User stats and feedback

Analysing user stats and feedback
Discussion: Analysing web logs

[Malin] ISP doesn’t send user statistics

[Sonya] Looking for better user statistics. Currently examine stats for

what documets are used more
where new information is needed
what need to be updated

Looking at options of what are usable statistics. Too much unnecessary information in current stat printouts. Doesn’t want states that are too long and useless information. Will look at options for useful concise stats.

Currently draws own diagrams based on text stats to report to management.

[Irene] Doesn’t know – thinks other users within organisation use statistics

[SK] Relies on users writing feedback to wwwboard

Demonstrations

Webalyser (c2o) - graphical summary on first page, more detail within

MKstats – (Toy Satellite) very user graphical interface.

Wwwstats – (AWORC) – comes as email. Will get larger as site gets larger, so will need to be delivered in another format. These have been highly customised, e.g. showing user search strings.

All the above stats software show what objects are requested.
Most are customisable. Up to us to communicate directly with the sysadmin to request customisations – eg I only need this certain information (doesn’t need server types, cares about where hosts are, browser types).
Ask the ISP / hosting service / system administrator - what they recommend, provide

Discussion: How do we get feedback?

Magazines/print, we know the audience /feedback via the mailing list.

Web sites can generate similar feedback, but who the individual or organisation is? you don’t know.

In participants experience how can we ensure consistent feedback?

[Sonya] Via mailing lists, response to specific things, can look at site and see via stats that mailing list (publicity) is working. People are directed to site for certain information.

[Cheekay] Aside from stats, can get personal feedback via mailing list, members of AWORC network. AWORC web survey (usage and vision for redesign) sent to went99 participants – unfortunately 10% response rate (survey design? language?).

Beijing infocon work – prepared a survey about type of information used. Asking whether aware of Beijing site. Latter part – what needs to be added in website. Respondants were intended audience. Important step to keep asking intended audience what more they need. Info you have may be obsolete.