After ManyEyes, Swivel is another site that another site that lets users upload and share information. It’s more explicitly modeled after Flickr and other web 2.0 sites, with community building and “fun statistics” (highlights today include chart showing number of people attending the burning man festival and the nationality of cyclists in the 2007 Tour of California).
Post hoc ergo propter hoc
The site illustrates one of my favorite issues with business intelligence: that having the data and being able to analyze it doesn’t mean that people will necessarily make the right conclusions and take the right decisions.
Much as I’d like to think that the chart implies that drinking more wine would help reduce violent crime in the US, I’m guessing there are other causal factors at work here.
Here’s another example, purporting to show relative crime rates for Wal-Mart and Target. Again, viewers should be cautious of how the data should be interpreted (if it’s accurate — one of the inevitable problems of this type of site). For example, it probably means that there are more Wal-Marts than Targets in high-crime areas, rather than saying that a Wal-Mart that’s right next to a Target store is likely to suffer more crime.
Stimulating debate
Luckily, In the comments area, users are free to debate competing interpretations of the information, and link to other charts that may shine light on the question. Organizations analyzing their own data could clearly benefit from this type of functionality, and some of the vendors in the commercial BI space have offered technology along these lines (e.g. Business Objects XI threaded discussions). Easier, more effective collaboration is clearly part of the future of BI.
Chart junk?
Personally, I’m not particularly interested Eximio’s distribution of pee frequency, but then again, that’s one of the benefits of the 2.0 approach — there’s full support for search and “most popular analyses” to help navigate through the vast number of different charts (more than 1.1 million at time of writing) that are already available (the site is only two months old).
Anything that gets people more interested in data, and more aware of some of the issues surrounding its interpretation, can only be a good thing for the BI industry.
Comments
2 responses to “Swivel.com: Flickr for Data Lovers?”
Timo, thank you for writing about Swivel. For me, correlations and comparisons are fun because they often ask a question rather than giving an answer. The Target/Walmart graph is a good example: what’s the real reason for the disparity? Is that in fact the case that Walmart stores are located in higher-crime areas? And by the way, who compiled this data and what is their agenda? Changes in wine consumption changing crime rates? That’s nonsense, right? Or is it? Who’s to say? And how do you know for sure? Probably only by digging deeper and thinking for yourself. If a graph at Swivel can make someone stop and think it over, it would make me very happy.
Thanks again for the thoughtful article and for posting our graphs, and please drop me a line any time with ideas or feedback on Swivel.
Thank you,
Dmitry Dimov
Product Chief & Cofounder
http://swivel.com
BI 2.0?
First, I agree with most of the coverage of business intelligence 2.0 so far: yes, the term is a little tacky — but it’s irresistible (and a great way of finding people blogging about Business Intelligence) An irresistible term A…