Sunday, April 26, 2009

Product development leverage

Leverage has once again become a dirty word in the world of finance, and rightly so. But I want to talk about a different kind of leverage, the kind that you can get in product development. It's a force that allows startups to build products at parity with much larger companies - cheaper and much faster. It's a key lean startup concept.

The idea of leverage is simple: for every ounce of effort your product development team puts into your product, find ways to magnify that effort by getting many other people to invest along with you.

Leverage was one of the big ideas we talked about early on at IMVU. We knew that we were only a small team, but we had big aspirations. We wanted to create a 3D product that could provide superior options for self-expression to millions of people. Other companies accomplish this by hiring a large staff of content production experts - and get a very good result. We didn't think we'd able to compete with that. Plus, we saw some of the intrinsic limitations of supporting such a large staff: slower cycle times, higher cost basis, and - most importantly - the ability to serve only a limited number of customer segments. For segments that were considered out of the mainstream or somewhat obscure, the ROI just isn't there for these established companies to serve them.

So we tried to craft a strategy that would give us the product development leverage we needed to serve all customers. We combined three tactics: extensive use of free software, an open platform for user-generated content, and leveraged distribution channels. Each of these tactics was effective, and I'll return to them in detail in a moment. The net result was powerful: within six months of starting the company, we were able to get a basic version of our product into open beta. Although customers didn't flock to this offering at first, we had enough of a developer program active to start recruiting early adopters to start creating 3D objects for sale in our catalog. That engine of creativity has led to a catalog of something like 2 million virtual goods authored by a hundred thousand developers. At no time did IMVU ever employ more than three full-time 3D artists. Most importantly, there is almost no niche or trend that is unserved by this community. Want emo shoes (172,190 available), goth earrings (152,996 available) or anime-themed furniture (55,240 available)? Yeah, we've got that.

Leverage is work, though. It has to be found and managed. For more on the specific trade-offs involved with IMVU's virtual good strategy, see Three decisions to make on virtual goods. In that same spirit, here are some suggestions for tactics you can use to increase the leverage of your product development efforts:
  1. Free and open source software (and even hardware). When you participate in an open community like these you take advantage of tremendous amounts of effort. Even as "just a user" you make the community better by adding momentum. Even better, if you engage with the community in a mutual relationship, you can increase your leverage further. For example, IMVU early on decided on using an open source library for our 3D file formats, skeletal animation, and scene graph. As a result, we were able to get started more quickly, avoid writing a ton of art path tools from scratch, and even hire from within that community. It's an amazing thing when you can hire an employee who knows more about your code base than you do, and this turned out to be a big source of advantage. Over the years, we've made many contributions back to this community; if its formats become standard, the company benefits further. (Of course, there are ethical reasons to prefer free software to proprietary software, too - but they don't bear on this particular discussion so I am omitting them intentionally)

  2. User-generated content. Our original mission statement for IMVU was "to fulfill the promise of online socializing and creativity." We hoped that user-creators would be part of our model from day one. Part of this was a values statement. We had been in a previous company whose pursuit of centralized control had proved damaging in many ways. But part was a recognition that we could gain substantial competitive advantage by leveraging a community of like-minded visionary customers to serve a wider (and more mainstream) audience than we could alone. Making UGC work requires good tools, open standards, and proper incentive design. Personally, the framework I've found most helpful is MTV's "create, share, validate" feedback loop. By focusing on giving creators all three, we were able to reap the rewards of their shared efforts. In the end, I believe they co-created our product with us.

  3. Leveraged distribution channels. It's now possible to gain massive distribution for almost any product without asking anyone for permission or signing a complex contract. This is what Google AdWords, Facebook Platform, the iPhone App Store, and Salesforce AppExchange all have in common. If you have the opportunity to use these channels to reach customers, you can iterate much faster and gain traction before more established competitors can move to check your growth. Of course, all of these mechanism have their own attendant risks, two in particular: 1) that the platform provider will itself decide to compete with you or just limit your growth (as Microsoft has a long history of doing, and as Facebook and Google have occasionally dabbled in), 2) that the ease of distribution empowers new competitors to chase you more effectively. Still, these risks are thoroughly mitigated if you can iterate faster than either set of competitors - and, as a startup, you shouldn't have any excuse for allowing that to happen. (For a specific application of this idea, see How to get distribution advantage on the iPhone.)

  4. Open API's and data-oriented architecture (aka "web 2.0"). The much-promised era of component reuse in software is finally upon us. Some of that is enabled by open source, but a lot more is enabled by simple services that allow apps to be composited in record time and without having to ask for permission. Many complex apps can now be prototyped as a simple mashup in order to prove market viability - and this is true beyond just software apps. For example, I recently created a customer validation exercise around the Lean Startup Workshop. It allowed me to assess the market demand for that offline product before I had the final product baked. Doing the market test required this blog, a SurveyMonkey account and a PayPal account - and nothing else. The response has been nothing short of amazing (thank you all so much!). I'll post more about the specifics of what I learned from this exercise in a future post, but for now I want to focus on how much learning it made possible for such a small amount of effort. And yet, even though it wasn't very costly overall, all of the players involved (including me, Google, PayPal and SurveyMonkey) were able to create and capture real value. Multiply that by the large numbers of similar "long tail" creators like me, and you can see how much leverage this ecosystem is creating.

  5. On-demand utility pricing for services (aka "cloud computing"). This is really a specific instance of many of the above trends synthesized together. But it's had such an impact that I think it's worth itemizing on its own. Naturally, everyone is using tools like Amazon's AWS and Google's AppEngine to prototype their startup's first version, thus lowering both time to market and capital equipment costs. But I'd like to call out special attention to services like Amazon's new FPS payment platform, which allows people to create new web services that are charged on a cost-plus basis. In effect, this allows you to create a new AWS service, profit by it, and distribute it alongside Amazon's other first-party services. I believe this is going to unlock a huge wave of innovative services available with utility pricing.
This is not meant to be an exhaustive list of leveraged product development opportunities. Instead, my goal is to give enough context to make this final point. The power of leverage is so high, it's a benefit that needs to be traded-off against explicitly in order to optimize total time through the fundamental startup feedback loop.

For example, I have wasted a lot of time in my career trying to argue for the superiority of open source solutions over comparable proprietary solutions. These arguments have ranged from the ethical, to performance comparisons, to feature-by-feature breakdowns. But I now realize that none of that was very useful. Instead, the real compelling reason to switch from proprietary platforms and vendors is the advantage you can gain in leverage, even if you give up other serious benefits. When proprietary vendors focus too much on value capture, their products become expensive, inflexible to change, and require too much permission to adapt to new contexts. All of which can slow down startups in just the places where they need to speed up. Thus, the right argument to make in evaluating proprietary alternatives is simply this: will this vendor speed us up or slow us down? If the latter, almost no benefit in terms of price, price/performance, product support, or lifetime total cost of ownership is worth it.

Once startups and vendors really understand this dynamic, we can all get past these legacy arguments and start focusing on building partnerships that truly make it easier to create companies that matter. In the meantime, I think we have a good explanation for what all those PR dollars are being spent trying to obscure. Don't fall for it - it's a trap.


  1. Great post!

    I'd love for you to elaborate more on MTV's "create, share, validate" feedback loop . . . hadn't heard of that before.

  2. Very useful post. A true follower of Wikinomics concepts.