12
November
2007

Flickr usage going down? 15:17 on Monday

It seems like all the camera usage graphs at Flickr are pointing down and look somewhat like this:

Are people using Flickr less than before? Or is camera usage more distributed by model than before? (I find this unlikely.) Is Facebook to blame?

9 Responses to “Flickr usage going down?”

    Mentions elsewhere:

  1. no sense of place :: links for 2007-11-16
  2. Comments:

  3. Asmo Says:

    I believe the reason is Facebook indeed. Quite many friends are nowadays uploading their photos there instead of good old Flickr. Even though there’s a Flickr application in Facebook (but it is not synchronized with FB’s albums section) and even though Facebook treats the photo files the way that they look a bit rough over there…

    As “everyone” is in Facebook it sucks life out of other services, even though the other services do the actual trick better. Sad but true. :-|

  4. Niko Says:

    Does anything beyond “good enough” actually make a service better? Or does featuritis start there? Obviously Facebook’s photo handling is good enough for many people and in some ways even better: the tagging of people in the photos is a very clever feature.

  5. mattw Says:

    The graphs on the camera usage are all shown as the percentage of Flickr members (afaict) which doesn’t give any data about absolute numbers. We can tell one or a combination of several movements:

    • The Flickr userbase is growing primarily in the area of people who don’t use SLRs (either cameraphones, or people who don’t take photographs)
    • Existing members are no longer using their SLRs (either changing camera or, if some element of ‘recency’ is included in the photograph, no longer posting photographs at all)
    • An error has been introduced, artificially boosting the total Flickr membership
    • An error has been fixed, which was over-counting the number of SLR users
  6. darren Says:

    i have a feeling it might have something to do with the importing of yahoo photos into flickr over the last few months.

    looking at this image (http://www.flickr.com/photos/gustavog/2025477227/) we see that when yahoo pictures are imported (and set to private) that they affect the amount of pictures that are available privately.

    the graph above from the flickr site seems to be for the “Top 5 Canon Cameras in the Community” (although on flickr.com no time scale is given, but it is different to that used in the november chart on the same page). so im guessing the importing of yahoo pictures swelled the numbers of members (and a lot of these members would have had older cameras than the current dslrs in the canon top 5.)

    the two graphs (in the original post and in the gustavog analysis) both have the same general shape so i guess one could suggest that they are caused by the same factors.

  7. Niko Says:

    Matt, according to the blurb at Flickr the graphs don’t (or shouldn’t) take into account users who don’t take photographs. The graphs basically represent EXIF data from images uploaded to Flickr, which already takes away a portion of images, either from cameras that do not support EXIF, or edited images which have the metadata stripped away.

    And Darren, I think you might be right. Graphs for all cameras from SLRs to cameraphones have more or less the same shape for the last few months, so the Yahoo import might be to blame.

  8. Ryan Shaw Says:

    Flickr also added a number of localized international sites this year (and I suspect like most of Yahoo is seeing the majority of its growth in these international sites). So in addition to the other factors mentioned above, you should expect to see more distribution among models, reflecting the different preferences in these new markets.

    One more reason why decontextualized graphs and charts should be treated as pretty pictures, nothing more.

  9. Ryan Shaw Says:

    See also the explanation of normalization:

    “The graphs are ‘normalized’, which is a fancy way of saying that they automatically correct for the fact that more people join Flickr each day: the graph moving up or down indicates a change in the camera’s popularity relative to all other cameras used by Flickr members.”

    So you would never be able to tell anything about Flickr’s growth or lack thereof from these graphs (otherwise they wouldn’t be making them public).

  10. Niko Says:

    Ryan you’re right. I read the explanation on normalization, but in a moment of .. stupidity I understood it incorrectly. :) So basically all the graphs for all the popular camera models going down at the same time could be interpreted in more camera models having been introduced to the database at the same time. Or…?