The book is based on the material from the pretotyping workshops that Jeremy and I have been conducting in the past few months (at Google, Stanford and for several companies.) In addition to introducing several new concepts (e.g., the Return on Pretotyping Investment (RPI) metric), the book does a great job of extending, augmenting and clarifying many of the ideas presented in Pretotype It. Download a free PDF or splurge and buy the Kindle version for $0.99.
I'm not sure how this factoid might affect a revision of this book, but GroceryGateway (http://www.grocerygateway.com) seems to be doing quite well up here in Toronto, Canada - they've been around for years now. What I'm getting at is that in your example, I'm not sure Webvan's failure was necessarily due to people unwilling to commit once the service was offered to them, but rather due to bad timing or execution failure.
ReplyDeleteRe: Grocery Gateway being successful where Webvan failed.
ReplyDeleteThank you for the comment.
Webvan makes a particularly good case because the 'it' that had to be tested (and should have been pretotyped) was not whether or not some people would order groceries online. That is, and was, a given. Some people with some frequency will undoubtedly take advantage of such a service. But how big is that market?
Some of questions that Webvan should have tested with much smaller scale pretotypes were: What percentage of people would order groceries online? How often would they buy online? What kind of groceries would the buy? Would people in cities use the service more than people in the suburbs? Etc.
Webvan's 'it' was not just an online grocer, but a massive play and bet on the assumption that A LOT of people would modify their behavior and shift their online grocery shopping online.
In addition to that there was, as you mention, a timing component. People were not as comfortable with online shopping in 2001 as they are today. Something that is a clear 'wrong it' in one era could very easily turn out to be 'the right it' in another era (or another market/geography.)
Finally, as you also suggest, there were also execution issues - not surprising since many of the key players in the company had little or no experience in the supermarket industry.
I hope this clarifies why we believe that Webvan is actually a very good example of something that should have been pretotyped to answer some key questions before committing $100M+ dollars.
Alberto
Far from religious about choice of font but the one used in the PDF simply hurts my eyes. Anyway, I can't wait to read it as a follow up to the first book.
ReplyDeleteZarko, thank you for the feedback about the font. I [Alberto] am to blame for starting the use of Courier to emphasize the manuscript/draft nature of pretotyping, but you may be right that it does not apply well to longer documents. Point well taken.
ReplyDeleteInsightful book(let), Alberto. I work as a test consultant in a B2B software integrator, and too often I see detailed functional designs, that lose their relevance once people experience the finished product for the first time. The whole process of functional design takes place in thoughtland. I would like for us to start pretotyping before we build.
ReplyDeleteI see it as follows. Right now we DESIGN, then BUILD, then TEST. I think we should DESIGN, then TEST (pretotype), then BUILD and then TEST. I'm leaving out the requirements phase, but the same goes there, I think.
I had no problem with the font. It looks dated, but the content is relevant.
Thanks,
Paul
You book was part of learning materials for the advance International Business course in Turku School of Economics (Finland). Thank you for the interesting book. Easy to read, easy to understand, clear and vivid examples. I can say that I certainly will use approaches I saw in the book in my future career.Thank you!
ReplyDeleteRegards,
Elkhan
(Azerbaijan)