Mailing List Archive
tlug.jp Mailing List tlug archive tlug Mailing List Archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]Re: [tlug] Browser share in Japan?
- Date: Mon, 30 May 2011 12:33:25 +0900
- From: "Stephen J. Turnbull" <stephen@example.com>
- Subject: Re: [tlug] Browser share in Japan?
- References: <4D7972F4.3070304@example.com> <4DDDD697.4060609@example.com> <871uzkanex.fsf@example.com> <BANLkTikPo_+UBGfjrMxFYnLX9j4sL+wY9g@example.com> <87oc2o8wum.fsf@example.com> <BANLkTindSoqUwu5XB6Y=gQjLrvM_SRTbvQ@example.com>
Martin G writes: > If so, then maybe "market share" is a useless and empty concept > when talking about browsers. Indeed. Actually, my argument is that it's harmful, in that it redirects attention from what's actually profitable, which is your market share of *customers*. The browser a customer uses is an interesting fact about the customer. I am quite sure that customers who use Safari are different from customers who use IE8, and both are different from customers who use Iceweasel. Some products (CDs by AKB48) probably are hard to distinguish by browser. Others (products that run only on the Mac) are going to have browser user shares of page views that are wildly different from global averages, and the conversion rates of page-views-to-sales are also likely to differ wildly, I suspect you will agree. So targeting your page by browser share is dumb, very dumb. What you want to do is target your page to potential customers. So you go with the browsers with high "correlation X user share", and ignore the others. If there's a significant correlation between potential revenue per customer and browser, the strategy of targeting high-user-share browsers will lead you astray. If there isn't, then the correct rule reduces to using browser user share as a *proxy* for customer market share. If you can get those correlations, then the correct rule is no more expensive than the simple one (after all, your spreadsheet will do the multiplications for you). Now, this is all obvious to anybody in the field, I suspect. The problem is remembering that it's *only a proxy*. The Harvard Business School casebook is full of cases where companies failed in part because some department got hung up (perhaps even with monetary incentives from the executive board) on a simple metric like browser user share. The simple metric probably was appropriate when originally established, but when the market changes (and it inevitably does), that metric is as useful as following the taillights of the lemming in front of you. As Bugs Bunny used to say, "Watch that first step. It's a doozy!" Avoiding use of the term "browser market share" is one easy trick to help watch the first step.
- References:
- Re: [tlug] Browser share in Japan?
- From: Darren Cook
- Re: [tlug] Browser share in Japan?
- From: Stephen J. Turnbull
- Re: [tlug] Browser share in Japan?
- From: Martin G
- Re: [tlug] Browser share in Japan?
- From: Stephen J. Turnbull
- Re: [tlug] Browser share in Japan?
- From: Martin G
Home | Main Index | Thread Index
- Prev by Date: Re: [tlug] Browser share in Japan?
- Next by Date: Re: [tlug] Browser share in Japan?
- Previous by thread: Re: [tlug] Browser share in Japan?
- Index(es):
Home Page Mailing List Linux and Japan TLUG Members Links