Opening Data to Innovation: stories, takeaways and twists

by Kate Hammer

Cross-silo Collaborations

Abhay Adhikari invited me to a beta workshop on cross-silo collaborations. The workshop has been designed to meet the needs of organisations that sit upon a mountain of data….and want to get inventive about how they make meaning from the data… and use insights it provides to create valuable new products, services, processes, etc.

While some data is proprietary and other data is open (available for use by all), using data for innovation means opening it by asking questions and seeing what insights the answers communicate. It’s those insights that make data-driven innovations meaningful.

The workshop was terrific, and it got me curious about how big data can drive innovation. Being a storyhunter, I hunted stories. Drawing on a network of Twitter conversations here in Britain, here’s what I found stories from Britain and Japan that fall into three groups. For each group, I’ve looked for takeaways and twists.

1. Visitor patterns drive site simple redesign with enormous immediate effect

Making sense of the breadcrumb trails left behind by Hansel and Gretel

Two axioms:

  • Digital behaviour leaves traces that are easy to capture.
  • Making sense of the patterns can throw up quick wins.
  • Example:

    This is a story by Jared Spool, founder of User Interface Engineering, a leading product and website usability research and testing firm. It first appeared in 2008-2009, before big data was quite such a noisy conversation.

    An e-commerce site was losing sales. It was the digital equivalent of trolley carts piled high with shopping being abandoned in the wide aisle perpendicular to the checkout lines. What was causing people to flee before completing a purchase? Visitor usage patterns were clear: a button requiring visitors to register before purchasing was at fault. First-time buyers didn’t want to start a relationship with the site. They wanted to buy and leave, end of story.

    So when customers were pressing “Check out” and being advanced to a form requiring them to register, they were bailing.

    Once this was obvious, it was such an easy fix. As Spool tells it:

    The designers fixed the problem simply. They took away the Register button. In its place, they put a Continue button with a simple message: ‘You do not need to create an account to make purchases on our site. Simply click Continue to proceed to checkout. To make your future purchases even faster, you can create an account during checkout.’

    The results: The number of customers purchasing went up by 45%. The extra purchases resulted in an extra $15 million the first month. For the first year, the site saw an additional $300,000,000.

    That’s why Spool calls it the $300 Million Dollar Button.

    Wouldn’t it be wonderful if every quick win were so easy to spot and so elegant to implement?

    Takeaways:

    • Use data to understand users’ experiences
    • Let the data show you what people are choosing not to do
    • Choice points can be re-designed to influence what people do next

    Twist:

    Zappos breaks the mold. To check out, customers must create an account or use an Amazon account. Perhaps the visibility the registration process gives Zappos is key to their legendary customer service?

    Like Herb Kelleher of Southwest Air declining to buy aircraft that tested higher in passenger satisfaction tests because a mixed fleet would be harder to maintain (see Mavericks at Work, Ch 1), this counter-example suggests that blind adherence to “best practices” is, as my friend Stephen Shapiro says: stupid.

    2. Social data reveals peer behaviours in useful ways

    Two axioms:

    High-infrastructure governments may hold data on personal finances.
    Benchmarking against solid data sets provides useful insights.

    Example:

    Compare My Spend is a British web app that brings benchmarking from the world of business and enterprise into personal life.

    Under the bonnet, the website has compiled monthly expenditure information across 11 areas. The data pool is segmented to reflect gender, age, family size and geographical size. So once you identify yourself in these terms, the app invites the user to gauge their monthly outgoings against the benchmark.

    The data probably hasn’t been collected with this use in mind, however. And that’s the pitfall for any projects driven by social data.

    So of course I outspend my peers on education….but that’s because my child attends an independent school with steep fees. What would be truly interesting is to know what my peers are spending on their own adult education and continuing development. But with just 11 categories, this granularity is lost. Or how my personal savings rate compares with other 1-child families who like ours rely on private schooling.

    The biggest omission is that a single category is used for savings and personal pensions. Britain is undergoing a slow revolution with regards to provision post-working age…and saving for next summer’s getaway is very different than long-term savings. Having visibility on how peers are saving would be truly useful.

    As it stands, the app throws up comparisons of monthly spends on groceries and electronics that make me stop and think. Thanks to app, I notice that I categorise “books” under “entertainment” and meals out in the neighbourhood under “groceries”. I wonder where smokers allocate cigarettes (which cost the sterling equivalent of $13 per pack of 20)?

    More importantly, since I filled in the survey I’ve also started thinking about how much better our household has done by shopping more often and not over-buying fresh food that then spoils.

    This suggests that benchmarking about spending can lead to some self-reflection about behaviours. But it’s hard to see why Compare My Spend would be worth regular visits or what the newsletter would offer me. Because what actually interests me about my personal financial situation is not what the aggregated data can show me.

    Takeaways:

  • The underlying data set’s structure will limit the questions you can ask
  • Just because you can measure it, doesn’t mean it’s meaningful.
  • Twist:

    The benchmarking impulse provokes my rebellious spirit. “Contrast me, don’t compare me!” says a voice in my head. “I’m unique!” it insists.

    The category of telematics insurance honours this uniqueness. Telematics captures actual driving behaviours and uses them to evaluate a driver’s actual risk, rather than relying on demographic identity attributes like age.

    Insurance aggregator Confused.com explains telematics in this video (click the title to play):

    Video: What is telematics or black box insurance?

    The logic underpinning telematics

  • Measuring actual car usage is straightforward, thanks to new technology.
  • Given this, behavioural attributes, not just identity attributes, can be used to segment drivers into risk bands.
  • Risk bands – as ever – determine pricing bands. But risk bands are enriched with behavioural data.
  • In this way, telematics combats prejudice and honours an individual’s initiative.
  • A truly usefully contrast

    Inspired by telematics and disappointed by “Compare My Spend” I’ve just dreamt up a “contrast” app that would cut through the noise, with a shot at tangibly change consumer borrowing behaviours. My idea has nothing to do with social data, and everything to do with transparency in lending and personal financial education.

    The idea hones in on the comparison between the price tag on any item/experience purchased, and the actual costs to the purchaser using a credit card.

    The comparison is dynamic, because it will vary according to time. And that’s where it gets interesting.

    The idea is an app that takes a feed from my credit cards, and transaction by transaction, shows me what a purchase is actually costing me depending on the APR and the length of time it took me to pay of any outstanding balance. (If you read the fine print on your credit card agreements, you may find that up to 56 days after purchase are interest-free.)

    Lenders have every reason to conceal this information…and consumers would have at their fingertips relevant insight worth considering before the next purchase decision. The app’s insights would also put the consumer in a stronger position to “stooz” – the slang name for transferring balances from a rate-bearing credit card to another credit card offering a 0% balance transfer.

    3. Crowdfunding to measure demand

    Crowdfunding allows inventors and product makers to communicate directly with purchasers, leaving out the intermediary relationships with distributors that are time-consuming and expensive to forge. It’s worth considering with the lens of how data drives innovation.

    There are two models in crowdfunding. On platforms like Kickstarter, the goal is make-or-break. Backers are not charged until the funding goal is achieved or exceeded. On platforms like Unbound or Indiegogo, backers part with their cash at the time of committing. So the risk to backers may be greater and funding may take longer.

    Some teams are very wise in how they use the crowdfunding platform to build relationships with backers and end-users. Uno Noteband exemplifies this. The product is a smartwatch that is a “one touch notifier and fitness tracker”. So you can stay aware of incoming emails and calls, social media alerts from your favourite apps, and calendar reminders…without looking at your phone. It also counts your footsteps and records activity levels. The notifier uses Spritz, a speed-reading technology delivering text at accelerated rates with the promise that readers quickly learn to comprehend.

    The Indiegogo page is transparent about the components of Uno Noteband (which are off-the shelf) and the multi-disciplinary team who have created it. It also conveys the sense that the product is expanding in step with its backer community:

  • Yes it’s a wearable for gamers (one of the first).
  • No, it’s not for Windows phone or Blackberry users, neither at present nor are there any plans in its future as currently envisaged.
  • Yes, there are clear rewards for backers who help Uno Noteband smash its current campaign goals (with 21 days left at the time I’m writing, backers have already doubled the goal).
  • Yes, there are meaningful choices about product features that backers will be invited to vote upon (in this case, wristband colours).
  • When Abhay Adhikari pointed me to Uno Noteband, he said:

    “Uno Noteband has an iterative, flexible crowdfunding strategy driven by demand and peer to peer recommendations.”

    The strategy works because the Uno Noteband team are using the Indiegogo platform to create genuine relationships with backers. This wouldn’t be possible without a robust user-centered product but it also treats what backer data as the byproduct of a set of meaningful conversations.

    Twist:

    When we circle back to the earlier examples, we see that data is the trace that behaviour or activity leaves behind and from which meaning is made. It’s data as residue. But data gathering can also be a performance, and data a response evoked by masquerade.

    Masquerade happens when a context that exists naturally is hijacked for research purposes.

    The context I’m thinking about is a Japanese crowdfunding site called Makuake. On Makuake, a novel watch made with electronic paper like that used in Amazon’s Kindle e-readers and the Pebble smartwatch appeared in September, 2014. The FES Watch was offered for about USD $167, and quickly surpassed its funding goal although the startup had no track record. That’s a familiar crowdfunding story.

    It turns out that, while the FES Watch is real, there is no startup. The project owner is Sony. No official statement from Sony has yet been reported. The Wall Street Journal quotes someone from Sony who is connected to the project:

    “We hid Sony’s name because we wanted to test the real value of the product, whether there will be demand for our concept.”

    Researchers have long used blind tests by target consumers to test product attributes. Savvy NPD facilitators may isolate product attributes and test each attribute with consumers. But thanks to open digital platforms where inventors, manufacturers and end users can interact, the research is now happening not simply “in vitro” but “in vivo”.

    For backers, part of the fun of crowdfunding is the gamble. Will the project you backed achieve its funding goal? Will the product delight as much as the teaser? It’s a chance to spot talent, and invest. Corporations can flop as ingloriously as any startup. But they’re not hives of as yet-undiscovered talent. They’re Hollywood studio machines not Sundance.

    Sony FES Watch crowdfunding campaign shows a goliath stepping into a territory made visible by mavericks, upstarts and renegades because crowdfunding offers a way to gain working capital and a customer base in one set of steps. The up-and-coming generation of makers needs crowdfunding and it’s becoming an important dimension for their survival when working capital from banks isn’t easy to access. A marketplace for ideas and possibilities that depends on trust risks being trivialised by the likes of Sony.

    Sony went into stealth as an experiment. The question was: what kind of demand can we stimulate outside the glow of our massive brand halo? The question at the heart of the experiment was not: what kind of conversations with backers or community might we build? If the other data stories offered different ways to visualise the relationship of purchaser to the e-shop experience, or the consumer’s identity (in relation to peers), this story is a one-way mirror….through which the corporation watches us.

    Sony's FES Watch Crowdfunding campaign was like a one-way mirror

    Conclusion

    These commercial stories range widely, from Britain to America’s Pacific Northwest to Japan. Learning about these projects has helped me sharpen some core beliefs. I leave you with these principles:

  • People matter more than numbers. Curiosity about people should underpin every data project.
  • Data must illuminate people’s experience, not overshadow it.
  • Great innovation isn’t constrained by rigid data structures.
  • The agenda for data-driven innovation matters. “Is it self-serving or generous?” is a great question for evaluating concepts before they become products, services or campaigns.
  • Acknowledgements

    Thanks to:
    Abhay Adhikari (@golpaldass) for suggesting the Uno Noteband story.
    John J Sills (@johnjsills) for suggesting the Compare My Spend app.
    Alastair Somerville (@acuity_design) for suggesting the Sony story.
    Matt Taylor (@mdtaylor) for giving me the link to Jared Spool’s “$300 Million Dollar Button” post and suggesting telematics insurance.

    Thanks also to Indy Neogy (@Indy_Neogy), Gregg Fraley (@greggfraley) and Abhay Adhikari (@gopaldass) for thoughts on this draft.

    image credit: youtube.com

    Follow @ixchat on twitter

    Wait! Before you go!

    Choose how you want the latest innovation content delivered to you:


    Kate Hammer is a joint founder of KILN, working with large-scale companies in the USA and Australia to transform their internal innovation processes. Kate works as a business storyteller. In 2012, she created StoryFORMs to help others articulate their commercial & organisational stories. Kate offers workshops & 1:1 coaching.