Seb Chan works for the Powerhouse museum, and on a lot of projects that need measurement in one way or another. This session was a rapid fire tour de force through measuring web visits and some of the fallabilities of old-style reporting of them. These are my notes from the session.
He explored the search problem – Powerhouse has up to 9 ways to enter its site and how more and more people don’t come through the home page – searches have visitors arriving on all sorts of pages from google.
85% of traffic comes through search – bringing a multiplicity of reasons for visiting with them. 16,000 variations of powerhouse, 18,000 of museum occur across the site.
Searchers spend different lengths of time – searchers 4 minutes, home page users about 7.5 minutes.
Log analysis is not always useful – can’t differentiate real from bot traffic for instance.
Powerhouse also uses video ads and Chan made an interesting comment about video ad monitoring – although the video viewed by 10,000, just 16% reached the end – but the museum paid for the 10,000 impressions.
Email newsletters, reporting tools, live tracking of websites. Comparative measurements and benchmarking were all covered in the first half of the lecture. The second half put the focus on web presences and combining them with other reporting.
One of the tools used was a photo of the day blog – about 500 days of photos. The museum encourage people to use subscriptions. Feedburner reports clients who don’t visit your site. For some blog comments are more important – for others citations are. Technorati measures blog posts about your blog – and in some cases the other comments occur on other websites – and perhaps that is better as it shows people may not be talking to you, but may be talking about you.
In that sense, content continues to move further and further away from us. We are crowding out traditional reporting with other stuff – you can learn more from Facebook than from web log analysis.
More intereactive experience with content than on our own website. Net promoter score is a one question survey. On a scale of 1 – 10 how likely are you to recommend this resource to another person.
7-8 – passives
% promoters – % detractors = NPS
Web is all about recomendations. To get content used more it has to be recommended to others. They have to be excited enough by the content to tell friends about it. Content has to be easily shareable
Set goals segment and observe trends – benchmarks, city first, trends in local community more important than international. Use appropriate tools, find benchmarks and aim to measure net usage growth.
Get better as you go
Need to improve content as a result of tracking web use – that is the benefit of doing the tracking in the first place and helps provide evidence base for trying things out.
Combine qualitative and quantitative – don’t report volume alone. Visitation does not equal satisfaction.