In the discussion of eBusiness / eCommerce / Web 2.0 companies you will frequently hear the word ‘friction’ being used, and it’s always something to be avoided. Friction is anything that will slow you down as you scale up your business model and increase revenues, from the movement of physical bits across the country, to a dependency on another firm in the middle of a transaction, to that most friction-full of items, pesky human staff. If one could only find the eBusiness model that would scale almost infinitely without the need to worry oneself with these mundane factors, especially increasing human staff, now that would be a winner. Some VCs use the term to a fault.
Certainly the growth of the web has overturned many industries, added new efficiencies and disintermediated others that no longer add value. Yet as many of the web names we know grow larger and become a bigger part of our lives, I find myself interested less in whether they are still true to their mission statement, and more interested in whether they have underestimated the real operational requirements to supporting large customer bases and millions of transactions.
Let’s take a look at a few examples:
- Google has increasingly sought ways to monetize their offerings beyond their lucrative search advertising revenue stream, including reaching into the mobile space and competing on office software products, areas where they will deal with business and retail customers that expect service. But let’s start with the original AdWords platform…is it frictionless at this point? Take a look at the help forum, where you will find thousands of posts by individuals (paying customers) trying to get assistance on the application. Many are pointed toward the web-based help tree, which was likely designed to serve the needs of >90% of those seeking answers. Yet if the answer to the system issue you are experiencing isn’t found there, what do you do? Is there a number to call or method to directly email a support representative? No. So you will likely try every path on the help tree until you find one that allows a ‘suggestion’ to be placed in a web form, only to receive an email back stating they will take your ‘suggestion’ under advisement. Have a real-life issue involving a real-life charge to a credit-card? You may be waiting a while. Google’s foray into the mobile space was met with frustrated customers after they seemed to underestimate the support infrastructure required for this consumer product:
….The real problem is figuring out which entity is supposed to be responsible for answering questions and providing support…If you buy a Nexus One manufactured by HTC, directly from Google’s Web site, and use it with T-Mobile’s wireless network–who do you call when you have a problem? Google is only accepting support requests via e-mail, and users are getting bounced between T-Mobile and HTC as neither seems equipped to answer complaints, or willing to accept responsibility for supporting the Nexus One.
What will the support load of all the new Google Apps for Business users look like over time?
- Rackspace, a leader in cloud hosting offerings, has built a solid reputation for hosting services and has many happy customers. Yet a small business doing research on hosting offerings for the CloudSites product will find no detail at all on what is included (disk, bandwidth, etc) for the $149 monthly price. Opening up a chat session will result in the agent first apologizing for the lack of detail on the site, then pasting in a long list of the technical specs for one to make a decision. You can’t view the specs anywhere on the site. Did Rackspace intentionally leave off all product detail in hopes that potential purchasers would start a chat session and, once in a conversation, convert at a higher rate? Perhaps, but many others will find it a curious way to sell a service, and move on. Certainly millions have been spent in automating the thousands of servers and every systems administration task possible within the Rackspace environment to enable scale without human ‘friction’ getting in the way. Yet one cannot even learn about the offering without a human chat session occurring.
- Amazon.com does $25B in eBusiness a year, is the definition of success for web businesses, and has built a world-class fulfillment and logistics platform, not just for their own offerings but for thousands of smaller vendors to leverage. Although Amazon has made web and phone support much more accessible over the years, in its early days customers would share unpublished support phone numbers on Blogs and forums like hot stock tips. Amazon.com still carries books that tout the direct support phone numbers for businesses, including Amazon.com.
- Apple’s App Store has been the foundation for success for the iPhone (and soon the iPad). The App Store approval process is designed to ensure that each new app has value, wont crash the device and is free of malicious intent. Once up and running, these 150,000 apps generate income for their developers (and Apple) with sophisticated distribution and billing systems. It’s an impressive infrastructure and works well in most cases. Yet that initial approval process requires direct human interaction for each app and each of its updates, and is therefore open to human subjectivity, human error, seemingly arbitrary delays, capricious decisions, and a great deal of frustration for iPhone developers (sample, sample).
- Outsourced help and community sites like Get Satisfaction seem to have as many eBusiness clients as they do traditional brick-and-mortar customers, and represent the dynamic tension between an organization’s desire to ‘own the customer experience’, and the realization that customers are not getting what they need internally, and will be out on the Web building communities around their frustrations with your product, so one might as well try to centralize it…
Clearly there is an unmet need for customer support for many eBusiness organizations, and no amount of salivation over a cool tech-laden business model will make this go away. Some drivers of this situation:
- Lots of investment and design in the transactional system or business idea, with too little focus on support and assistance for the back-end
- An assumption that the entire system is largely bug-free, and that there are few use cases or modalities that will fall through the cracks. Those customers that do fall outside the parameters may feel like they are the only customer it has ever happened to, as the online help docs don’t cover them, despite spending hours looking.
- That a simple 1-time build of a help tree will suffice, without updates. How are people actually using the tree? Are they getting answers from it, or are they giving up and going elsewhere?
- That paying customers will gladly help out other paying customers for free, and that this method can serve as the primary support function for your new web business.
- That the customer should be prepared to spend substantial time navigating poorly-designed help/support resources rather than inconvenience you with an issue. “If those pesky customers were only smarter about their navigation of the help system, and saw the brilliant logic in my design….”
Its important to note that these aren’t Facebook users complaining about a Scrabble application misbehaving on their news feed. These are real, paying (or potential) customers with money on the line, money they are giving you, and credit card charges to pay. As is often the case with tech-focused enterprises, the sexiness of the model or the tech being developed often drowns out discussion on the non-sexy but critical customer support requirements. Is this even a topic of discussion in start-up pitches to VCs? How many questions are asked about the customer support model in those meetings? Its cost over time?
How long before one or more major web operations finds their customer satisfaction ratings down in the basement with cellular providers and cable companies?
Your comments are welcome. If this post was helpful, you might like to subscribe to the RSS feed, sign up for weekly updates via email or follow me on Twitter.