SaaS & Minimizing Buyer Risk

Andreessen Horowitz had a podcast recently on software company valuations. All of their podcasts are excellent by the way and definitely worth listening to when you have some time. This one discussed the fact that, traditionally, big enterprise software deals were sold as "perpetual" licenses. This meant that the enterprise would pay big money upfront for software that could be used forever. This was a nice thing for the seller from an accounting point of view. You'd get big bucks on day one that you could use to pay your engineering team and sales force. Your financials would look really good in that period. The software as a service model (SaaS) is much different. With SaaS, the license is sold as a subscription and revenues and costs are spread out over the life of the agreement. At first glance, the SaaS model doesn't make a seller's financials look so good. When the deal is closed the seller has to pay their engineering team and sales force upfront. All that cash is out the door but the revenue is collected and realized over several years. This is why Castlight Health was called the most overpriced IPO of the century when they went public a couple months ago at a valuation of $1.4 billion on only $13 million in revenue (an outrageous 100x revenue multiple). I don't have a strong opinion on the company or their valuation but what many people in the media missed is the fact that most of that multiple was being driven by the company's "deferred revenue" -- or deals that have been closed but not yet realized from a revenue recognition perspective. Deferred revenue is a critical measure of a SaaS company's health.

I say all of this to make a related point. One of the challenges in selling enterprise software is the buyer's concern about risk -- e.g. a buyer might say "what if we make a big investment in your product and we find that it doesn't work for us, do you provide a guarantee?" When you're selling SaaS (as opposed to a perpetual license) it's important to explain to your buyer that you are totally aligned on risk. The entire SaaS model is built around getting renewals. During the initial contract period, the cost of selling the software likely exceeds the revenue collected so it's critical for the seller to get the buyer to renew. The good news for the seller is that over time the costs are amortized and when the client renews the relationship becomes quite profitable. So the entire model is setup to drive customer renewals -- in many cases, most of the risk is actually on the seller. It's worth explaining some of this simple accounting to a buyer when they push back on risk.

Unlike a perpetual license, the buyer and seller's interests are completely aligned: the buyer needs great software and the seller needs a renewal.

4 Things That Big Healthcare IT Companies Must Do To Stay Competitive

The healthcare IT space is possibly the most exciting and dynamic industry in the United States right now. Healthcare is going through a total transformation driven by massive regulatory change, the “consumerization” of healthcare and the important shift from a system that manages sickness to a system that manages health. Underlying all of this change is the software that runs large healthcare organizations -- specifically, the big EMR systems. Given all of the rapid change in healthcare, the EMR industry -- and the dominant players that lead it -- are ripe for disruption. It's not unlikely that there'll be some big names dropping out of the race over the next several years.

With that in mind, here are 4 things I think the large EMR players should do to remain competitive amidst all of this change.

1. Move to the cloud. Healthcare IT is all about big data. And the large EMR companies host loads of it. In the traditional database space, Oracle and SAP waited much too long to move their data to the cloud. And it seems that some of the big EMR companies appear to be waiting too long as well (though, it's possible they could be making this transformation behind the scenes). Regardless, the fact is that health information is going to have to live on the cloud in the long-term. There is no way around this. Patients are going to demand interoperability of data between their primary care doctor and their gastroenterologist and their dermatologist and their dentist. And there is no way that all of those providers are going to be running the same EMR – the space is way too fragmented. I'd argue that not moving to the cloud is a bigger risk for EMR companies now than it was to the large database companies ten years ago. Patient advocates and regulators are simply not going to allow a big EMR vendor to keep their data in house. Larry Ellison said it took 7 years of development to get Oracle on the cloud. EMRs vendors can't continue to put this off.

2. Open up platform APIs (I mean, really open up platform APIs). I've used the BlackBerry versus Apple's iOS example in the past when discussing this topic. Apple opened up its app store early (effectively employing hundreds of thousands of app developers) and as a result made the iPhone 1,000 times more valuable. Meanwhile, BlackBerry dragged their feet and eventually ended up near bankruptcy. There are a number of reasons why the analogy isn't perfect (EMRs aren't consumer products, there are HIPAA restrictions around exposing personal health information, etc.) but EMRs should take a close look at what caused BlackBerry's demise. Part of the reason they dragged their feet on opening up was that their corporate customers were hesitant to allow their employees to download apps. They let their own customers slow down their development. Now some will tell you that the EMRs have created APIs and are adding services on top of their products all the time. This is not true. Even the most open EMRs are tightly policing the products that plug-in to their platform. The first EMR that takes a true "app store" approach will have a massive advantage. There are a ton of well-funded developers building amazing things that these EMRs can tap into if they open up.

3. Focus on usability. I'm not a doctor and I don't work in a doctor's office. But I've seen enough of these systems and I've heard enough complaints from users of them to know that the usability of most EMRs is not up to par with high quality B2B software tools. This is the classic case of B2B software being bad because it can. These companies have high talent sales teams that only need to sell a handful of executives and the rest of the health system is forced to use it and deal with the usability problems. With the emergence of B2E2B (business to employee to business) sales strategies a lot of this is changing. Staff members expect B2B software to work the same way their consumer tools work (Facebook, Gmail, Amazon, etc.).  Granted, due to high switching costs, the big EMRs can get away with poor usability for a while -- it'll be a long time before EMR software is sold the way Yammer is sold but when big contracts come due in a few years, usability will be a massive competitive advantage.

4. Get out of the B2C business. Many big EMRs are rapidly creating direct to consumer products, mostly in the form of a patient portal. This is being driven by 1.) the belief that consumers will continue to be more and more engaged in their care and 2.) the government is requiring it as part of meaningful use; though it’s mostly being driven by the latter, which is a recipe for really weak consumer products. Take a look at the app store ratings of many of the big health IT apps – consumer expectations of what makes a good app are much too high for an enterprise-focused vendor to meet at this point.  To compete in the consumer space you have to be totally focused on the consumer. It has to be an obsession. Take a look at a company like Oscar Health that has built their entire business around consumer experience. This isn’t a criticism of the EMRs, they do lots and lots of things really well. The point is that they should focus on those things and double down on them. Moving to the consumer space is too hard and too competitive and too much of a distraction.  The better approach is to buy or partner with an organization that is built around the consumer.

Facebook's Defensibility Is Gone

Traditionally when people have thought of Facebook and their defensiblity, they've pointed to its ubiquity and the size of its network – at last check they had something like 1.1 billion active users. People reasoned that Facebook would continue to dominate social because it's the one place that has profiles for all of your friends. All other social  networks would be forced to plug-in to the Facebook ecosystem. But as Facebook’s defensive purchase of WhatsApp shows, this is no longer the case. Users are bouncing from social network to social network. Social apps are much, much less sticky than initially thought.

Benedict Evans and others have pointed to the seemingly minor but incredibly impactful fact that any newly launched social app can easily tap into your mobile phone's address book and instantly build out a network equal to -- or better than -- Facebook's.

This wasn't a big a issue when most users accessed Facebook through the desktop site, but now that most users access it through their mobile app, Facebook's unbundling has accelerated.

More and more users are migrating to WhatsApp for messaging, Vimeo for video, Instagram for photos, Foursquare for location sharing, etc. And there are niche players internationally that are focused on badges, stickers and other features valued in those communities.  There are now dozens and dozens of social apps in the app store with more than one-million downloads.

Facebook's strategy of running the social ecosystem seems to be shifting more rapidly than they had planned. Because of the mobile phone's address book, the approach of plugging social apps into Facebook may be losing steam. Instead of just letting them plug-in, the better approach, it seems, might be to buy them.

Attackers & Defenders

A few years ago I had the pleasure of meeting Steve Case -- the original founder of AOL and current CEO of Revolution. He was considering an investment in our company and I was lucky enough to be able to pitch him our business. In the short time that I spent with him I could tell that we were dealing with an extremely savvy investor. He got right to the key issues surrounding our growth and his questions were extremely challenging and relevant.

I came across an interview that he did recently with Adam Bryant from the New York Times. In talking about different types of businesses, he said this:

...I realized the world of business really separates into these two groups. The attackers are the entrepreneurs who are disrupting the status quo, trying to change the world, take the hill, anything is possible, and have nothing to lose in most cases. They’re driven by passion and the idea and intensity. Large organizations — and it’s true of Fortune 500s and it’s also true of governments and other large organizations — are defenders. These guys aren’t trying to pursue the art of the possible, how to maximize opportunity. They actually are trying to minimize the downside, and hedge risk. They’re trying to de-risk situations. Entrepreneurs can’t even think this way. It’s not even a concept they understand.

For the traditional executives running these large companies, of course they want to grow, of course they want to innovate, of course they’d rather have revenue grow faster than slower, but they mostly don’t want to lose what they’ve got. But entrepreneurs are deathly afraid that they won’t be able to change the world, and that somebody else will. Again, these generalizations are a little unfair, but corporate executives are all too often deathly afraid that the business they inherit will be less valuable when they leave than when they started.

This is so true and exactly why no company will last forever. Even the best eventually flame out. The cycle of disrupt >> succeed >> defend is unavoidable and, frankly, perfectly logical. When companies reach a certain level of success, innovation becomes too risky and the smarter, rational move is to protect your turf.

This is why companies like Apple are so impressive and so rare. They somehow continue to attack and innovate despite their immense success.

Internet Marketplaces: Buyer Utility & Seller Reviews

Charles Hudson had a good post this week titled, Marketplaces, Rating Systems, And Leakage. In it, he talks about leakage in online marketplaces. Leakage defined as a user coming to a marketplace to transact and then completing subsequent transactions off of the marketplace.

Once they’ve acquired a new customer through a service, there’s a significant financial incentive for sellers (Open Table restaurant owners, Uber drivers, Task Rabbit workers) to try to get the user to make their second transaction offline – to avoid paying the marketplace a commission.

But these marketplaces aren’t seeing this type of behavior. They’re seeing that subsequent transactions are staying in their marketplace.

The reason for this is twofold:

1. The user values the utility of the service (it’s much easier to book a restaurant reservation on Open Table than it is to call, wait on hold, and find they don't have any tables tonight).
2. As Charles points out, sellers place a high value on reviews from the marketplace. A commenter notes that he once offered to pay for his Task Rabbit project offline and the seller declined. The seller would rather the transaction happen on Task Rabbit so a review gets logged for his work, improving his Task Rabbit reputation. For savvy sellers, a good review on a trusted marketplace is like gold.

Internet marketplaces are always at risk of becoming a lead generation service instead of the central spot where transactions happen. To keep people transacting in the marketplace, it’s important that buyers value the utility of the service and sellers value the reputation gained through post-purchase reviews. Open Table, Uber and Task Rabbit do both of these things well.

Open Conversations

Back in 2005, Union Square Ventures -- the well-known NYC-based venture capital firm -- converted the homepage of their website into a blog. Brad Burnham, one of USV’s partners explained their reasoning at the time.

"We realized that our thesis evolves incrementally as a result of our dialogue with the market, and that the best way to manage that was to accept that we would never get to an answer, so we should just publish the conversation. The best way to do that is with a blog. So here it is."

A few months ago, they took this a step further and turned their website into a conversation, allowing anyone to share links and discuss topics related to the firm and the firm's investments. They also now cross-post their own blog posts and even take pitches from entrepreneurs on their site. Really cool.

In some ways, it’s surprising that an institutional investor would be so open and willing to have a public conversation about their investments and their investment strategy. VCs don’t have hard assets, they don’t have engineering talent, and they don’t have a product. Their entire value is really their investment thesis and their ability to execute on that thesis. So it’s a pretty bold move for them to open up all of that intellectual capital to the world.

But as Brad noted, he believes that opening up the conversation actually puts them at an advantage.

I’d love to see more companies be as open as USV, and to begin having open conversations with their employees, vendors, partners and customers. Personally, I’m constantly having conversations with my colleagues and with the market about the things I’m working on. These conversations help me get better at what I do. Part of the reason I write on this blog is to help me think things through.

What USV has done is scale their conversations and their ability to get better at what they do enormously. Instead of just having conversations with their colleagues that sit across the hall, they're having conversations with (potentially) anyone in the world. That kind of scale has to put them at an advantage over other VCs.

The obvious concern with this approach is that opening up the conversation about your work and what your company does will give away sensitive, proprietary information that would put the company at a disadvantage against the competition.

But I think there are two critical insights here that strongly counter that concern.

1. With very, very few exceptions, companies don’t have some secret and final solution that will drive their success. As Brad notes, most growth and success comes incrementally as a result of perpetual interaction with the market. The thesis is never final, it is always evolving. This is true of nearly every company.
2. Just because you can view and participate in the conversation that a company is having doesn't mean you can recreate what that company is doing. When I write about a new approach I'm taking, by the time someone reads it, internalizes it, and acts on it, I've already moved on and improved on that approach. In addition, my approach is probably wrong for you anyway. You're in a different situation, have different resources, have different connections, have different opportunities and different constraints. It's useful for us to have a conversation, it will help us both. But it doesn't put either of us at risk.

So with very, very few exceptions, I think more companies should begin to open up their internal conversations, challenges and ideas to the public. In the book The Wisdom Of the Crowds, James Surowiecki talks about the fact that across multiple applications (business, military, psychology) large groups of average people are much smarter than any small group of elite thinkers.  I think it's a mistake for companies to think through their challenges in private. A company's likelihood of success is much greater if they open up their challenges to the 6 billion people outside of their walls -- in addition to the small group of individuals inside them.

Put simply, in most cases, the long-term benefits of open conversations are far greater than any potential short-term risk.

Healthcare Reform & Prioritizing The User

I was really interested to read the other day that the president of Comchart Medical Software (an EMR vendor) just announced in a blog post that his product is no longer going to be certified for Meaningful Use. For those readers that don’t work in healthcare and don't know what I'm talking about, Meaningful Use is a really important qualification program happening right now in healthcare.

Some background. As part of healthcare reform, the government wants healthcare providers to use software (as opposed to paper) when providing care. Specifically, they want providers to invest in and use an Electronic Medical Record (EMR) system. The hope is that the use of these EMRs will enable interoperability between providers; improve care quality, safety, and efficiency; engage patients in their care; and improve overall population health.

With that in mind, the government has laid out a five year plan and three stages of Meaningful Use implementation and compliance that EMRs must meet. Just like the name suggests, the government wants providers to use their EMRs “meaningfully." In each stage of the implementation, the usage requirements of electronic healthcare become more and more significant.

The government is pretty serious about this effort. In the short term, they’re providing financial incentives (specifically, higher Medicare reimbursements) to providers that meet Meaningful Use requirements.  In the longer term, those incentives will turn into penalties.

As you can imagine, this change in the law has led to massive technology investments on the part of healthcare providers. They’re all scrambling as fast as possible to implement their EMRs -- and vendors that make software for healthcare have seen their sales skyrocket. On a side note, this is a large part of the reason that you’re seeing more and more independent doctors becoming employed by large hospitals and health systems. They can’t bear the cost of installing an EMR on their own.

But now that EMRs have gotten some traction with providers (Stage 2 goes into effect in 2014), things are starting to get interesting. As providers are further along in their meaningful use certification, they’re finding that they actually use (and need) these products to run their businesses. Like most users, they want the software to be user friendly and to align with what's important to them and their patients.

And of course, the good EMR vendors -- like most good software companies -- are learning, iterating and releasing changes and improvements to delight these providers.

But, wait a second, not so fast. Maybe they're not.

Remember, the priority and goal of the EMR vendors isn’t necessarily to serve their customers (the providers) and, by extension, patients. The priority and goal of the EMR vendors is to help their customers reach a specific list of objectives as laid out by the government 4 years ago. The EMR vendor's goal isn't to make a product that helps providers and patients, their goal is to make a product that complies with a series of strict government mandates and timelines.

Anybody that knows anything about product development, especially software development, knows that the the product you set out to build in the beginning is always wrong. You have to launch and iterate and iterate and iterate to get it right. You can't know in the beginning what is right so you must change and release, change and release.

But given that the government is likely the least agile organization you'll ever find, they can't change their product requirements to meet provider needs. Or at least they can't do it quickly. So it was just a matter of time before Meaningful Use requirements and what's good for providers and patients began to diverge.

And that’s what we’re seeing with Comchart’s decision to halt their product’s Meaningful Use certification. Take a look at this excerpt from their President's blog post:

While the individual people involved in promulgating these EMR mandates (mostly) have the best of intentions, they clearly do not understand what transpires in the exam room, as many of the mandated features confer little or no benefit to either the patient or the healthcare provider.

And this:

As a result of the mountain of mandates, ComChart EMR  and the other small EMR companies will have to choose to implement the mandates or use their resources to add “innovative” features to their EMR. 

So, in short, a software vendor has decided to prioritize its users over government mandates.

Now of course I don’t know enough about the clinical value of Meaningful Use requirements to understand how off base they actually are, but I’m confident that we’re going to see more of this in the months to come. You just have to assume that, despite their good intentions, the government missed the mark with these mandates. And because big government mandates aren’t at all agile – like software development needs to be – you just know that Meaningful Use mandates are getting further and further away from what’s best for providers and patients (they just didn’t know what they didn’t know whenthe requirements were written).

Related to this, I’ve written a quite a bit about how bad enterprise software is when compared to consumer software. For the simple fact that, traditionally, big enterprise software companies could get away with it – they just needed good salespeople that could sell an individual or a small group of individuals on their product and those individuals would force their employees to use the product. Enterprise software companies can survive (and thrive) with a weak product.

But what's happening here is even worse. The government, who’s even further removed from the needs and wants of the end user, is mandating what the software must do with virtually no ability to iterate on it as priorities change and new discoveries are made.

Despite everyone's best intentions, this is a recipe for a terrible product. It is so far removed from what's good for the end user.

In the long term, I think we'll see more and more of these small, user-driven EMRs abandon Meaningful Use certification. And this will result in two types of products, or two different somewhat radical product directions: one that meets Meaningful Use requirements but is painful to use, and one that doesn't meet those requirements but is a delight to use.

In my view, in the long, long term, as Meaningful Use requirements are scaled back or phased out completely, the lighter-weight, user-driven EMRs will be the vendors that win. They'll have such a strong and inherent product advantage over those that were forced to rely on the government to design and dictate their product roadmap.

That said, I recognize the challenges for EMR companies that go their own course. This is going to create a major client management problem in the short term. And I recognize that it's likely going to take years for these vendors to win back clients. But physicians are no different than any other consumer; they want great products that are beautiful and intuitive and easy and seamless. Eventually they'll demand it. And eventually they'll get it.

As I wrote about in my post about the business to employee to business sales strategy, this is the same course that companies like Yammer and DropBox and Xobni are taking. They've prioritized the user and built a sales and product strategy that relies on user satisfaction and product quality to succeed. These companies are winning because they're bypassing the bureaucracy and misplaced priorities that lead to large, lumpy sales and mediocre product offerings.

They've prioritized the user. And the EMR vendors that do the same will be the ones that win.

This is going to be fascinating to watch.

Amazon Drones & Offline Spend

[youtube=http://youtu.be/98BIu9dpwHU]

There’s been a lot of hype around Jeff Bezos’ announcement on 60 Minutes that Amazon is building drones (see above) that will ship packages from warehouses to consumers’ homes within 30 minutes.  While this may seem crazy, it’s really a very well calculated strategy aimed at getting access to one of Amazon’s only untapped markets: offline consumer spend.

Offline commerce remains a trillion-dollar industry that most online sellers have yet to tap. It’s still true that most consumer disposable income is spent locally -- restaurants, bars, coffee shops, gas stations, salons, malls, etc.

Amazon and other online retailers have struggled to tap into this marketplace. But lots of them are trying hard...

Consider a company like Trunk Club that sends consumers a trunk full of clothes every once in a while, allowing the consumer to keep what they like and send back what they don’t. With free shipping.

Or Dollar Shave Club that sends razors and other toiletries each month at rock-bottom prices. Check out their brilliant promotional video that went viral a while back.

Or Warby Parker that ships you three different styles of glasses at no cost so you can try them on. And you ship them back for free. For every pair that's sold they donate a pair to a person in a underdeveloped country.

Or Groupon that sells access to local salons or gym memberships or karate lessons at steep discounts.

All of these companies have built strategies that are attempts to tap into local, offline spend -- and the Amazon drone is no different. By shortening delivery times they're hoping to keep people out of stores and in their homes buying goods online. This is a fun trend to keep an eye on.

Your Product Doesn't Sell Itself

Blake Masters posted his notes from a class that Peter Thiel taught at Stanford a while back. The class was focused on distribution for startups and the notes are awesome, awesome, awesome. I've been meaning to write about them for a while. They're a must read for start-up sales & marketing professionals. The whole thing is great but the piece I want to talk about today is where he points out that the idea that a product can sell itself is a complete myth.

Given all of the focus on product lately – particularly in the consumer internet space – you might be surprised to hear this from Peter Thiel. But he’s spot-on. Here are the key paragraphs:

People say it all the time: this product is so good that it sells itself. This is almost never true. These people are lying, either to themselves, to others, or both. But why do they lie? The straightforward answer is that they are trying to convince other people that their product is, in fact, good.  They do not want to say “our product is so bad that it takes the best salespeople in the world to convince people to buy it.” So one should always evaluate such claims carefully. Is it an empirical fact that product x sells itself? Or is that a sales pitch?

The truth is that selling things—whether we’re talking about advertising, mass marketing, cookie-cutter sales, or complex sales—is not a purely rational enterprise. It is not just about perfect information sharing, where you simply provide prospective customers with all the relevant information that they then use to make dispassionate, rational decisions. There is much stranger stuff at work here.

To emphasize his point, he uses this framework:

Consider the quadrants:

Product sells itself, no sales effort. Does not exist. Product needs selling, no sales effort. You have no revenue. Product needs selling, strong sales piece. This is a sales-driven company. Product sells itself, strong sales piece. This is ideal.

If you believe that your product is so great that it can sell itself you’re either delusional or your aspirations aren't nearly high enough – and it’s great to see a hugely successful, product-driven investor make that point.

15 Reasons Why Healthcare Has A Business Model Problem

I’ve worked in start-ups my entire career. Part of the fun of working in a start-up is creating, testing and iterating your business model. In order to succeed you need a business model that both makes sense intuitively and works financially. If it doesn't make sense, it doesn't work. When I entered healthcare not too long ago (an industry that has existed in one form or another for thousands of years) and started to dive into the detail and understand the motivations of the different players in the market (providers, patients, payers, government, etc.), I quickly realized that something wasn't right.

The business model of healthcare didn't seem intuitive and clear to me like it did in other industries. So I sat down with a pen and paper and tried to scribble out the model. I was sure I could figure it out. I drew a box for each of the stakeholders and labeled them with their different motivations and then drew lines between them indicating the different money flows.

Trying to make the model make sense for all of the stakeholders literally made my head hurt.

I felt a little bit better last week when I heard Jonathan Bush, the CEO of AthenaHealth, state clearly in an interview at Duke University's business school that the problem with healthcare has never been a clinical problem or a scientific problem, the problem with healthcare has always been a business model problem.

This is so true. There are so many inherent challenges with healthcare that make the business model extremely difficult to figure out and get right. I came up with a list of 15 reasons why healthcare has a business model problem.  I'm sure there are many, many more, so feel free to add them in the comments.

1. The vast majority of care is paid for directly by a third-party, so patients have little incentive to shop around, effectively averting supply and demand and a reliable market equilibrium
2. If patients did have an incentive to shop around, on a practical level, it would be nearly impossible to do (without a clinical background, it’s extremely difficult to measure quality or evaluate what care is or isn’t necessary)
3. Because of the way care is paid for, patients plan procedures months in advance (pregnancies, knee surgeries, etc.) but finance them like they're unexpected emergencies
4. An enormous amount of patients have their care subsidized or paid for by the government (~21% of federal tax dollars go healthcare)
5. Many patients are irrationally loyal/unloyal to their doctors because they're not impacted by costs or able to measure quality
6. Sick patients jack up the price for healthier patients (payers pass on costs)
7. There are for-profit, non-profit and government-funded providers competing to provide the exact same services to the exact same patients
8. An endless number of moral issues surrounding care make consensus on a business model extremely difficult
9. It's an extremely local industry and difficult to scale across geographies
10. Financially speaking, providers and patients are at odds with one another (providers do well when patients are sick, at least in a fee-for-service market)
11. It’s an extremely fragmented  industry (patients see different providers for different services – a patient could have more than 10 different doctors that don’t have to talk to one another); consolidation is desperately needed to bring down cost
12. Death is an outcome of poor or insufficient care, so it’s impossible to cap prices -- healthcare prices are extremely elastic
13. People don't want care unless it's urgent but when they want it they want the best
14. By law, providers cannot turn some patients away and in many cases must provide care for free
15. Unlike most insurance providers, health insurance providers generally pay for everything, not just catastrophes (this would be like your car insurance company paying for your oil changes and your gas)

B2B Pricing & Malcolm Gladwell's New Book

Last Sunday night on a late plane ride back to New York I finished reading Malcolm Gladwell’s new book, David and Goliath. It’s not his best work, but it's still thought provoking and worth the time.

The cliff notes version of the book is that it turns out that David wasn't such an underdog after all. He was actually at a huge advantage in his battle with Goliath because he was able to change the rules of the game. Lots of underdogs win by changing the rules of the game so that they become the favorite. To prove his point, Gladwell discusses the civil rights movement, World War II, middle school basketball and many other examples. It's an interesting concept and a pretty good read.

Anyway, as part of his argument, he spends a lot of time talking about the difference between linear curves and U-curves. He argues that there are some things that correlate and may seem like they should sit on a linear curve, but in reality they sit on a U-curve.

Perhaps the most interesting and classic example that Gladwell uses is the correlation between class size and test scores. You might think that the class size/test score graph would be linear -- as class size increases (the X-axis), test scores go down (the Y-axis). That's the conventional thinking.

But it turns out that's not true. The reality is that when class size gets really small, studies have shown that test scores actually begin to decrease again. When there isn’t a range of opinions and when the teacher is too focused on one student, the quality of education goes down. Students benefit from the energy and discussion that comes with having lots of other students in the class. So the correlation between test scores and class size really looks more like a U-curve. Test scores are optimal when the class is around 25 students. Too big is bad and too small is bad.

This is also true of crime and punishment. Many believe that as the severity of punishment increases, the amount of crime goes down. But studies have found that there’s a U-curve effect here as well. As punishment becomes more and more severe for smaller crimes, citizens begin to believe that the system isn't fair. That it’s us against them. And they commit more crimes as a result. So you obviously don't want punishment to be too lenient, but you also don't want it to be too strict. Like a lot of things, eliminating crime isn't simple. It isn't linear.

The corollary back to the story of David and Goliath is that you can't improve education simply by adding more teachers and reducing class size and you can't reduce crime by simply making punishment more severe. With complex problems, brute force doesn't always win.

With the U-curve in mind, I've been thinking a bit about price increases and how they impact client relationships. Traditionally, companies like to have modest annual price increases of, say, 2% to 5% per year. Companies like these small increases because they reduce the risk of putting a large strain on their client relationships or losing clients all together. And generally, due to inflation and other factors, clients usually find these increases acceptable.

But as you begin to raise the annual price increase more and more you'll generally see clients resist more and more; 2% is acceptable,15% is not. Like the examples above, on the surface, it seems that the correlation between price increases and client resistance should be graphed on a linear curve. As the price goes up (the Y-axis), so does resistance from clients (the X-axis). Keep your price increases low and your clients will be happy.

But I don't think this is true either. As price increases, client resistance certainly increases, but only to a certain degree. At some point, say 10% and up, in order for the increase to make sense there has to be a substantial and fundamental product change. A 20% increase isn't caused by inflation, increased labor costs, or a greedy salesperson. The increase is happening because there's a substantial change in the product and the product's value. As a result, clients should be much more accepting of this change and client resistance will start to go back down -- creating a U-curve. If the increase is sold and communicated properly, clients will begin to recognize that they're buying a much different product, or at least a much more valuable one. They'll understand that the rules have changed. In fact, I'd argue that at some point most clients would welcome these large but less frequent increases more than they would the small, annoying increases that don't show a corresponding increase in product value.

Increasing price is a requirement of any sustainable long term B2B relationship. And like most things, it's not simple. It's not linear. So when considering a strategic approach to pricing, companies should consider the U-curve and not rely on small, inflation-based annual increases that barely impact revenue. They should cancel those increases. That energy is better spent finding ways to change the rules -- to completely rethink product function, value and positioning. To make a splash. To delight rather than satisfy.

In short, the lesson of the U-curve is actually pretty simple. Don't avoid client resistance by keeping price increases small, modest and frequent; get clients behind them by making them big, bold and rare.

Internet Marketplaces Should Be Seller Agnostic

When you're running an internet marketplace (Etsy, OpenTable, eBay, Uber, KickStarter, Yelp, etc.), it's very tempting to give sellers the opportunity to buy premium placement. This could be things like homepage placement, better placement in search results or enhanced profile pages. Selling this stuff is certainly a very logical way to monetize your user base.

But the problem with this approach is that every time you give priority to a seller that pays you more money, you've muddied your value proposition and taken an equal amount of value away from your user base.

A good marketplace is one that creates an easy, beautiful, seamless and open buying experience that enables rating and recommendation systems so that users can decide the best sellers and the worst sellers. Selling premium placement effectively circumvents your users' preferences putting you in a race to the bottom.

Sure, by charging for premium placement, in the short term, you'll generate some cash. And you can use that cash to do some marketing so that you can generate more users. But over time, if you want to continue to grow, you'll need more and more money and more and more marketing. That will force you to create more and more premium placement options (subsequently devaluing your marketplace).

The better approach is to prioritize the user and compete in the race to the top. Give your users the best possible marketplace experience and let them decide which sellers should win and which sellers should lose. Those great experiences will result in buyers telling their friends to use your service and that will result in more people going through the experience and more people telling their friends to use your service and on and on.

That's the way to scale. And you can't do it favoring one seller over another. Give sellers a great experience too, but don't prioritize them. Prioritize the user.