Tech Club Summer Stories
We asked a few interns "How's your summer going?"
Here's what they have to say...
Latest updates coming soon!
Follow Darden Tech Club on Instagram to stay updated
Alyssa Apolonia, '17
|
Jimmy Figueroa, '17
|
Linden Schult '17
|
Peter Bergen, '17
|
Making Data Meaningful Through Socialization
This summer I'm a product marketing intern on the Office 365 team, focusing on assessing "market fit" for a new data analytics product called Delve Analytics (DA). DA has really gotten me thinking about self-quantification and what people actually do with data, so instead of talking about how my internship is going I wanted to pour some thoughts out on this topic. I've been working in tech and data analytics for over six years now and do lots of self-quantification so feel pretty well qualified to do this. Companies are pumping out products that track and quantify everything. Wearables are all the rage, and I can now track things like sleep quality and heart rate variability without needing a doctor. Delve Analytics tracks what you spend your time on at work (more on this later). You can even get analytics on your dog now (the screenshot below is from a company called Voyce). |
|
I often wonder how far this craze will go. What is my body's max threshold for Chipotle in any given week? What is the optimal response time to my mother's phone calls so that she doesn't worry too much? Which angle in my OkCupid profile pic gets me the most likes? More importantly, how can I do anything at all with all this effing data to actually make my life better?
One way to make data actionable is to set thresholds, i.e. some level above or below which you need to start worrying. At Whereoware, the digital marketing company I spent 4 years at before Darden, we sent a lot of emails out on behalf of our clients (like millions per day). Periodically some bad email addresses could get into our lists, and this could kill our deliverability rating. So naturally it was pretty important to track this type of deliverability data regularly, which the dashboard you see below helped us do. Over time we figured out that a deliverability score below 90 would indicate that we need to fix something lest we get placed on a dreaded “blacklist.” Without that threshold, we'd have no standard for when and how to take action, and the data would have been meaningless. |
|
The challenge is knowing what to use as your thresholds. There are some scenarios in which thresholds are clear-cut, like the deliverability case above. In other cases this is not so straightforward. For example, I track my sleep with my Microsoft Band and it tells me I wake up around 10x per night on average. Is there a certain specific number above which I should freak out and run to a doctor? (If so, actually let me know, because 10x does feel kind of high).
|
|
Socialization is one way to make data meaningful in the absence of simple absolute thresholds. The basic idea is that you take some metric and compare it with a relevant group of peers (note the use of the word "relevant"). These comparisons can help companies set benchmarks and tell stories to themselves and their stakeholders about how "average" or "not so average" they are. These types of metrics won't tell you how to achieve "above average" status, but human psychology seems to inspire us to action in more immediate and substantial ways than any other type of metric I've encountered.
At Whereoware we were constantly pressed by our clients to provide industry averages on website and email activity. Fortunately our marketing automation vendor, Silverpop, published these averages every year, and it proved extremely helpful: for clients that had far below-average open rates, for example, we were able to upsell subject line A/B testing or distribution list cleanups. We also, of course, use industry averages all the time in finance - "comps" are just industry averages, and the art is all in picking the right companies to include in the average. |
|
But socialization is even more powerful amongst individuals. Opower, for example, is a utilities technology company that tracks how much energy you use, then shows you how you stack up against your neighbors. They've found consistently that people are far more likely to reduce their energy consumption when they see how much more they use than their neighbors (read more on this here).
|
|
Omada Health provides another great example of socialization, this time in the field of chronic health management. Omada worked with the NIH to conduct research on reducing risk factors for pre-diabetics. They found that, through a combination of smart wearables and online social communities, individuals could achieve incredible health outcomes - in this case, an average 4.7% weight loss after one year in the program, a sustained drop in markers of blood sugar levels, and an unusually high 80% program completion rate (more on this here). Social groups were a fundamental element of these results. Each participant was paired with an online peer group based on age, BMI, and location. These groups were formed to maximize similarity amongst group members, again demonstrating the importance of 'relevance' when choosing comparable external metrics.
|