Observations from Segment Synapse 2019

posted: September 28, 2019

tl;dr: Like most technologies, analytics is not inherently good or evil - it depends on how it is used...

I was fortunate to be sent by my employer, meltmedia, to attend Segment Synapse 2019 in San Francisco, my first tech conference in the city since GraphConnect SF 2016. This was the second annual user conference for a growing startup company, Segment, that develops SaaS solutions for collecting user analytics across a variety of digital properties: websites, mobile applications, emails, digital ads, etc. Segment’s solutions make it easier for companies to get data from user activity on their digital properties, and to send those data to a wide variety of downstream tools and storage systems, where the data can be acted upon and analyzed. In this post I’ll share some of my thoughts about user analytics, which were brought to the forefront by this conference.

Segment sits squarely in the midst of various political debates and regulatory activities concerning user privacy on the Internet. There is much heated rhetoric in this area, with some people using the loaded term “surveillance capitalism”. Segment, however, is a much different company than Google, Facebook, Amazon, or Apple.

Segment’s products cost money and are sold to companies. Unlike some of the aforementioned companies, Segment doesn’t give away or sell low cost consumer products for the purpose of collecting data on users. Segment does not store user data other than very briefly, for the purpose of reliably forwarding it to other end destinations. In fact, Segment can be used to collect data and keep them within the control of the company whose digital properties the user is actually using. If desired, it can help replace a very popular free service, Google Analytics, which many websites use to collect user analytics. With Google Analytics, the website owners get to see reports on user behavior and digital ad effectiveness, but the data end up in Google’s databases.

Segment Synapse 2019 day two keynote address by Fatemeh Khatibloo of Forrester Research

There’s a lot of bad press in this area, and far be it from me to try to defend what happened between Facebook and Cambridge Analytica. I also have had my data leak to third parties with whom I’ve never transacted business. I get constantly bombarded with fraudulent emails and phishing scams from senders claiming to be a bank, a government agency, or a major retailer, but whose email addresses actually are or, to cite just the two most recent examples from my junk mail folder. I get so many scam phone calls that I don’t answer my phone unless it is a number already in my contact database. I often wonder why it is so hard to bring these obvious scammers to justice.

Yet, on the flip side, there is a real desire that I have, and which most people have, to receive personalized, relevant content. I found my last two jobs thanks to personalized emails. When it was time to start looking at new opportunities, I went to a legitimate job posting site, LinkedIn, and set up some filters that would find the kinds of jobs I was interested in, in the locations that I wanted to be. I configured LinkedIn to send me regular emails anytime their software found job postings that matched my filters. That is how I first learned of the two most recent jobs that I’ve taken. Certainly there was much more that was necessary to do after I received the personalized email, but the email was indeed the very first step towards a new job. This is far from an atypical situation: there are many situations where I and other consumers relish receiving personalized content that is tailored to our expressed desires.

This tension between privacy, including the desire to remain anonymous at times, and the desire to receive tailored content, was called the “Privacy-Personalization Paradox” by several speakers at Segment Synapse 2019. There are ways to ensure privacy and to give users more control over their data, while at the same time allowing that data to be used for legitimate purposes as directed by the users. Many of the problems that make the news happen when data leak outside the boundaries that the user has agreed to, although some of the problems happen when users aren’t clearly told how their data might be used.

Collecting user data is not inherently good or evil; it depends on how the data are used. In this respect it is like many technologies: the good or evil comes from how it is used. A gun can be used to stop a mass murderer or to commit mass murder. A Macbook can be used to create music that brings joy to millions, or it can be used to launch a DDoS attack that inconveniences millions.

At meltmedia we most definitely are not a shady offshore outfit trying to sell knock-off Viagra or to bilk elderly people out of their life savings via a phishing scam. We actually help our clients comply with privacy regulations and best practices, and help them implement opt-out solutions for their email campaigns. Segment makes tools that legitimate companies can use to present their users with more relevant content, and I believe that the company will continue to grow and succeed.

As for San Francisco itself, it seems like every time I return there are even more troubled people in the streets. I don’t know which cases are the saddest: the older burnouts who appear to have been living on the streets for years, or the younger people who appear to have already thrown away their lives for the temporary solace provided by drugs or alcohol. Sometimes I’m not sure whether the person I’m stepping around on the sidewalk is alive or dead. San Francisco used to be one of my favorite cities to visit, but I can no longer say that.