1. Not the best case study in the world, but a fun one:
    I was bored at work seven years ago today. I can tell that because I was just looking through old emails and seven years ago today I was emailing Seb with an 'audit of recent text messages'. Here is who I was text messaging seven years ago:
    What fun to have that data and be able to look back - not at who I remember being in touch with - but who I was ACTUALLY in touch with. There are some names on there I had completely forgotten about! One good reason to gather data and keep it laying around, I guess. I'll probably dig it out again in another seven years and then I'll really appreciate it!
    (If I hadn't swapped phones so often I'd have kept all text messages since then and I could do some trend analysis. But technology has sadly not made this possible, but I'm starting again now with my iPhone and this helpful site
    (As shown in a previous post, there is a neat tool to do this kind of analysis on your Gmail, but I'm not enough of a nerd to be able to get it to work. Rubbish.)
    1

    View comments

  2. Since someone leaked the British National Party's membership list to the internet, I thought I'd take a look.
    Here are some common first names:
    And some common last names:
    No surprises there then!
    I couldn't find a Singh or Patel. Odd.
    3

    View comments

  3. I'm currently working on advertising strategy and someone just forwarded me a document arguing that when a major brand 'goes dark' on advertising for a long period, it significantly damages the brand. But I drew the opposite conclusion from the document!
    The article is by an organisation with -to put it politely- a vested interest in keeping the advertising dollars flowing and so there are a few misleading things in there. But I thought one misleading conclusion was particularly fun. Here it is:
    Note that they chose to not visually display the 'no change' result, because it is by far the biggest segment and does not help their case! As a result their tag line is plain misleading: The most likely thing to happen when a brand goes dark is actually that there is no impact on the measures they show! That should be the headline message: "brand measures are unlikely to change when you go dark."
    And the chance of going dark and having no impact or of an impact being favourable is 72%-76%! That is pretty good odds, I'd say. I'm willing to bet that we've done smarter analysis than the competition and so if we were to do that we'd get even better odds ...
    1

    View comments

  4. I saw another article today on how supermarkets waste too much fresh food. I thought I'd post a quick chart showing an easy way to reduce waste: buy more stuff!
    Slow selling products lead to waste. They always will. So look for products that don't sell very well and buy more of them!
    The following chart shows what waste might be expected for fresh food, at different sales rates. Slow selling fresh food (2 singles per week in the chart) might be expected to usually waste more than 150% of what you sell. So for every two you sell, you waste more than three! (And some products waste way more than this, if they are very short life, or if they must be ordered in large cases from suppliers for example.)
    You'll see that this decreases massively as sales increase slightly. Go from sales of two to four singles per week and your expected waste falls below 100% of sales. Increase the weekly sales to eight and you're back at a reasonable level: expected waste of about 20% of sales.
    So if you are worried about the ammount of food that supermarkets waste, simply look out for the slow selling products and put a couple in your basket. a couple of extra sales will make all the difference to supermarket waste.
    0

    Add a comment

  5. I was playing around with a really cool and easy to use analysis tool for Excel, Analyse It. And I thought I'd show how easy it is by running an old data set through it. So here is some quick analysis of an old political campaign I worked on. Seriously, this took about 30 minutes total. 
    We won the race overwhelmingly. In fact we won every precinct in the city. One thing the tool allows you to do is compare groups of data. Here is our vote share by city ward, with each blob representing a precinct:
    This type of chart can be created with about three button presses. It shows we won big in our home ward (yay!). It also shows the distribution of precincts. They were pretty bunched together, but it might be interesting to look at the outliers. Why were they so different?
    We did an ok job of predicting where turnout was going to come from. I can tell this by running a regression of our predicted turnout by precinct and the actual turnout by precinct. The tool does this very simply, with no specialist knowledge required:
    This graph also takes about three button presses to generate, and comes with a neat set of stats telling us that we did an ok (not great) job of predicting turnout by precinct. But then we should have done, since historical turnout by precinct was widely available, and the best predictor of future behaviour is past behaviour.
    Now here is the interesting bit: we were crap at predicting how people would actually vote. I should stress that this was a Democrat versus Democrat election, so it is a notoriously difficult thing to call. But either way, we didn't do a good job! Here is our categorisation of precincts in to how we thought they would vote in advance versus how they actually voted:
    You have to say that's pretty bad. In the precincts we called for our opponent, we actually did better than those that we thought were too close to call. Its good to know that we did better in the precincts we thought would support our candidate. But in those precincts there is a really wide range of support. They could probably have been better broken out in to different categories.
    We did try to break them out further, but without much luck. Here is a more detailed look at what vote we thought we would get versus what we actually got:
    We did a good job of predicting the very best precincts, but beyond that group, all of the other five groups we decided on behaved very similarly.
    We also broke the precincts out that we thought our opponent would do well in. That also didn't turn out to be a good predictor of the actual vote:
    Pretty depressing really!
    Data and targeting has played a huge part in many political campaigns. That is particularly clear from '08 and organisations like Catalist have played a huge part in this. In politics I had the pleasure of working with a fantastic group of people who would honestly evaluate their own work, but ...
    ... usually after elections finish, candidates' political campaigns close their doors and everyone goes off to the next campaign, tired and often ready to forget about the campaign. There is no will to evaluate what worked and no money to fund evaluation. And besides, the people who did the analysis often don't want to evaluate for fear that their analysis that they were paid handsomely for wasn't actually helpful after all. This is a tragedy.
    Even if predictions aren't that helpful, having lots of data on a campaign makes people believe. Everyone has confidence in a well presented graph and they feel that those extra phone calls and those extra door knocks are being well targeted. In reality, I often believe that when you are lost, any old map will do. 
    However if there were more evaluation after the effect, perhaps campaign analysis would improve quicker and it would more often actually help the campaign to win, rather than just boosting morale.
    0

    Add a comment

Labels
If you like this you'll like:
Info Clarity Archive
Loading