Against growth: Building a slower, more thoughtful internet

Unlearning growth-by-default like our lives depend on it

Growth is a dangerous ideology. It’s one that’s taken hold across the globe — in our financial systems, in business, in the internet’s architecture, and in software development. Endless growth is not possible, or desirable, and it’s killing us — emissions related to the internet and its data centers account for 3.7% of global green house emissions, as much as that of the airline industry. In today’s short post, I’ll unpack the idea of growth, offer a few alternatives for a more sustainable internet, and end with a note on growth and social impact.

The fruit of the tree of knowledge? We ate that up
Growth’s original sin starts is the First Industrial Revolution in the 1760s, when mankind discovered large-scale manufacturing. We added in fossil fuels with the second IR (1850–1930) and globalization with the third (1950s and onward). And here’s how that played out for the planet:

At the same time, the global financial system developed and took root, doling out debt and resulting in companies ballooning beyond britches and borders. (Okay, no more alliteration.)

Fast forward to 2021 and every tech company fashions itself a VC-fueled rocket ship on the trajectory of exponential, hockey-stick growth. Publicly listed tech companies are basically required to demonstrate continuous growth as they metastasize around the globe. Companies that miss analyst targets or don’t show enough user or revenue growth see drops in their share prices. Hence the pressure to juice.

The global content platforms provide a cautionary tale. Platforms default to yes: yes to new users, new content, new groups, more engagement. It’s this dynamic that means Youtube can’t possibly moderate all of its videos for radicalization, or Facebook can’t moderate all of its groups for terrorism, or Twitter can’t prevent harassment. This underpins everything else in the stack.

Alternatives to growth
Growth-by-default is so ubiquitous that it’s hard to imagine an internet without it. Below are some alternatives for slower, more thoughtful products. It’s a bit of a hodgepodge, spanning business strategy to product features

1/ Slow content creation (Dispo). One way to hinder growth is to throttle content creation. Dispo is the hot new photo app that aims to make phototaking a little more analog. The idea is that you take photos but have to wait for them to develop at 9am the following day, at which point you receive a batch of photos that filtered to reproduce the oversaturated effect of disposable film. It’s a gimmick, but an interesting one. After using it for a week, I can tell you that I’ve found a little more joy and serendipity shooting photos with my phone (which I otherwise do all the time). I’m not clogging up my phone’s storage with a million dupes. And I feel less beholden to instant gratification and put my phone away instead of review photos or posting them. Designers have long reviled skeuomorphism — when a digital product is meant to look analog, like the lined paper in the Notes app — but in this case the skeuomorphic dynamic feels like a productive direction. More digital should feel analog.

2/ Limit virality and engagement (WhatsApp). Growth-powered dynamics like virality — and features like Facebook’s Share button and Twitter’s Retweet — fuel harms like misinformation. WhatsApp recently limited the times a message can be forwarded, which “cut forwarding by 25% globally and ‘highly forwarded messages’ by 70%.”

3/ Distribute decision-making (Wikipedia). Wikipedia offers a nice solution to the growth problem by distributing most of its decision-making. It empowers its community to make all content and also to moderate it — which a lot more democratic than having a twentysomething doing that from Palo Alto. It’s not a perfect system, as Wikipedia can inadvertently reproduces societal biases like underrepresentation of women. It’s worth reading this piece from Wikipedia’s CEO, Katherine Maher, on lessons from their first 20 years.

4/ Design for breaks and breakups (Hinge). Many products are designed with the assumption that its users can and will hang around forever — and badgers their users to come back. What happens more often is people use a service for as long as it is valuable and then move. Postlight’s Liz Tan looks at Hinge, which markets itself as “the dating app designed to be deleted.” She writes about how more services being designed with an end in sight, to optimize for “good churn.”

5/ Build a sustainable business (Mailchimp). You can build a lemonade stand and sell lemonade to 100 people every day. You don’t need to double by Wednesday, franchise by Thursday, and IPO on the Lord’s day. Mailchimp provides a nice counter to the typical Silicon Valley narrative. In this great piece, Farhad Majoo notes: “You can simply start a business, run it to serve your customers, and forget about outside investors and growth at any cost.”

Growth and social impact
I have obvious mixed feelings on this topic. While growth can be problematic, it isn’t all bad, especially for tech nonprofits. Nonprofits want to have an impact on as many lives as possible, and they can learn a lot (for free) by watching big companies scale. It’s something I think and write about a lot.

Perhaps the secret is moderation — to grow mindfully and to own the impact that you do have. Small numbers can be really powerful for a nonprofit. At Crisis Text Line, we power 3,000 conversations a day. That doesn’t sound like much in the tech world — people get that many pageviews on LinkedIn posts. But each of those conversations is a life saved.

It’s important to remember that as a mission-driven organization, your outputs are more nuanced than a share, retweet, or pageview. They represent some sort of unique, sometimes life-changing impact — an intervention that isn’t provided by something in App Store.

So my final advice is to combat the ideology of growth by embracing the small. Think about the lives impacted and lead with stories rather than big-sounding data. Data is human behavior at scale, but each data point is a person. Here’s feedback from one of those 3,000 texters we served one day last week:

For weeks now, I’ve thought about ending my life, recently I thought up the easiest way for me to go. A measly bottle of pills and the consistent amount of time my parents leave me alone at home. I was very close to my end.. This text line was my last resort… I may not know what you look like, or the sound of your voice. But I know you’ve got a big heart and a beautiful soul. Thank you.

— XML

— — — — — — — — — — — — — —

Links this week

  • Algorithm auditing is going big business. Ironically, the industry is as much a black box as the algorithms it aims to understand. (The Markup) 📚
  • The racist roots of the Body Mass Index (Good Housekeeping) 📚
  • Clubhouse creates space for the taboo. How Chinese users use the app as an opportunity for dialogue on the Uighur genocide. The New Yorker’s Jiayang Fan in conversation w/ David Remnick (The New Yorker podcast) 🎧
  • La Brega, On The Media’s first bilingual podcast, dives into Puerto Rico’s history (WNYC) 🎧
  • Product Strategy resources: a dive into the “product strategy stack” (Reforge 📚) and a hands-on case study of Netflix’s 2020 product strategy (Industry Product Conference 📺)

Product Director @ Crisis Text Line