoying, are you going to remember its name? Was it torrentroom, torrentreactor, torrentz, torrentscan, torrentornottorrent … ?

In fact, the business model of torrent download websites can be summarized as follows:

I’ve been searching for my sanity and ended up on this page. The only piece of information here is found at the bottom: «total 0 torrents found» and basically everything else on the page is just a clever setup engineered to trick the visitor into believing my sanity can be found at these five locations. The blue «sponsored search results» heading is a blatant lie—those five lines are not search results, they’re going to appear exactly as shown regardless of what you’re searching for (except the initial bold text, replaced with your search keywords), and it’s all designed to be as desirable as possible: TRUSTED DOWNLOAD (trusted by whom?), Fast Download, Full Version, Rapidshare, sizes of 1.31 GB or ~700MB that are fairly typical of movies, and near-100% health.

Clicking through leads to a web site that will try to extract as much money as possible from you before you finally notice that my sanity is nowhere to be found.

I can accept that some people online are stupid enough to pay an anonymous shady website in the hopes of watching that movie without paying the movie publisher. I refuse to believe that anyone would be fooled more than once by this tactic. But just to be on the safe side, you don’t want people to remember you after you’ve done this to them.

This no-branding, screw-the-customer approach works pretty well for areas where:

  • Relying on repeat business is impractical
  • There are no dominant brands that own the market.
  • There is little to no contact between potential customers.

Torrent search websites are one example: repeat business is impractical for the same legal reasons as the absence of dominant brands (as soon as you become noticeable, the RIAA/MPAA shut you down), and people tend to remain anonymous while searching for torrents so there is no contact between such people.

Another example would be plumbers. Repeat business is hard to get (the better you are, the longer your customers can go on without calling for you again), there are no top-of-mind plumber brands around that people can think of, and there are no communities for people with plumbing needs to talk to each other. This is an excellent environment for all kinds of shady deals. A recent example a friend had to deal with:

  1. Your water heater takes its own life.
  2. Being a creature of the night internet, you google around for a nice replacement model.
  3. You look for a plumber who can deliver and install that replacement model. No helpful plumber reviews online, so you call several and pick one.
  4. The plumber comes, removes the dead heater and carries it away.
  5. As he works, he asks innocently why you picked that specific model, and you answer without suspecting anything.
  6. He goes back to the warehouse to fetch the new model.
  7. You get a phone call from him, explaining that he’s looking at the model right now, and it doesn’t have the feature you are looking for. But he has a cheaper model with that feature.
  8. You agree for the cheaper model to be installed. Unsurprisingly, it sucks.
  9. You call the maker of the model you initially asked for, and find out it did have that feature.
  10. The plumber does not return your calls.

What probably happened was that the plumber never had the model you were looking for in the first place, so he resorted to bait-and-switch to funnel you into buying the model he did have. Annoying, but clever nonetheless.

As Soon As Possible

Published
by
Victor Nicollet
on November 15, 2010
in Strategy
. 4 Comments

As you might expect from a Start-Up team, we of RunOrg work alone on whatever task we have to do, with fragmentary communication during the day and frequent one-hour meetings to discuss the state of the project. We have a pretty clean road map ahead of us, but there’s a lot of unexpected or unpredictable things going on in the early months of a start-up: new, cheaper service providers on the market let us do a 180° turn on our production strategy ; an administrative process requires immediate submission of even more forms (oh God, let it end!) ; important customers show up and want to discuss our product-to-be ; something important happens in the life of one of the three associates, crippling 33% of the work force for a short while …

After a few weeks, there’s one lesson to be learned : in a start-up, there is no such thing as ASAP.

Suppose I’m your core product developer (not a very bold assumption) and you need me to fill in some administrative form, because you’re in charge of asking for a government grant. There’s a pretty tight deadline, so you tell me to send back that form ASAP. The problem being that almost every single other thing I need to do needs to be done ASAP: set up the production server backups, correct that bug that sends ten thousand password recovery e-mails for every single request, submit some critical information for a conference we’re attending, and give some developer from FoobarCorp technical details about our API. So, there’s a chance that by the end of the week, I’ll be done with your form, unless the server dies at 11 p.m. the day before and I spend the night bringing it back to life.

You know that the administrative form is a 10-minute task that needs to be done by Tuesday evening and that could win or lose us $25,000 but I don’t have any information about it other than its urgency. Why is it so important? I know why the server backups are important because I’ve put them on my todo-list myself and I have an intimate knowledge of what is at stake. I know what it means to receive ten thousand e-mails for a single password request. I just cannot feel the same about an administrative form that you dump on me. In fact, the only reason you said it was to be done ASAP might be that for the last few weeks I’ve been passively filtering out any non-ASAP requests from you. The oral equivalent of writing V!agra instead of Viagra to get your spam e-mail past the spam filters.

It’s nothing personal. It’s just a consequence of three environmental factors:

  • Everyone is overworked, so tasks that are not blatantly important might be postponed forever.
  • We’re equals, so you can’t give me «drop everything and do this now» orders.
  • You didn’t give me enough information to decide how important your request is.

This is a nasty mix. To get over it, there are two simple rules you can follow.

The first rule is the requester’s responsibility: always add a deadline to your requests. If you tell me that you need those forms for Tuesday evening, there will be no ambiguity as to the urgency of the matter — I could have questions about the priority (is this form more important than setting up the server backups?) and it’s up to you to defend the importance (this could win us $25,000) until we agree on whether the action is must-have ($25,000? of course we’ll do it), nice-to-have ($250? I’ll see if I have time…) or ridiculous ($2.5? are you kidding me?). Basically, we’re working together to help me fit this new task in my own mental priority queue.

The second rule is symmetrical: always ask for a deadline. «We need to frobnicate the product as soon as possible» could equally mean «drop everything and frobnicate the product immediately or we’ll be sued to hell and back» or «make a mental note to add frobnication on the Q3 2011 wishlist» and whatever level of urgency you thought you heard is probably not the level the requester intended. You can ask for a deadline explicitly, or you can provide a plausible deadline based on your current load — do whichever you believe fits the situation, both are ultimately equivalent.

Low-Cost Software

Published
by
Victor Nicollet
on November 5, 2010
in Strategy
. 2 Comments

Low-Cost is all about having a dirt cheap base product, such as plane trips, combined with additional fees for almost every single thing you can imagine, such as carry-on luggage, using the restrooms during the flight, or wearing a blonde wig. In the B2C and B2IB (Business-to-Itty-Biz) worlds, the Low-Cost model works when it seems fair. Asking people to pay for what they use is fair in two ways:

  • Share the cost: What I use costs the company money. They went to great lengths to drive the base price down, so it’s only fair that I participate in any additional expenses I cause.
  • Avoid free riders: If I don’t have heavy luggage or use the restrooms, I won’t have to pay for the kerosene spent carrying around other people’s luggage or the water for their restroom usage.

That’s why a Low-Cost model cannot charge for wearing a blonde wig : the wig doesn’t cost the airplane company anything, so there’s no reason they should charge you for it.

This obsession with fairness goes head-to-head against traditional economic theory. In a traditional economic world with purely rational agents, everyone agrees that a printer that prints 100 pages/minute costs more than a printer that prints 20 pages/minute. In our real world, if the 20-pages printer is the exact same hardware as the 100-pages printer with a speed inhibitor tacked on, bloody murder will be screamed. Of course, big businesses don’t care — it’s a game everyone plays, so they’re going to negotiate discounts and settle for the higher price anyway — but if you’re targeting the B2C/B2IB market and they find out you’re into that kind of backstabbing, you’re done for.

If you’re going for a product with multiple prices, or with additional features available for a fee, remember that customers will be looking at the price difference through two very different lenses:

  • Is it worth it? Does that $5 feature provide me with happiness or productivity that exceeds $5?
  • Is it fair? Does that $5 feature cost the provider anything close to $5, or are they just gouging me for the fun (and gross margin) of it?

In the software world, we’re all concentrating on the first question, but we had better keep away from the second one : writing software is dirt open-source cheap.

It’s fair to ask for some money when you’ve spent weeks working on a feature. It stops being fair when you cripple your software to create a cheaper version. And by «cripple» I do not necessarily mean removing or blocking existing features: willingly straying from the optimal design to be able to monetize a feature that would have been an immediate consequence of that optimal design.

Are you providing some of your users with daily backups, instead of guaranteeing that you will never lose their data?

Are you trying to sell data recovery features by selling a product without a revision history tool?

Are you actively fighting third party data conversion tools to force customers into buying your own data extraction plug-in?

Are you limiting the number of anything (users, posts, files…) that the users are doing with your software on their computers?

Know Thyself

Published
by
Victor Nicollet
on November 3, 2010
in Strategy
. 0 Comments

What do you enjoy about this project? What do you utterly hate about it? What’s the long-term motivation that keeps you working on it : fame, money, passion, something else?

What are the day-to-day activities that you love doing? What would you rather not do? Do these match your actual skills, or are you unskilled in what you love and hate what you’re skilled in?

When and where are you the most productive? Are you a regular worker, or do you have infrequent bursts of productivity?

I’m glad you took the time to know all of these tidbits about yourself. It really helps.

Have you told the people you work with? Have they told you what makes them tick? If not, how do you know you’re working with them and not against them?

Prove It for Breakfast

Published
by
Victor Nicollet
on October 18, 2010
in Strategy
. 4 Comments

You’re saving an important document on your computer and all of a sudden, it explodes. Maybe the power went out, maybe the hard drive joined the choir invisible, or maybe the application just bluescreened. Either way, you lost your work for the past five minutes (because you’re hitting Ctrl + S every five minutes, right?)

I have some bad news for you: the computer went down while it was writing on top of the file. What used to be a though-provoking bullet-based presentation has now turned into the software equivalent of a week-old roadkill.

In the software industry, we need higher reliability. You don’t want your bank to say «your payment is safe, unless a random crash occurs in the next five minutes», what you expect is nothing less than «should the apocalypse strike within the next five seconds, your financial transaction will be safe with us» and this causes us technical types to run around in circles looking for better ways to keep your data crash-proof.

So, obviously, every single database vendor out there advertises itself as being completely and irrevocably crash-proof. Usually as part of a huge bulleted list of features. Boring.

Here, CouchDB is the one with the outstanding marketing message. The CouchDB server cannot be shut down by any safe means, the only way to stop it is to cause it to crash (or kill it with a task management tool, or power off the computer). They’re basically implying that their crash protection design is so good, crashing the software is the normal shutdown procedure.

One way to be outstanding is to make difficult things appear effortless — routine events no more remarkable than tying your shoes or drinking water. Where every other software package makes crash recovery sound like an exceptional effort worthy of being included in a feature list, CouchDB shrugs, an unimpressed «I eat crashes for breakfast» look on its face, and goes back to manhandling your data.

So, what do you eat for breakfast?

The Outstanding Economy

Published
by
Victor Nicollet
on September 25, 2010
in Strategy
. 2 Comments

Not so long ago, this world was a world where everyone died within miles of the place they were born.

If you wanted to have dinner, you could eat stale bread at home, or you could eat stale bread at a friend’s home, or you could eat somewhat less stale bread at the village inn or tavern. If you wanted to eat something really outstanding, you would go to a large city and spend what probably amounted to your life’s savings to eat a meal by a top chef. And you would be happy because it had meat in it. And spices. Then, the industrial revolution happened, and its main accomplishment was that it turned masses of people living in the country with access to plain, bland commodities into masses of people living in the cities with access to plain, bland commodities. It also had a very interesting side-effect: it sparked a century-long downward trend in transportation costs.

In 1610, even the richest merchant in Western Europe could not get their hands on a plain mango, because mangoes only grew in the east indies and were months away by ship or caravan. Until the advent of refrigeration, most of the western world knew of mangoes only in their pickled form (this led to the apparition of «to mango» meaning «to pickle» in some parts of the US).

In 2010, advances in refrigeration technology (fruit last longer), air travel (fruit travels faster) and cultivar development (the main cultivar of mangoes, Tommy Atkins, was picked because it provides the longest shelf life) mean that everyone in the western hemisphere can buy a fresh mango for the equivalent of ten minutes of work at the average wage.

On a local scale, inhabitants of major cities have hundreds of restaurants available to them for dinner, conveniently placed within an hour of their home by car or public transportation.

And of course, using the internet, anyone has access to any piece of information available online, and can buy unique hand-crafted pieces from an equally anonymous person on another continent.

Too Many Choices

When you only had access to a handful of options, you could become knowledgeable about each one. You could know how good was the food at every restaurant in town, because there were only five. You could have an opinion on everything, and if you didn’t, you could ask your friends, which is how reputations spread.

As the number of options increased, the reputation system could not follow because there are too many alternatives for any individual or tight social group to handle. This led first to the apparition of reviews (a trusted third party specializes in having an opinion on everything, and shares it with everyone), but the number of product categories has increased (meaning you now need a reviewer reviewer to determine whom you can trust) and they now differ on several variables (weight, battery life, size of the app store, price…).

So, we end up subject to the availability heuristic (people only remember and discuss exceptionally good and exceptionally bad things) and satisficing (people will settle on the first thing that’s not too horrible).

There are two obvious strategies one could follow from here. One is the classic approach to marketing, which you probably experienced on a daily basis for decades: create something that isn’t exceptionally bad (so it isn’t knocked out by the availability heuristic), advertise the hell out of it so that it’s the first thing people try (and settle on it through satisficing), and watch the cash come in. If you can set up a subscription-based system with lock-in, that’s even better. This is a great strategy for conquering a new market, but it’s extremely inefficient at stealing market share on existing markets.

The other strategy relies on creating an outstanding product—something that is so exceptionally good, the availability heuristic will kick in and customers won’t even remember there were any competing products in the first place. While this strategy can be used for conquering new markets, the classic approach is cheaper because it can afford using a cheaper product. On the other hand, the only way to steal market share is to create something that is obviously and undoubtedly better than the alternatives.

So, it’s a matter of outstanding vs. satisficing. What side are you on?

Related Posts

Picture credit: Sergey Prokudin-Gorsky under Creative Commons.

1170 feed subscribers
(readers who polled a feed this week)