What’s NEXT Dinner Series: How to be a force for good in technology
Purpose-driven technology has become a huge focus in tech for the past year given the rise of privacy/data breaches and adverse effects of persuasive technology. With the rise of social enterprise startups, B-corps, and humane tech advocacy groups, the tech industry is at an inflection point. After seeing the negative externalities that came out of Web 1.0 and Web 2.0 business models, many in the industry are seeking to course-correct and find new ways of building and innovating mindfully.
To explore this topic further, we convened a group of builders and founders from Human Ventures, Condé Nast, Motivote, Hire Tech Ladies, Breathing.ai, The Village, Juxtapose, Enigma, and Deed for our fourth What’s NEXT dinner on Tech for Good.
A big culprit for the internet’s myopic focus on growth and data collection stems from the adoption of the “time spent” metric. Large advertising and social media platforms adopted the metric to measure user attention, which created an incentive to optimize their products to keep users engaged.
While this metric is an indicator of usage and attention, it does not accurately measure an app or service’s value to the user or the quality of a user’s experience. Did the user learn something new? Do they feel closer to someone (or the world at large) after using the product? Was the content they viewed useful in any way?
Could new metrics do a better job of measuring user value?
The overreliance on time spent as the defining metric for so many businesses created a ruthless optimization that did not consider the consequences of consumer health or product integrity. However, what you measure, you can change.
Ryan Cooley from Human Ventures shared a story about building a group live-streaming app, which found that greater emotional discomfort in users correlated with longer viewing sessions. That is, the more emotional tension users felt, the longer they spent using the app. The product ended up shutting down in part because growth stalled while trying to optimize for more usage. Ryan warned about how important incentives are to an organization because that becomes the North Star for everyone.
We asked, “What new metrics could replace time spent?” But finding new metrics to supplant time spent is a challenge because each product is designed to do something different. The group agreed that moving away from engagement metrics is a start. Defining metrics specific to a startup’s core business model is a better way to measure actual value delivered to users.
Breathing.ai CEO Hannes Bend recommended measuring psychosocial factors in digital consumption affecting a user’s subconscious and wellbeing. His company personalizes the UX/UI of screens and audio devices using patented “adaptive interfaces” that are designed to improve existing screen time by individualizing the user experience within the applications used. For example, the colors and fonts on a screen can change due to a user’s neurological activity, which is detected via web/smartphone camera and AI.
Hannes believes that understanding the impact of time spent on devices on user physiology can lead to more positive behavioral changes and the design of more humane digital interfaces. By informing users of the health effects of their devices, Breathing.ai hopes to center device usage around the user’s wellbeing.
Jun Harada, who is working on digital strategy & innovation at Condé Nast, is bullish on using “loyalty” as a new KPI to measure subscription and readership. Condé Nast’s digital innovation team is still early in measuring loyalty in users and is experimenting with new metrics. The team is exploring longitudinal data analysis of its audience and the frequency of site visits over a time horizon as long as 10–15 years.
Who and what will get surveillance capitalism to change?
Many at the dinner were skeptical that regulation could keep up with the speed of technology. Others who had worked for big tech companies doubted the efficacy of employee advocates making an impact against corporate management.
Rachel Konowitz, who works on a voter engagement platform called Motivote, believes that industry change for more humane technology practices will come out of a collaboration between regulators and companies. Rachel pointed out that perhaps the catalyst for change will rest on a combination of consumers, nonprofits, ethicists, and lawyers rallying together, pushing for activism and change outside of tech companies.
Michael Brent, Data Ethics Officer at Enigma, has founded a grassroots organization called the Technology Ethics Coalition (TEC), which connects people working in the local tech industry. They address the social, political, economic, and moral issues related to the current and future state of our industry.
Their ambitious goal is to protect individuals and communities from the harms caused by the uncritical use of data and machine learning technologies. They seek to achieve this through building an educated and engaged base of coalition members comprised mainly of local tech workers, fostering industry adoption of regulatory frameworks that protect companies from ethical risk, and by playing an integral role in the inevitable development of future legislation regulating the tech industry.
It’s easy to be pessimistic about how things that are broken in tech today, but this group of passionate change-makers is hoping to make an impact on the way the technology industry defines success. Today, creating a success is no longer based on blind optimization for a single metric.
It will be a challenge to address the industry’s metrics, incentives, and old habits, but the future of how technology will evolve depends on us solving the unforeseen consequences of how it’s measured.
One of the basic tenets of building technology is iterative learning, which means learning from our past mistakes. What we’ve learned from the consequences of blind optimization for engagement and growth will inform how we build a new generation of technology that’s more humane, mindful, and truly willing to put the user first.
