Well, Nik and Rob just beat me to it, but this might be worth adding. I think you don't have to worry too much about accuracy, as long as you understand that accuracy is tough to define and harder to achieve. But, if you want to be most honest and fair, you can:
1) only compare like against like--"this section of our site, using this version of this analysis software, did 10% better after we redesigned its index page" (ie follow trends, like Nik said) b) be aware, to the extent possible, of when your analysis software changes. For an interesting write-up, see http://www.imediaconnection.com/printpage/printpage.aspx?id=16342 (we don't report "time on site" outside the website department--I don't want our stats going out with complicated disclaimers). c) use page views as your primary measure, since it's the most reliable of the primary measures (as long as you're careful to count them the same way). Don't use Time on Site, and don't use visits. On the other hand, lots of page views is not always a positive thing (think search, or your donation section--really you want them to find what they want and take care of it). For what it's worth, we use visits, even though I don't really trust the numbers, because it's just so effective at communicating how popular the site is. I'm happy to explain what it means, if people ask ... --Matt On 3/11/08 5:27 PM, "Nik Honeysett" <NHoneysett at getty.edu> wrote: > Google Analytics is a page tagging service, its accuracy will depend on you > and what you want. If you have comprehensively tagged all your pages, then it > will comprehensively report on your traffic, but that traffic will include > bots and spiders i.e. not real people. You can filter these out - up to a > point - but are real-people numbers important to you? Its unlikely that you > will ever get an accurate number of real people visiting your site, so its > best to accept that. > > Assuming, you did have an accurate number and that number suddenly doubled or > halved what would you do? What would happen? In either case you would want to > know why, but you're not really interested in the number, only the change. You > can figure that out whether you have real-people numbers or all-inclusive > numbers. Your real concern should be trends and Google Analytics is fine for > this, as long as you know what you're reporting and you don't change the > filtering. > > -nik > >>>> "Jeff Tancil" <jtancil at tenement.org> 3/11/2008 12:58 PM >>> > That seems to beg a question: what stats service is useful? As a fairly > dinky Museum, we use the best free service, GoogleAnalytics. How badly > do people think that skews? > > -----Original Message----- > From: mcn-l-bounces at mcn.edu [mailto:mcn-l-bounces at mcn.edu] On Behalf Of > Nik Honeysett > Sent: Tuesday, March 11, 2008 3:53 PM > To: Museum Computer Network Listserv > Subject: Re: [MCN-L] Website benchmarking > > Like you say, these sites are ok for trends but do not give anything > close to accurate figures for traffic your numbers. > > -nik > >>>> Russ Brooks <RBrooks at mus-nature.ca> 3/11/2008 10:43 AM >>> > When we noticed a shift in our web statistics we wondered if it was just > our > site or was it something that was affecting all other museums. > > We found the two following sites very useful in providing us with an > opportunity to compare our performance to that of other museums. > http://www.alexa.com/ > http://www.compete.com/ > > These two sites allowed us to see the exact same patterns in traffic > affecting nearly all other museums. > > These sites can also be useful when trying to determine Internet usage > trends. Is Facebook still hot? Type in their address and you can see the > results. > > > > On 3/11/08 1:26 PM, "Leonard Steinbach" <lensteinbach at gmail.com> wrote: > >> I was wondering whether anyone uses any particular web traffic > statistics to >> compare the performance of their website to the websites of other > museums. >> In effect is anyone benchmarking their website against others, or know > of >> any studies or papers which address this issue? >> >> Thanks >> _______________________________________________ >> You are currently subscribed to mcn-l, the listserv of the Museum > Computer >> Network (http://www.mcn.edu) >> >> To post to this list, send messages to: mcn-l at mcn.edu >> >> To unsubscribe or change mcn-l delivery options visit: >> http://toronto.mediatrope.com/mailman/listinfo/mcn-l > > _______________________________________________ > You are currently subscribed to mcn-l, the listserv of the Museum > Computer Network (http://www.mcn.edu) > > To post to this list, send messages to: mcn-l at mcn.edu > > To unsubscribe or change mcn-l delivery options visit: > http://toronto.mediatrope.com/mailman/listinfo/mcn-l > _______________________________________________ > You are currently subscribed to mcn-l, the listserv of the Museum > Computer Network (http://www.mcn.edu) > > To post to this list, send messages to: mcn-l at mcn.edu > > To unsubscribe or change mcn-l delivery options visit: > http://toronto.mediatrope.com/mailman/listinfo/mcn-l > _______________________________________________ > You are currently subscribed to mcn-l, the listserv of the Museum Computer > Network (http://www.mcn.edu) > > To post to this list, send messages to: mcn-l at mcn.edu > > To unsubscribe or change mcn-l delivery options visit: > http://toronto.mediatrope.com/mailman/listinfo/mcn-l > _______________________________________________ > You are currently subscribed to mcn-l, the listserv of the Museum Computer > Network (http://www.mcn.edu) > > To post to this list, send messages to: mcn-l at mcn.edu > > To unsubscribe or change mcn-l delivery options visit: > http://toronto.mediatrope.com/mailman/listinfo/mcn-l
