"The more I find out, the less I know."

Ten Business Lessons

Friday - February 22, 2008 10:08 AM

Rob at Businesspundit is signing off today, and he ends with a list of ten lessons he's learned about business after ten years of entrepreneurship.
I don't normally link to Businesspundit (mainly because he's loaded up the site with zillions of craptacular advertisements, widgets, and other blog-flotsam), but this article is excellent and worth reading for anyone thinking about starting a business.

There's an eleventh lesson not mentioned, however: every entrepreneur will have to learn these lessons for himself. We're a stubborn lot by nature, and there's no substitute for getting burned a few times.

Posted at 10:08 AM | Permalink |

The Service Quality Tracker

Thursday - October 12, 2006 02:32 PM

I haven't blogged about my company in a while, though not for lack of activity. Its just that I get plenty of chances to blog about work on my work blog, so why do it here?
But today we started something pretty cool which I'm eager to share. We've launched a project, called the Service Quality Tracker, to measure the quality of customer service at different companies and post the results to our web page in real time.

Anyone can be part of this: all you need to do is use our alternate toll-free phone number instead of the company's number. When you call, we'll forward the call to the company, and call you back after you hang up for a survey.

The first industry we're tracking is computer technical support: Apple, Dell, Gateway, and HP. There's no data on our web page yet because we don't have any data, but as soon as we have enough you'll be able to see how the companies stack up against each other for technical support. In real-time, with the freshest possible data.

So, please, pass these phone numbers around. Print them out and keep them handy. The more the merrier!

To Call Dial Instead of
Apple Technical Support 800-894-3218 800-275-2273
Dell Technical Support 800-894-3639 800-624-9896
Gateway Technical Support 800-894-3552 800-846-2301
HP Technical Support 800-871-4439 800-474-6836

Posted at 02:32 PM | Permalink |

Advice for Amanda Congdon

Saturday - July 08, 2006 09:32 AM

In my previous article, I posted advice for Andrew Baron, owner of a controlling stake in the empty shell that is Rocketboom.

This article has my advice for Amanda Congdon, the former host of Rocketboom who now finds herself without the platform and audience she built over the past 18 months or so.
At first blush, Amanda appears to be in a better position than Rocketboom. She's got a big fan base (including Yours Truly), brains, looks, and talent. She's proven that she can write, act, and do comedy.

What she doesn't have is offline stardom.

Amanda wants to pursue an acting career in Los Angeles, and I wish her all the luck in the world. She won't need as much luck as most aspiring young actresses, but the odds are still very much against her.

The cold hard fact is that for every Julia Roberts, who churns out hit movie after hit movie, there are hundreds of actresses who have one great role and never do anything of consequence again. And for every one-hit-wonder, there are thousands who tried and failed to get that one great role. Success in Hollywood is still very much a long shot.

There may even be some places where Rocketboom counts against Amanda. Just as TV stars have a notoriously hard time transitioning to movies, the Internet has yet to be a springboard to any kind of lasting fame for anyone in the offline world. A director may decide that he would rather have a fresh face (there are a lot of them out there) who doesn't bring the baggage of fame among geeks.

As a woman, Amanda has one other thing working against her: time. In probably five years, she'll be "too old" by Hollywood standards.

So what to do?

My advice to Amanda is simple, and has the benefit of being something she's already said she'll do.

Amanda should stick with what works.

Resume doing a daily video blog (she shouldn't have much trouble finding someone to back it financially), a medium in which she has proven talent and a big fan base. I would guess that at least a quarter of her old audience will find her within a week, and maybe more.

Use the video blog as a way to grow as an actress, keep her name in front of the public, and shoot for something bigger.

But don't count on that Something Bigger coming along to pay the bills. It might happen, but the odds are still stacked against her.

It is probably a lot easier to figure out a way to monetize her existing stardom and turn video blogging into a career than it is to turn a popular video blog into an acting career.

Look at it this way: 300,000 people watch Amanda online every day. That's about 299,700,000 people in the United States alone who don't watch Amanda. Her stardom counts as a mass audience in the small pond of the online world, but it's nothing compared to the draw of an honest-to-goodness movie star.

While video blogging, Amanda can also make time for an acting career if that happens to work out. If she plays her cards right, she won't have to give up the blog when the casting director phones with the Big Call. So she has the chance to have it all: undisputed stardom online, without giving up the chance to pursue a "real" acting career.

So my advice to Amanda is this:

1) Don't give up video blogging. You're good at it, and you have already built an audience.

2) Find a way to make video blogging pay the rent, because the odds are still against you in Hollywood.

3) But don't give up your dream.

Posted at 09:32 AM | Permalink |

Advice for Rocketboom

Saturday - July 08, 2006 08:56 AM

Perhaps my way of dealing with tragedy is to strategize.

I'm posting a pair of articles with my advice for the former Rocketboomers, one with advice for Andrew Baron, and one with advice for Amanda Congdon.
First up is Andrew Baron, the owner of a controlling stake in what used to be the most popular video weblog on the Internet, now a name, a website, and a bunch of archives (my advice for Amanda is here).

Surprisingly, Andrew may actually be in the better position here (I'll explain why in the next article with advice for Amanda), but he'll have to act fast and act smart to capitalize on it.

Andrew's goal at this point is to recover from the loss of Amanda and turn Rocketboom into a profitable, sustainable enterprise.

The first thing to understand is that Rocketboom has been built on a lie, and you need to undo the lie. The lie is that Rocketboom cost only $25 per episode to produce.

It may have cost $25 out of pocket to produce, but it also took multiple talented people the equivalent of a full-time job to create. If Rocketboom was just an interesting hobby you could discount that, but in a profit-making enterprise you can't ignore the people costs, even if you're not paying them at the beginning.

Don't feel bad about this. Lots of entrepreneurs make this mistake. The problem is that it lulls you into a false sense of the economics of your company, and you underestimate how much revenue you need to bring in to be profitable.

If you were to pay everyone on Rocketboom something like market rates, plus compensate the correspondents for their work appropriately, I bet you'd find that Rocketboom actually costs something like $500 to $1,000 per episode. Most of this money is being invested by the people in the form of time.

It is important to understand this now, since (unlike when Amanda started) Rocketboom now has a demonstrated ability to generate at least some revenue. It isn't just an interesting experiment anymore, and whoever you hire to replace Amanda will know this and demand to be compensated appropriately. Worse, because Amanda owns 49% of the company, you can't give her replacement any equity and still maintain a controlling position.

(As an aside: About that stock Amanda holds: I'd let her keep it, for several reasons. First, this whole affair is ugly enough as it is, and you probably won't be able to get the stock back without a big public fight. That will just distract you, cost you tons of money, and further stain your personal reputation. Second, Amanda is still a very visible and popular personality, and it is to your benefit that she have a stake in making you succeed. She can still help Rocketboom fly, and you want that. Third, if you force Amanda to give up her stock, even assuming you can, you will have a hard time using Rocketboom stock to compensate talent in the future, since nobody will trust you to let them keep it.)

Okay, now that we understand what Rocketboom really costs, let's talk about Amanda's replacement.

Your goal is to turn Rocketboom into something bigger than the talent. Ideally, you want it to be a platform from which careers are launched. That will give you the upper hand in finding the replacement for Amanda's replacement: nothing works better than "I can make you a star!"

The one thing you don't want is a clone of Amanda. If you hire another Quirky, Perky Blonde, then you send the message that you need someone Just Like Amanda in order for Rocketboom to have an audience. That's not true, since there are lots of interesting, talented people out there who can build Rocketboom. But if people think that you need another Amanda, then they conclude that Amanda is the reason for the show's success, not the format, the platform, or the Rocketboom brand.

You want to show that Rocketboom can succeed without Amanda, and that it can make a star out of whoever sits behind the desk. So go for a strong contrast to Amanda, there's lots of talent out there.

(Unfortunately, early rumors are that the new Rocketboom host will be Joanne Colan, another Quirky, Perky Blonde. I think this choice is a big mistake. Near term you may keep more audience with an Amanda clone, but longer term it turns Rocketboom into the Quirky Perky Blonde Show, rather than a platform for great talent to grow.)

Then you need to start working on finding the next Rocketboom host, and the next one after that. Start building a bench, as they say. Have some backup hosts for when your main host is on vacation, sick, or quits. Changing faces for a week every now and then will let your audience discover new people, and make Rocketboom less about the star and more about, well, Rocketboom.

Finally, let the talent explore their own directions. Resist the temptation to meddle. Give the host 3-6 minutes each day to explore. It can be hard when you have a different vision for the show than the talent, but if you hired the right people, then the talent is usually right (that's why they're called "talent"). One of the great things about Rocketboom is that it is small enough to make mistakes and learn and grow.

So, in summary, my advice to Andrew Baron is:

1) Figure out what your real costs are, even if you're not paying them, and build the business around that. $25/day is a lie.

2) Let Amanda keep her stock. She'll be a good ally in the future, and Rocketboom can't afford another messy fight right now.

3) Hire someone who is not Amanda. If you go for the Quirky Perky Blonde, then the show never becomes bigger than the host.

4) Find a handful of backup hosts, and give your main host a break every now and then.

5) As long as the results are good, keep your hands off. Let the hosts do their own thing.

Posted at 08:56 AM | Permalink |

The Killer Question

Friday - April 21, 2006 09:47 AM

There was a time, beginning about 15 years ago and ending about 5 years ago, when an entrepreneur's hopes for funding from a VC could be dashed with a single question:

"Could Microsoft incorporate this into Windows or Office?"
Microsoft had such a dominant position in the operating system that the presumption was that it could own any market it wanted, and that if a software startup was successful then Microsoft would want that market for itself. No venture capitalist wanted to invest in a company which Microsoft might squash almost as an afterthought.

Today, the killer question for web-based startups is "Can Google offer that for free?"

Google has such a dominant position in the search market, and has been spreading to other online services (mail, maps, blogging, etc.) so aggressively that the presumption is it can own any service it wants. There are exceptions, of course: Flickr remains a more popular photo sharing site than Picasa, and YouTube is winning big share in video.

But long-term, how can a startup with a few million for development compete against a billion-dollar behemoth growing 80% per year and generating almost 50% operating margins. The amount of cash Google can throw at a new market is nothing short of staggering, and the company has a history of executing well.

So any venture investor today has to presume that if a new online service is modestly successful, and Google can offer it for free, then Google will attempt to dominate the market. That gives the startup three options: sell to Google, sell to someone stupider and poorer than Google, or get crushed.

Posted at 09:47 AM | Permalink |

The Death of Wal-Mart?

Thursday - February 09, 2006 01:52 PM

I don't think the imminent demise of Wal-Mart is near. But I do think that right now the company is at the very pinnacle of its reach and power. When business historians write the history of Wal-Mart, they'll write that the company never again enjoyed the heights of market power that it saw between 2000 and 2010.
I say that because I read an article today (no link, sorry) that Wal-Mart is going to start renovating some of its stores to appeal to a more upscale crowd and compete better against companies like Target.

Not, mind you, that Wal-Mart has anything to fear from Target, since Wal-Mart is many times bigger. But Target has managed to attract a more affluent and desirable demographic, and Wal-Mart apparently feels it needs a piece of the action.

But it is almost impossible for a company that has built its entire business on being the lowest price to move upscale. The brand will always have a taint. If you don't believe me, try saying "Luxury Yugo" without laughing. What's more, Wal-Mart has made the shopping experience in many of its stores so very unpleasant that many shoppers (myself included) won't even consider going in the door. Fancy decor doesn't help if your prospective customer never sees it.

And the simple fact that Wal-Mart wants to move upscale (despite the challenges, which should be obvious to the company) means that the company knows that it can't sustain its growth in its current niche. Competing only on price is probably just about tapped out for Wal-Mart.

So what's next? Probably decades of very modest growth, and Wal-Mart will no longer be a Wall Street darling, just another big safe company for widows and orphans to invest in. Some other company (maybe even Target, or possibly Amazon.com) will define the next era of big retail. Or they might screw up somehow and go the way of K-Mart.

But in five years I don't think we'll be hearing much about how Wal-Mart is ruining America.

Posted at 01:52 PM | Permalink |

Business Book Review

Tuesday - January 31, 2006 12:01 PM

This critique of "Good to Great" sounds like it can be applied to almost every popular business book ever written.
And that's the reason I generally avoid the genre. It seems that any time someone tries to distill strategic management wisdom into a form which can be applied to companies across the board, they wind up with vague generalities that don't provide much actionable advice, and that are of questionable accuracy.

At some level, I'm surprised that anyone even thinks there is such a thing as universal rules for running a business. Why should we expect that the right way to run Northwest airlines (a bankrupt service company locked into ruinous union contracts and with a reputation for poor service) would be the same as the right way to run GE (a conglomerate which can't decide if it's a manufacturer or a financial services company) or the same as the right way to run Google (a media/advertising masquerading as a hypergrowth technology company).

Every company has its own unique circumstances and challenges, and so any universal advice will be either so vague as to be useless ("Don't Do Dumb Things") or so full of exceptions that it won't be universal anymore ("Focus on your core business. Or core businesses. Or businesses which have synergies with your core businesses.").

I think it would be much more useful to compile negative advice: a list of things companies shouldn't do, and which will often lead to ruin. For example, don't be creative in your accounting. Don't break the law. Don't fire your best people.

But even that advice will come with caveats and exceptions ("Don't break the law? Does that mean that nobody can get a parking ticket while on company business?"), so I'm not sure how universal it would be.

Posted at 12:01 PM | Permalink |

The Job of the CEO

Tuesday - November 08, 2005 09:14 PM

Running a 4-person startup generally means one thing:
I get stuck with all the unpleasant tasks nobody else wants.

Sometime last night, the thermostat in our small refrigerator stopped working, and the fridge went into full-cold mode. Two cans of soda exploded, jars of olives, mayo, salad dressing, and relish were frozen solid, and there were bits of broken glass from a bottle of iced tea which had frozen and shattered.

Somebody had to clean it up.

The options were:

1) Take a developer off of developing our new service (the one which we have a major customer practically begging to buy),

2) Take our salesperson off of doing e-mail follow-ups after a long weekend, or

3) Me.

They say that the CEO is the most important person in the company. It's true. I'm the most qualified person to clean up the skanky mess in the fridge.

Posted at 09:14 PM | Permalink |

New Service

Friday - September 16, 2005 06:15 PM

My company announced a new service this week called Express Feedback. I won't go into great detail (you can follow the link if you're interested, but unless you're in the customer service business you probably won't be), except to say that we've devised a dramatically improved way to do call recordings and satisfaction surveys in a call center.
We would have preferred to have a few more weeks to put some more gloss on our marketing, but there's a major trade show next week, and we wanted to get the announcement out before the event. In any event, everybody works faster under deadline.

The result is that this has been a tiring, but exciting week.

We've been cooking this up since July, and expect to go into pilot phase by the end of October (a very preliminary date, to be sure), with revenue at the beginning of 2006. For the past six weeks or so, we've been talking about it with a handful (less than a dozen) companies, and going into the announcement we had three pilot sites already signed up. We consider that very positive early feedback, especially since we expect that a significant fraction of pilot sites will go on to subscribe to the service.

So when we made the announcement on Wednesday, we already knew that Express Feedback was probably a winner (another reason, by the way, why I am dead-set against the Cult of the NDA: early feedback from prospective customers is probably the most important marketing tool in existence).

Within 24 hours of our announcement, we had another two pilot sites lined up, and a couple more companies expressing a lot of interest. These are not small clients, either: of the five pilots we have so far, three are major household names--the kind of customer VC-backed startups love to list in IPO prospectuses. Both of the tentative pilot companies are also Fortune-500 sized companies. Any one of those big companies could give us enough business to be nicely profitable.

Then, just as a topper, one of our smaller pilot customers told us that they plan to ask for a budget next year to subscribe to Express Feedback, at twice the level we had pegged as their maximum. Then they dropped a hint that they might go up to three times the already-increased level. This before we've even set final prices (though we gave this company some idea of the range we expect).

All I can say is, um, wow.

After nearly four years of struggling, this is what it feels like when things start coming together.

But now the pressure really begins, since we have to finish developing Express Feedback, build infrastructure, and solve various unanticipated problems, all before the momentum and interest runs out. I think my developers are starting to really feel the heat now. This project matters.

As hard as the work ahead is, though, I'd much rather have this problem than the opposite. This is fun.

Posted at 06:15 PM | Permalink |

eBay Jumps the Shark

Monday - September 12, 2005 08:06 AM

The big business headline today is that eBay is paying $2.6 billion to buy peer-to-peer Voice-over-IP software company Skype.
(Pronunciation guide: "Skype" rhymes with "hype")

What's missing from this deal is any semblance of business logic or sanity.

I will grant that, as an outside observer I don't have any special insight into Skype or the thinking of eBay management. But the optics of this deal are really really bad.

Starting with the price. $2.6 billion harkens back to the frothiest days of the dot-com bubble, when a whiff of an idea plus a couple engineers could command a billion dollars. Skype has been very close-mouthed about its financial performance, and estimates range from $15 million in revenue per year all the way to $100 million.

$100 million in annual revenue is pretty respectable for a startup, and would be enough to go public on, but even at that high end of the range eBay is paying 26x revenue. Not 26x earnings, but 26x revenue. At the low end of the range for revenue estimates, eBay is paying over 100x revenue.

(Though by the end of the day today we should have a better handle on Skype's financials, since eBay will have to disclose at least some information.)

The real head-scratcher is the business case. As in, I can't think of one. As near as I can determine, eBay thinks it can push the Skype service out to its millions of members, somehow turning the auction company into a next-generation phone service.

It should be noted that this is very different than the PayPal acquisition. PayPal was already used by many (perhaps most) PayPal members before eBay acquired it, and was an essential part of many transactions. eBay was essentially buying a piece of its own value chain that it didn't already own.

With Skype, this simply is not the case.

Just to make things worse, if Skype is ultimately successful, it will look more and more like a phone company. Already much of its revenue comes from providing access to traditional phone networks, and from traditional value-added phone services like voicemail.

The operational realities of a phone company are much different than those of an auction service or payment service. For example, if eBay management thinks that governments won't try to regulate Skype as they currently regulate phone companies, they're dreaming. Regulation will quickly become the driving reality of much of that business.

Finally, both eBay and PayPal are web-based services, while Skype is largely desktop software. As someone who runs a company doing the former, whose wife works in a company that makes the latter, those are two very different kinds of businesses. It is an open question as to whether the experience running a web-based service helps you at all when trying to publish desktop software.

In fact, the only similarity I can find between Skype and eBay and PayPal is that all three companies grow through a "network effect," which can be very powerful when it works in the company's favor. Unfortunately, there isn't much about the network effect that makes it a real synergy between two companies.

I think the reality is that eBay and PayPal are saturating their respective markets, and eBay management wants to use some of its currency to buy something--anything!--which can help keep it growing and maintain the stock price.

But that's not a good reason to spend $2.6 billion.

UPDATE: The actual purchase price is $1.3 billion in cash, $1.3 billion in stock, and up to an additional $1.5 billion in earn-outs if Skype meets performance goals over the next three years. Skype's revenue forecast for 2005 is revealed to be $60 million. So the price is at least 40x 2005 revenue, and possibly as much as 68x 2005 revenue. Or, as someone else calculated, it comes out to something like $25 for every person who has ever downloaded the Skype software, whether they use it or not. Ouch!

Posted at 08:06 AM | Permalink |

Shameless Plug

Wednesday - August 24, 2005 03:49 PM

Just to prove that there is no marketing fad I won't jump on, my company now has a blog.
We call it The Customer Service Survey. It is a group blog (written by company employees) about providing good customer service.

Posted at 03:49 PM | Permalink |

CEO, Psychopath

Monday - August 22, 2005 08:58 AM

Reading this article in Fast Company, I suddenly realized that as a CEO, I am very likely a psychopath.
No, really.

Apparently you can't be a CEO without being a psychopath, at least if you believe the authors of the article. Many corporations are apparently also clinically psychopathic (whatever that might mean when applied to a legal construct).

Can we do a quick reality check?

To begin with, "psychopath" is a clinical diagnosis, and making the diagnosis is supposed to require a two-hour interview with a trained clinician (this checklist is really a joke). I seriously doubt that anyone involved in the article did much more than rely on media reports of bad behavior for most of their "diagnoses."

But even assuming that the diagnoses are correct, something like two percent of the population are psychopaths (which is basically defined as someone who utterly lacks empathy for others but fakes it well). If there are 2,000 companies in the U.S. of reasonable size--probably a low estimate depending on your definition of "reasonable size"--then there are 2,000 CEOs of good-sized companies around, and you would expect to find around 40 clinical psychopaths among that group even if they mirrored the population as a whole.

Coming up with a handful of examples is hardly proof that CEOs are more likely to be psychopaths.

But even stranger is the idea of "corporate psychopathy," the notion that corporations can be clinical psychopaths. Corporations are legal constructs, groups of people working towards some common goal, so it is hard to see how a corporation can exhibit empathy (or any other emotional state) to begin with, much less some deranged variant.

The idea that corporations are inherently psychopathic is even odder. That's like saying V.W. Beetles are inherently friendly, or kindergarden classes are inherently happy. They may exhibit outward signs that we would associate with friendliness or happiness, but the car itself, or the class itself does not have an emotional state (even if kindergardeners are, on average, happier than the population as a whole).

But for the sake of argument, let's accept the idea that a corporation could possibly exhibit something like what we would call the emotional disorder of psychopathy. You can certainly cite examples of companies which behave badly, but I can cite counterexamples of companies which behave very well.

For example: Target and General Mills, both of which give 5% of pretax profits to charity (and actively encourage volunteer work by employees). Medtronic, which places more emphasis on saving lives than making money (though the company does both very well). Costco, which pays its employees above-market wages despite pressure from competition and Wall Street. And I could go on. These are all big companies which have done well by doing good.

Companies are not inherently good or evil: they are collections of people working toward some common goal, and their actions are determined by the collective actions of all the people in the company.

So if you're going to make a claim like "Corporations are inherently psychopathic," or the more nuanced "Corporations tend to exhibit behavior typical of psychopathy in people," or even the clinically testable "CEOs are more likely to be psychopaths than the population at large," you should back that claim with actual hard research, not anecdotes and hand-waving. Otherwise you're just blowing hot air.

Meanwhile, I need to go work on my evil laugh.

Posted at 08:58 AM | Permalink |

Northwest Airlines Strike

Saturday - August 20, 2005 10:09 AM

Mechanics for Northwest Airlines, our very own Twin Cities airline monopoly, have gone on strike. The official word from Northwest is that they plan to fly a full schedule using outsourced maintenance and replacement workers.
Barring a sympathy strike from the flight attendants or pilots, they can probably do it.


Because I've reached the conclusion--based on what I've read in media reports, so don't think I'm particularly insightful or knowledgeable--that Northwest's real intent all along has been to break the mechanics union. The way to break a union is to demand such extreme concessions that the union goes on strike, and then replace all the striking workers.

After some period of time, either the union will cry uncle (in which case the airline wins), or the strike will have gone on so long that Northwest will be able to argue that the replacements are now its actual mechanics and ask for a vote to decertify the union (I'm not up on the finer points of labor law, so please correct me if this is wrong).

I can hardly blame Northwest in this case. The airline, like most of the older carriers, is battling for its life against low-fare newcomers....and in the case of Northwest, not doing very well. Northwest has not been my airline of choice for a couple years now, thanks to high prices and often customer-hostile service. Of course, in Minneapolis, sometimes you have no other choice, but I probably fly on other airlines three times as often as on Northwest, despite the fact that Northwest owns 80% of the departures from the Twin Cities.

So I believe Northwest when they claim they can keep flying without its mechanics union. I think they've been planning to do exactly that for a long time.

Posted at 10:09 AM | Permalink |

Stealth Mode Startup

Friday - May 27, 2005 09:21 AM

If you've read my Cult of the NDA article (which, by the way, remains one of the most-read articles on this blog), then you know that I think good luck and good execution are far more important to success in a startup than a Really Big Idea. Ideas don't have a lot of value by themselves, they tend not to be unique, and by staying secret a startup gives up the chance to get valuable feedback from prospective customers and partners.
Let me add another reason to that list: it isn't always that hard to figure out what a stealth-mode startup is really doing. Let me give you a real-world example....

I'm at a tradeshow in Dallas this week, and we were visited by some very bright people from a stealth-mode startup. During the two days of the show, one of their programmers sat in on my speech about building better customer service automation; I had an hour-long chat with the CEO and another of the founders; and their PR manager stopped by our booth for a 15-minute briefing about what we do.

During this whole time we didn't learn anything about what they do. Nada. Bupkis. This was, in fact, one of the most secretive startups I've ever encountered (and I've known a few). They told us that the name on the business cards was a code name, and that the logo was deliberately misleading. They didn't even offer to sign an NDA.

Normally this kind of secrecy just makes me smirk a little and think "probably another online pet food company." But in this case, all the attention they were paying me made me want to know more--especially since they were asking were the kinds of questions you'd ask if you wanted to invest in or acquire a company. And I hate being at an information disadvantage.

So I sat down and did some research.

And here are the facts I turned up:

* They think their business is a huge opportunity, probably a billion-dollar plus market which nobody else is currently addressing (this last point could easily be naivete, which is common enough among entrepreneurs).

* The basic business model is a business-to-business service bureau (this from their website, and is consistent with what they told us in discussions).

* They want to build a big brand fast after they come out of stealth mode.

* One of the cofounders has been active on a mailing list for Asterisk, an open-source Linux-based PBX.

* They were at a tradeshow for speech recognition technology and IP-based call centers.

* The same team started another company which successfully sold services to the Fortune-1000.

* They are actively hiring Perl programmers. In fact, they are quite emphatic that they specifically want people who can write Perl and don't much care for any other skills.

This last point is intriguing since, as my VP of Development puts it, "Perl is the duct tape of programming languages." You use it for quick and dirty patches, stuff you don't need to be scalable or maintainable. In fact, Perl seems like a downright odd choice for a company planning to build a big service bureau, where scalability and maintainability are critical. This means that either (a) the person making the key architecture decisions is a moron (unlikely, but always a possibility, and there are some rabid Perl fans out there); or (b) there is some other consideration forcing them into Perl.

Coincidentally, the programming hooks into the Asterisk PBX are Perl-based.

So it seems very likely that they're building a service bureau based somehow around a PBX (using Asterisk as the core). Combine that with their presence at a call center show, and the fact that Perl seems well-suited for doing lots of little bits of customization around Asterisk--which is necessary for building a call center--and it seems most likely that they're building a company to provide IP-based call center infrastructure.

In other words, they'll use Voice-over-IP technology to provide all the technical infrastructure of a call center (automated self service, agent queueing, etc.) on a hosted basis. You just give the agents a PC with a web browser and a headset and this company will do everything else. Our intelligence assessment puts it at a 60% probability that this is their business plan (or something very similar). This, by the way, is a really cool idea and one which could easily be a billion-dollar business if it is properly executed. It isn't unique, though. I would guess that there are 3-4 venture-backed startups working on similar ideas.

Looking a little more broadly, we think there's an 85% chance that their business plan has something to do with providing advanced business telephony services through voice-over-IP. If not the call center business, then something like an IP Centrex (though even a casual Google search reveals dozens of companies already offering IP Centrex).

Now that I've told you what we think this company is doing, am I going to tell you who it is?

In a word, no. I respect this company's desire to remain in stealth mode, even if I don't think there's much point.

But if I'm right, you will have read it here first.

Posted at 09:21 AM | Permalink |


Friday - April 29, 2005 02:59 PM

I'm a total upgrade whore. I admit that.
Well, that and the fact that I'm going to be at the Mall of America anyway this evening means that I may just swing by and pick up a sparking-new copy of Tiger.

But while browsing Apple's website, I noticed that they're going to be giving away scratch-off cards this evening as part of their "world premier." Curious, I downloaded the schedule of prizes. Prizes range from an iTunes download to a free 15" laptop.

But the really strange thing is that every store has different odds of each prize! Not just because the number of cards at each store varies. For example, the odds of winning an iTunes download vary from 1:3.6 (at the North Michigan Ave. store in Chicago) to 1:1 (every card is a winner...at the smaller stores). The odds of winning a coupon for a 10% discount (not by the way usable for a Tiger upgrade) vary from 1:1.4 to 1:6.1

What I'm wondering is: what is Apple doing here? Are they constructing some grand marketing experiment to see what the optimal odds are for generating revenue? Or do they have a very expensive consultant telling them that a 10% discount gets them more benefit in Chicago, whereas a free iTunes download is optimal in Minneapolis?

In any event, each store seems to be giving away exactly one free laptop, so you want to go to one of the smaller stores to have the best chance at the really big prize.

Posted at 02:59 PM | Permalink |

1. Aim Gun At Foot. 2. Pull Trigger

Tuesday - March 01, 2005 09:35 AM

The release of some old TV shows on DVD has been held up by problems getting licenses for all the music. A number of TV shows used pop songs as background (or even foreground) music, and now the rightsholders of those tunes want additional royalties for DVD release.
None of this should be surprising. When the shows were originally created there was no market for consumer sales of TV shows (except maybe for made-for-TV movies or things like National Geographic specials). As a result, nobody thought about this licensing issue until suddenly DVDs created the market.

In other words, what had been assumed to be a nearly worthless asset now has new and unanticipated value.

Given the state of affairs, one can hardly blame the companies which own the rights to the music for wanting to extract maximum value for their copyrights. But I can't help think that they're shooting themselves in the foot by making it difficult to negotiate royalties. Here's why:

1) Some money is better than none, and no money is what the music rightsholders will get if the producers of the TV shows either can't release them on DVD or decide to substitute different music.

2) Now that this issue is out in the open, TV producers are going to be very careful to make sure they own all the rights to the music they use in a show. That means that either familiar tunes will disappear from TV shows, or music rightsholders will be forced to give up a lot more rights for TV use than they used to. It would be stupid for a TV producer to use an Elvis song in a 2005 production without being absolutely certain that the song won't slow down DVD release.

3) The net result is that some TV shows will highlight new music from unfamiliar artists (where the rights are cheap); and other shows will insist on getting broadcast plus DVD rights for what they used to pay for broadcast rights alone. Either way, the rightsholders of the most popular music get less money than they would have, all because they insisted on making music royalties a stumbling block for DVD release.

ASIDE: I'm becoming more and more convinced that while most companies are very good at tactical thinking (i.e. getting what they want right now), very few companies do a good job of strategic thinking (i.e. anticipating other companies' reactions and working for the best long-term outcome).

A classic example is the common practice (especially in technology) of cutting deals at the end of a quarter to make a sales goal. Good tactics, since you meet the short-term target, but lousy strategy since it trains customers to wait for the end of a quarter and makes sales forecasting very difficult.

Good strategy sometimes requires giving up a short-term goal for a long-term benefit. Companies (public companies in particular) seem to have a very hard time doing this.

Posted at 09:35 AM | Permalink |

Economics of Book Publishing

Tuesday - February 22, 2005 11:02 AM

I'm at the SpeechTEK West conference/trade show in San Francisco this week, and I was chatting with a consultant I know about our respective books (his forthcoming, mine published). An industry analyst of our mutual acquaintance interjected the too-obvious-to-ask question:

"What are the economics of writing a book? How do you make any money off it?"

After a few seconds of blank stares, he clarified: "You spend six months or more of your life writing and editing a book, and get to keep fifty cents or a dollar from every copy sold. How do you get paid for all your time?"
It's a good question, especially considering that for a technical book selling 2,500 copies is considered pretty good.

In our case, where we're our own publisher, we had to front about $4,000 in design and production costs. Our gross margin is anywhere from $6/copy (for books sold through Amazon.com) to $19/copy (for books we sell at full price). We would have to sell about 700 copies through Amazon to cover our costs. If we sell that many I'll be a very happy camper: we expect to take a financial loss on our book (not even counting the time we spent writing and editing it).

So why bother?

The answer for us is the same as for the consultant (who may ultimately get a couple thousand dollars for half a year's work): Marketing.

Of course we all want to have a best seller on the business list, but that's unrealistic. There are thousands of business and technical titles published every year, and only a handful ever make it to the top ten.

But for me and my consultant friend, the payoff comes (we hope) through other business and enhanced credibility in the market:

* Simply having a book published is a mark of competence. It says that we take this seriously, that we're smart people, and that we know what we're talking about.

* Every copy sold equals someone willing to pay money to read our ideas.

* Every copy has my name and our company logo on it. This links us with the core ideas of the book (i.e. good customer service) in the minds of the readers. Some of these readers may be in a position to buy services like ours in the future.

When the stars align, the marketing power of a book can be amazingly powerful. There's one consultant in the call center industry (who shall remain nameless, but industry insiders will know who I mean) who has parlayed academic credentials and some research twenty years ago into a sizable business just co-authoring other people's books. His business is basically this: companies pay the consultant a fee to co-author and publish the books they write (he self-publishes everything). In other words, he gets paid to put his name on a books in order to lend credibility to a company's marketing message (he also does other sorts of endorsements; the model is sort of like the pro athlete who gets paid to drink a particular brand of beer in public). Of course, every book he publishes with his name on it also enhances his own credibility as an industry expert--after all, look at the stack of books he authored! He must be a smart guy!

Not a bad line of work, if you don't mind selling your soul.

(P.S. Our book is currently outselling all but two of this consultant's titles on Amazon.com. But that's not saying much.)

Posted at 11:02 AM | Permalink |

Transparent Horizontal Integration equals Bad Customer Experience

Monday - February 21, 2005 05:41 PM

"Vertical integration doesn't work" is a relatively recent management insight. Recent enough that there's still a lot of people debating the question (especially those clinging to failed vertical integration strategies). But the most successful industries of the past few decades--personal computing, networking, and anything Internet-based--are all aggressively horizontal. I expect the debate to die down within a few years as it becomes "obvious" that with only a few exceptions, horizontal is the way to go.

Mostly this is good. Horizontal integration allows maximal innovation and competition in every piece of a product or service. This drives down costs, advances technology, and generally makes the world a happier place (except for middle management, which is more likely to get fired as part of a lean-and-mean strategy).

But there are some significant risks to horizontal integration, and the biggest one is also the most painful one: if not handled carefully, horizontal integration leads to a really bad customer experience. It is no coincidence that some of the most infamous examples of bad products and services today--personal computing and mobile phones, for example--and also industries which are either the most aggressively horizontal, or industries which are moving from being vertical to horizontal.
Vertical vs. Horizontal Integration
Imagine a diagram showing how raw materials get transformed into products or service consumers buy. At the top are the rawest of the raw materials: energy, ores, silica (sand), oil (for both energy and making plastics), and so forth.

As you move down the diagram, the raw materials are transformed into more and more refined products: sand is refined into silicon wafers; which are combined with numerous chemicals to become Pentium processors; which are incorporated into personal computers along with memory chips, hard drives, DVD recorders, and Microsoft software to make the product you buy at Circuit City.

On this diagram, no one company makes everything. Intel, for example, buys silicon and chemicals, and sells Pentium chips and motherboards. Dell buys plastic (or metal) cases, LCD displays, chips, motherboards, disk drives, and Microsoft Windows to make a Dell PC. If you were to draw a box around the products which a given company makes, vertically integrated companies would have tall boxes (spanning many steps in the process from raw material to consumer product), and horizontally integrated companies would have short boxes (with only one or two steps).

The classic example of a vertically integrated company is the old AT&T (before it broke up). It was not too much of an exaggeration to say that AT&T bought sand as its raw material and sold dial-tone as its service. In its heyday, AT&T manufactured its own chips, built its switches and telephones, strung the wires, and managed the network operations. AT&T is not the only example, though: way back when Ford Motors built a factory in St. Paul, one reason that location was chosen was because it sat on top of a high-quality sandstone deposit. The sandstone was important because Ford mined the sand, refined it, and made it into windshield glass, all on site.

Why Vertical Integration Doesn't Work
It has been a long time since Ford made its own windshields from sand mined at the St. Paul factory, and the reason why is the same reason vertical integration generally doesn't work as a business strategy: it is cheaper to buy glass made elsewhere by factories which specialize in making windshield glass.

Vertical integration, as a rule, short-circuits the competitive process which leads to better and cheaper products. Ford's in-house windshield factory had no incentive to become more efficient or produce better windshields, because it would be guaranteed to "sell" its production to Ford's car factory. No doubt when the factory was built it was state-of-the-art, but third-party glass factories had a continuous incentive to keep getting better. At some point, it became cheaper to buy the glass elsewhere than to make it in-house, even allowing for the profit margin a third party would demand.

Horizontal integration, by contrast, gives each company up and down the product cycle the chance to buy its raw materials competitively (at least in theory), which gives everyone from the ore miner to the retailer an incentive to improve their products and reduce the costs. If someone comes up with a better way to carry out any step in the production chain, that improvement can quickly benefit the entire industry.

There are some circumstances where vertical integration makes sense: for example, if no outside suppliers exist for some part or raw material, if the transaction cost of working with an outside supplier are too high (excessive transportation costs are a common reason this might happen), or if the shortcomings of a supplier negatively impact the end product (i.e. a company's step in the manufacturing chain is transparent--I'm getting to that). Monopolies--like the old AT&T and the current Microsoft--also often try to expand vertically, since they don't have room to expand within the confines of their primary market.

Horizontal Opacity
The key to making horizontal integration work well is that each step in the chain from raw material to consumer purchase has to be as opaque as possible. By "opaque" I mean that the customers of Company X should not need to know or care who the suppliers of Company X are.

Chip manufacturing is an excellent example of an opaque process. The buyers of memory chips or Pentium-class processors don't need to know or care where the fabrication plant purchased its silicon wafers or etching chemicals. All the buyer needs to know is what's the price? and does it perform up to specifications? The chip manufacturer is free to change suppliers, innovate, and otherwise improve.

Opacity also means that there's no finger-pointing. If a computer manufacturer gets a bad batch of memory chips, the chip manufacturer doesn't tell the computer manufacturer to talk to the company that supplied the silicon ingots. Rather, the chip manufacturer takes responsibility (after all, it should have caught the defects) and remedies the problems. Similarly, when you buy a car, it doesn't come with separate warranties from separate companies for each component (body, timing belt, upholstery, heater) even if the car manufacturer bought those components from other companies.

Transparent Horizontal Integration
The opposite extreme is where problems happen. If opacity is good in horizontally integrated companies, transparency is bad. Transparency means that the customers of Company X know and care who Company X's suppliers are. In the extreme, Company X's products aren't really its own products, they are a hybrid of Company X and Company Y.

Mobile phones are a great example of this. Are you buying a Nokia mobile phone with Verizon service? Or Verizon wireless service with a Nokia mobile phone? Since the customer (often) cares about both the service provider and the phone manufacturer, neither company is completely free to change its product or service (since that might require changing or severing the relationship), and neither company is completely responsible when things don't work properly.

Another great (and wonderfully convoluted) example is the personal computer industry. You buy a Dell PC with Microsoft software and an Intel processor. Which of these three companies is responsible when the computer doesn't work? (answer: they'll each point fingers at the others) Is Dell free to substitute Apple or Linux software if it works better and is cheaper? (answer: no) Is Microsoft free to come out with a new version of Windows which works better but doesn't work on Dells? (answer: no) Which company has the ability to improve the overall customer experience? (answer: none of them, at least not individually) [Aside: strictly speaking, the choice of an Intel vs. a competitive CPU doesn't make much difference in the overall customer experience, so the processor choice is opaque. But customers do care about it, and it does impact the market dynamics.]

The end result is a mess. No one company is responsible for the complete product, and as a result, nobody has complete freedom to innovate or improve. Consumers are never sure which company is at fault for the state of affairs (good or bad) so they have a hard time making informed buying decisions (classic example: I have a Samsung mobile phone with SprintPCS service. Lately I've been having a hard time placing or receiving calls in areas which used to work well like my office. I don't know if this is because the SprintPCS service has gotten worse, or because the phone is getting old. The net result is that when my contract expires, I plan to get a new phone with a new carrier, and will not use either SprintPCS or Samsung).

Getting Out of the Transparency Trap
So far, everything I've said can be boiled down to: "hybrid products like mobile phones and PCs will always suck because no company is completely responsible for the customer experience." Maybe I should have just written that and skipped the first thousand words of this essay....

The problem is that this state of affairs is a trap for the companies involved because no company wants to give up the relationship with the end customer. Intel loves the fact that (some) customers care that its processors are at the core of most Windows PCs. That gives Intel a competitive advantage over other processor manufacturers, even though it reduces Intel's need to be the fastest or cheapest processor company. Intel spent a lot of money on branding to get that market position.

Similarly, Microsoft will do anything in its power to keep the operating system from becoming generic, even though compatible (if obscure and somewhat rough) alternatives exist. Microsoft makes a ton of money off the fact that 95% of PC buyers won't buy a PC that doesn't come loaded with the latest version of Windows.

And of course Dell--which is the natural candidate to own the customer relationship, since it sells the computer to the consumer--might want to find alternatives to Intel and Microsoft, but its ability to do so is limited by what the customers want.

Even though the inevitable result of this state of affairs is that the customer experience sucks, nobody has the ability to change things.

How to change things?

The most obvious way is for a company to vertically integrate just enough to make its product opaque to the customer. In other words, seize responsibility for everything which impacts the customer experience. This is what Apple has done in the personal computer market: when you buy a Mac, everything about the Mac experience out of the box is controlled by Apple, both hardware and software. This gives Apple the ability to integrate the operating system and basic software with its own hardware in a way Microsoft and Dell can only dream about. And you can't argue with the result: current versions of Mac hardware and software are light-years ahead of anything in the Wintel world in terms of the quality of the overall customer experience (the fact that some earlier versions of Mac hardware and software sucked also proves that merely controlling the customer experience isn't enough to make it good). The downside is that Apple's market is limited to the 5% or so of buyers who don't care if their computer doesn't run Windows on an Intel processor.

Another way out of the trap is to disconnect the parts of the hybrid product or service, and make them separate buying decisions. In Europe, for example, GSM mobile phones aren't tied to a particular carrier's network they way mobile phones are in the U.S. The customer makes two decisions instead of one: buying a phone and buying service, as opposed to buying a phone+service package. Each company can concentrate on making its product or service the best possible, and consumers can switch either the phone or the service when something better comes along. I suspect that it is no coincidence that Europe has long led the U.S. in mobile phone usage.

Industries changing from vertical integration to horizontal integration is probably one of the best things to happen in the business world for many decades. The benefits in terms of efficiency and innovation are immense.

But this has also led to some industries where the customer experience is spectacularly bad, usually when no one company controls the entire end product. The result is that no company is responsible or has the freedom to improve.

Unfortunately, when these situations arise it is usually the case that none of the companies wants to give up its relationship with the consumer for the sake of a better overall product.

Posted at 05:41 PM | Permalink |

Not Yet Published and Already In Its Second Printing!

Sunday - January 09, 2005 09:46 PM

Pre-sales of my company's book have been astonishing. Recall that we're planning to print 1,000 copies, and hope to do two things: get a lot of marketing bang for the buck, and cover our expenses. Make a profit (on the book, at least) isn't really a priority.
Nevertheless, we haven't sent it to the printer yet, but we've already sold over half the copies we plan to print.

How the heck did that happen?

Amazingly enough, one of our prospective customers (not even an actual customer) decided that they like the book so much, they want to hand it out at trade shows.

Yes, you read that correctly. One of our prospective customers is going to pay us to give away our marketing material. Whoa!

We sold them 500 copies (at a steep discount, naturally, but still above our cost of printing). Even better, 500 copies of our book (with our logo plastered all over the cover) will be in the hands of 500 industry decision makers, at no cost to us.

The next question is: do we increase the size of the print run from 1,000 copies to 1,500 or 2,000 based on the early demand? Or do we wait for it to sell out and do a second printing?

There's something very nice about saying our book is in its second printing....

Posted at 09:46 PM | Permalink |

1.9% Sold Out!

Thursday - December 16, 2004 02:57 PM

Writing a book is a lot of work. Writing a short book is even more work, since every page has to be tight. I'm in the process of putting the finishing touches on a short book my company will publish in January. Called Gourmet Customer Service, it is about using a rigorous and scientific approach to improving customer service in call centers. Today we started taking preorders through our website.
We started this project six months ago, with the idea that it could serve as a vehicle for marketing, customer education, and building a reputation. You'll notice that "making a profit" isn't on the list, though we would like to sell enough copies (about 300) to cover our expenses.

We sent out an E-mail announcement to our internal E-mail list of 1,000 names about three hours ago. That's a psychologically important step: until you actually start taking orders, there's always the chance to back out, to change your mind. Now we can't do that anymore. For better or for worse, we're committed to publishing a book in January.

When the orders start coming in, there's a moment of truth. It's gratifying that someone actually wants to read our baby. But when we look at the names of people preordering--clients, prospects, respected people in the industry, plus my wife's boss--there's a sinking feeling that this better be good. We can no longer afford to publish a book just for the sake of publishing a book. Someone's actually going to read it!

We're planning on printing 1,000 copies. After three hours, we've got preorders for 19. That's 1.9% sold out, and better than we thought we'd get on this initial mailing. Now we just have to get it off to the printer on time...

Posted at 02:57 PM | Permalink |

Powered By iBlog, Comments By HaloScan
RSS Feed