Analytics and inspiration

October 31, 2008 — 11 Comments

Geoffrey Moore (of Crossing the Chasm and Dealing With Darwin fame) and I spent yesterday speaking with a large group of analytics folks brought together by analytics software company SAS. I have to say that I have always considered analytics to be a long way from design thinking but I left yesterday’s session with a new point of view.

One of the biggest problems in design is knowing what questions to ask. You can take the intuitive approach but this seems to be very random when you are thinking about more strategic and upstream problems. At some point the potential area of exploration covers 360 degrees and goes to infinity. It seems to me that using clever pattern recognition through software analytics might point out interesting areas to explore. I heard about one example yesterday that is connected to a project IDEO did a few years back. When we helped Bank of America develop the savings service, Keep the Change, I had assumed that the consumers we observed were picked based on intuition. One of the BofA analytics folks was in the audience and she told me that her team had worked on the project and analyzed consumer data to identify some target groups of users that the company were interested in understanding better. So it turns out that analytics helped point us toward a segment that ended up inspiring us to create a pretty successful service.

I wonder what interesting patterns might emerge from the use of analytics and be a source of inspiration for design thinkers if we put some thought to it.

By the way, the graph above is the Google trend for use of the term design thinking over the last four years.

Tim Brown

Posts

11 comments on “Analytics and inspiration

  1. Dear Tim,

    Not only I agree that there is a lot of inspiration to look for with the use of analytics, but I also believe that being able to find new and better ways of visualizing data allows you to understand with better detail – or find new knowledge – in some information that might otherwise be ‘lost’ within the data. Hans Rosling’s statistics on TED are a eye opener on this (http://www.ted.com/index.php/talks/hans_rosling_reveals_new_insights_on_poverty.html )

    So, for me, visualizing data can also be a way of design thinking. What do you think of this?

    All the best
    Pedro Monteiro

  2. Tim and Pedro are both making good, albeit orthogonal, points. To Pedro’s, visualization is critical to statistical perception. Scatterplots and histograms allow the data to jump out at you. Ditto for geo-coded visuals–they always make you see something you didn’t know. The best person for this sort of design thinking is probably Edward Tufte.

    On Tim’s point, data mining and personalization have shown how effectively statistics can focus innovation on the right segment, issues, etc. That’s how Amazon gets to recommend you the right book (or the wrong one if you have just bought a book for a really wierd friend). At the end of the day, a direct encounter with novelty is still at heart of innovation, but creatures are made up of more than just their hearts. So design thinkers everywhere–take a statistician to lunch, explore if not the dark side, the gray side.

  3. It is astounding to me what we could know but don’t. I working an architecture firm (the bricks and mortar kind) as a researcher whose job is to find out all existing research on how space affects the way we live. In practice this is a mix of environmental and industrial psychology, anthropology, and a dash of building performance science thrown in for good measure.

    It is astounding to me what little information is out there with regard to the effect the built environment has on our everyday lives. “Life-cycle analysis” is the new buzzword for material and system selection (as it should be), yet average baseline information for water use, energy use, and basic maintenance and operating cost data for different sectors and is startlingly difficult to find. Though there are a few exceptions, there is not currently a good tool for architects to make informed life cycle decisions from a predictive analytical system based on good data.

    Just playing around with the numbers that are available we’ve learned of some interesting opportunities that are a bit counterintuitive. For instance you can save more on energy costs than water/sewer costs by investing in water savings devices where hot water is used. Water costs are extremely cheap in the US, which makes a financial argument for water use reduction investments difficult. However, restaurants can save thousands/year by investing in a low-flow pre-rinse spray valve that costs less than $100. Similarly cheap changes in lavatory sinks in hotels yield huge dividends in energy savings. The whole premise is the less hot water you use the less you have to heat.

    Not related to architecture but applicable to the analytics discussion – there’s an interesting Wired article on a psychological approach to winning the the Netflix Prize.

  4. I agree with the post’s main point as long as we remember that data don’t speak for themselves, as much as stimulate meaningful conversations between design research, users/customers, and client management. Those conversations can result in interpretations that make a difference for the design of products and services offered by the business. I like Gerald Zaltman’s take on the use of statistics in his book, How Customers Think. He noted, “the various pieces of information that we gather through statistics, personal observations, and other data sources…are all stimuli that influence our thoughts, feelings, and behaviors. By viewing data in this way, most managers suddenly see the value of collecting multiple kinds of data” (p. 275).

  5. Curiouser and curiouser. First, about the last post and full disclosure. Just as you, Tim, worked with PNC, I also must fess up to having worked with ING Direct when they launched into the U.S. I guess we really do talk about what we know, eh?

    But now this? SAS? Analytics as a way to human insight? I had to laugh, and then wonder. You see, at the very same time we was developing the launch campaign for ING, our firm was also working on the re-position campaign for the SAS Brand.

    Just wanted to let you know that I’m obviously engaged in the subject matter and will be alert to opportunities to contribute where appropriate. Keep the smile. — Scott

  6. “…but i left…with a new point of view.” was the best line in this post.

    here’s my take: if something helps generate these “new point of view[s],” then aren’t we moving in the right direction of design thinking? my one caveat is that whatever it is that we consider needs to be seen within context, then afterward perhaps pulled out of its context, shredded, pieced back together, turned upside down, flipped around, cut in half….

    just some thoughts from a stranger,
    liz

  7. One of the problems with “analytics” is the difficulty we all have in visualizing what is going on when there are multiple variables (both dependent and independent) and feedback loops with time lags. I have found that the employment of system dynamics is a great help in dealing with this problem and built a very successful one-man consulting practice using this methodology in the gas and electric utility industry — one of the least innovative business sectors owing to heavy government regulation. Using SD models, utilities have been able to save millions by watching simulations unfold in time and adjusting their systems’ design and operating parameters to see what happens. After multiple iterations the insight gained is quite remarkable.

    The process is a synthesis of analytics and creative intuition. I highly recommend it.

    Ed Curtis
    E J Curtis Associates, Inc.
    EJCInc@aol.com

  8. Hi Tim

    >> I wonder what interesting patterns might emerge from the use of analytics
    >> and be a source of inspiration for design thinkers if we put some thought to it.

    I’m coming at this as an IT nerd with a passion for design. I ‘do’ business intelligence and data warehousing in an Australian regional university. What I also try to do is get people excited about the numbers through report design and presentation. Tufte has been mentioned and his stuff is great although I can’t help but feel his latter volumes are a bit of a rehash (but I still bought them!). I also see links to Hamel in your posts – he is inspiring too but in a different way (his ‘Future of Management’ is well worth a read).

    So what is exciting about numbers? Check out ‘Super Crunchers’ for some great real world examples of what numbers can do for us akin to Hamel’s stories of innovation strategies. From memory there are examples in there about designing google-type advertising links and prototyping them to see what makes a successful (ie clicked) ad line. The huge volumes of data associated with today’s Internet traffic quickly allow us to run parallel analyses using control groups and see what works or doesn’t work.

    Of course what people click on isn’t necessarily ‘good design’ but perhaps it is ‘fit for purpose’. Is that the same thing? (I know I’m being overly and overtly simplistic here)

    Another great bit in Ayres’ book involves taking a psych quiz where you have to provide a two-number range in which answers to questions such as ‘how much does a Boeing 747 weigh?’. The point here is we tend to be over confident about our level of understanding or knowledge instead of relying on data to help us. The best combination is when data and experience come together.

    So what’s all this got to do with design? I think design can be informed through data, data can assist our natural creative thinking to produce a product that is both inspiring and beautiful and yet functional. That product could be a website, it could be this blog (have you checked stats on usage when you try one theme over another?) it could be vehicle colours, sales of toothpaste before and after packaging mods or re-branding.

    With RFID tagging you can get huge volumes of data that can inform consumer product design (which products did a consumer pick up but put back on the shelf only to take a more (maybe?) attractive alternative?

    So I didn’t answer your question but I maybe I elevated the nerdy IT world of Business Intelligence and Data Warehousing up a notch. Happy to talk more about this if you are interested.

    Cheers,
    Rob

  9. Interesting post, and it’s true that there is definitely a place for analytics in design.

    While I would say the greatest strength of design thinking is its ability to explore “to infinity”, I always assumed its second greatest strength was to match this out of the box thinking with solidly grounded empathy and contextual awareness. In the case of BofA, it appears that someone provided this context (set the parameters of what was “inside the box”) and fed it to the analytics group. In all likelihood design thinking had nothing to do with it!

    My greatest challenge with clients is the need to balance these two strengths of design thinking. Exploring far and wide to find the right questions might be my first impulse, but as with many things, the answer is often in my own back yard.

  10. With vast amounts of data, sometimes the most daunting task is to find the right question to ask or know where to begin. I’m wondering if rather that using data to inform design thinking if we should be talking about using design thinking to inform the data analytics process.

Leave a Reply

Your email address will not be published. Required fields are marked *

*

HTML tags are not allowed.