How do you build service to scale?

I read @codybrown‘s excellent analysis of network growth and scaling with regards to MySpace, Facebook and Twitter this morning and got to wondering if the concepts he’s talking about can be applied to broader concepts in Customer Service. Brown presents the idea that services like MySpace have failed because they couldn’t channel their explosive growth into a system that worked not only for the core adopters but also for mainstream users. Later-arriving, “fast-follower” services with more focused missions are stealing the MySpace user (Last.fm for music, Facebook for social networking, etc.) because you don’t have to be an insider to figure out how to find the service useful. What does this mean for service concepts in general?

The recent purchase of Zappos by Amazon is a good example of a company with fanatical customer focus acquiring another company with fanatical customer focus. Both Zappos and Amazon have built service to scale — and, I believe, will be able to avoid the problem that @codybrown references — because they focus on three things: Customer Service above all else, Mass Customization, and Back-end services that lower the cost of transactions.

Both Zappos and Amazon practice fanatical devotion to Customers. Zappo’s motto, stated directly in its logo, is “Powered by Service”. About.zappos.com states that “Customer Service is everything.” Amazon has similar focus, calling itself “Earth’s Most Customer-Centric Company.” There are many customer testimonials that might provide additional detail here — the point is that both of these companies have made it possible for customers to tell other customers whether a product or service is good or bad. By empowering the customer and by pledging to fix it when the customer’s expectations aren’t met, both Zappos and Amazon have created a situation where their best customers will shout the company’s name from the rooftops and their worst customers will have a strong process for recourse. Customer Service, and policies that reinforce that customer service, make it possible for both of these companies to build the rest of their brand around that service.

To deliver that service, both Amazon and Zappos practice what I call “mass customization” rather than making an individual web site, store, or product offering for one customer or set of customers. This may sound contradictory, because each customer does get an individual order, and that customer may have an individual concern that sounds different than another customers. But if you take the perspective of looking at groups of customers who buy shoes (in the Zappos world) and then you segment them so that they all buy shoes in roughly the same way (even if front-end design templates and brand segmentation make it seem different), and you fulfill these orders and return them in roughly the same way, then you’ve built the system to have a relatively constant cost per order, even though your product offerings may be myriad in their breadth.
Build once, deliver many different flavors with a common back-end system, and you’ll be able to satisfy the Customer Service needs of all of those different customers.

To deliver this vision of build once, deliver many flavors, both of these companies had to build back-end logistics, supply-chain, and fulfillment services that lowered the overall cost of doing transactions. For Amazon, this meant building software systems to manage inventory, present and build stores, make and fulfill orders, and then building a physical distribution system to implement this vision. For Zappos, the implementation was slightly different, but the effect was the same: build a system that as it scales provides learning curve improvements, opportunities to apply LEAN and greater efficiencies and profitability overall per order. Only now as their services are growing and growing do these investments look particularly shrewd, especially when new product lines (Clothes.com, e.g.) are introduced for the combined company to market. Amazon is leaving Zappos alone to operate as an autonomous unit. Good move — I think it will pay off for both companies.

What does this mean for Twitter, Facebook, MySpace, and other new product entries? Build your service to planet scale — at least in your head. If you don’t know how your service will function when it’s really really big, then you will miss the opportunity as MySpace did to shape the service and prevent fast-followers from cherry-picking the best parts of your model. Make your offering a defensible position — as both Zappos and Amazon have done — but focusing on a service offering your present and future competitors can’t easily duplicate. And finally, a paradoxical idea: you need to try many ideas and fail fast, but remain focused on the small goal that’s easy to explain to your most fanatical and dedicated customers, as well as the mainstream customers you don’t have yet. Keep it simple, please the customer, mass-customize by configuring your offering rather than building net new code for each segment, and build your systems and your ideas to scale.