Welcome to Brilliant Reads. This week we have two stories on privacy and marketing, lessons we can all learn about product development from Google+, and thinking on how to make commenting a more productive process.
Can you hide from big data?
In this interview with Jessica Goldstein for Think Progress, Janet Vertesi, assistant professor of sociology at Princeton University, explains how and why she tried to hide her pregnancy from big data.
Pregnant women are a marketing gold mine; their data is highly sought and worth 15 times as much as the average person’s. Out of professional curiosity, Janet decided to see if it was possible to keep the fact that she was pregnant from marketers.
She and her husband used a traceless browser for researching anything to do with the pregnancy or baby. They bought baby things in real life using cash, or online using a fake email, gift cards paid for with cash and then had them delivered to a mail locker. They avoided discussing the pregnancy in email and on social networks, and asked friends and family to do the same.
The process was difficult: it led to extra work; it meant they had to pay more for baby products; friends and family struggled to grasp the concept that even private email and Facebook messages would be tracked; and Janet was worried that the kind of subterfuge she had to engage in would make it look like she was actually involved in criminal activity.
In this article for the Guardian, Luke Dormehl looks at the latest developments in facial recognition technology and the ethical issues arising as result.
Dormehl highlights several new innovations in facial recognition technology: Facebook has created a tool almost as accurate as the human brain when it comes to recognising people in photographs; Apple has investigated the possibility of using facial recognition as a security measure for unlocking its devices; the NameTag app lets users identify people they photograph; Tesco plans to install video screens at checkouts to capture the age and gender of customers; and a US startup called Emotient is developing a generation of TVs that monitor facial expressions and viewer engagement.
These kinds of developments have huge implications for privacy. While privacy is an issue with every form of data mining, a lot of information online is anonymised. Facial recognition meanwhile is precisely the opposite. In addition, because facial recognition takes place in public spaces, it’s not necessary for the person being surveilled to actively ‘opt in’ to the service.
A further issues lies in the fact that while facial recognition algorithms may be neutral themselves, the databases they are tied to are anything but. Whether a database concerns criminal suspects or first-class travellers, they are still designed to sort people into categorisable groups, and have to make judgements based on appearances alone.
This Journalism.co.uk article shares three ideas from ProPublica’s senior engagement editor, Amanda Zamora, about how to make online comments a better experience for readers and content creators alike.
Comments, the murky ‘bottom half of the internet’, don’t have a great reputation for interesting or reasoned debate. Amanda suggests that the commenting process could be made more productive by thinking about the following three factors:
- Design – if you’re trying to elevate the level of discussion, think about what you want readers to contribute and design your commenting process accordingly. Rating or upvoting comments can be beneficial, as it provides an incentive for people to be thoughtful in their contribution.
- Context – contextualising comments by letting readers annotate stories with their reactions or thoughts (like Medium) is another way to encourage thoughtful, relevant commenting.
- Structure – creating a system where comments and contributions (like polls or surveys) feed directly back into the story means the commenting experience takes on a more positive and productive role for the writer as well as the reader.
Writing for The Next Web, Paul Adams, who used to work on Google+, suggests some broader product lessons we can learn from the inception, launch and evolution of Google+.
- Build around people problems, not company problems – Google+ was built to solve a problem of Google’s, not a need of the user.
- Perceived benefits need to be greater than perceived effort – Circles might have been a useful feature, but people won’t use something if it seems like the effort (assigning every single contact to multiple Circles) outweighs the benefit.
- Be patient – other than outliers like Instagram, social networks take a long time to build out and deepen, just as real life relationships do.
- A fast follow product strategy doesn’t work when you have network effects – with a social network, people want to be where their friends are. A fast follow strategy where you copy a competitor and improve on their functionality won’t work, because the network effects take time to build.
BrightonSEO: crawl space, semantic search and Turkers
In this post, Digby from Brilliant Noise provides a write up from the recent BrightonSEO conference.
Rather than the common ‘SEO is dead’ cliches that often dominates at industry conferences, Digby found some fascinating talks on search strategy and SEO’s role in media.