Out With the Old Process, In With the New – Completely

Remember Webvan? Those pretty brown trucks. We saw one all scuffed up and worn down by the side of the road today and it perfectly symbolized the promise of an Internet-based grocery company that never made it. Lots of theories on its failure have been created, but I am always amazed that most people fail to recognize one of the biggest factors that doomed Webvan. It is a lesson that too few people seem to recognize when running their own business units.
Replace the competition or be restricted to a niche
Webvan’s biggest failure was not offering same-day grocery delivery. To say that this single factor caused an epic business failure might be going too far, but it certainly restricted Webvan to being a niche player. Designed and capitalized as a replacement for traditional grocery stores, as a niche player its success would probably never have produced significant financial results for Louis Borders and its other investors. This was more than a customer satisfaction issue. It forced Webvan’s customers to frequent the very brick and mortar grocery stores that the company sought to replace.
Why was same-day delivery such a critical failing? Without same-day delivery, Webvan couldn’t become the grocery resource of choice for even their own loyal customers. The most organized, once-a-week shoppers often make additional trips to the store for a gallon of milk, fresh fruit or the like. And while there, they typically pick up additional items. With Webvan, if you’d run out of milk or realized that you needed something for dinner tonight, you could order it for delivery… tomorrow.
Parallel systems don’t reinforce desired behavioral change
What occurred at Webvan is just the tip of the iceberg–the crux of the problem is far deeper. Anytime one technology or process replaces an incumbent technology or process, there is a period when both processes are in use. For most products and services this is not a problem. Banks get incremental savings when people use the ATM rather than a teller, but other than adjusting the number of ATMs and tellers, it does not make a huge difference.
However, sometimes the new process needs to completely replace the old process in order to be economically viable. If the same bank moves from a paper-based record keeping system to an electronic record system, it has to operate two record keeping systems–with lots of fixed costs–until the paper system can be completely eliminated. In the meantime, instead of achieving a cost savings, the bank’s expenses explode. Ouch!
Where does complete replacement matter?
For consumers, maintaining two separate grocery acquisition channels is no big deal. Many of us already shop periodically at multiple grocery chains. However, if we are talking about a manufacturing process, a product or service integrated with an ERP or CRM system, or another product or service that requires a significant on-going investment by the user, it’s a very different story. Salesforce.com is not going to sell many seats if they cannot get potential customers to give up what they are currently using.
Whether you are buying or selling, next time you see a product, service or process justified on the basis of replacing something currently in use, think about what has to happen in reality to achieve the promised benefits. If not, think about that lonely old truck and don’t make the same mistake Louis Borders made with Webvan.