Thu, Aug 8, 2019

A Retrospective on Structural Shifts in Consumer Internet

I think a lot about the history of the consumer internet industry. The big structural shift of 1990s was the internet. The trick to finding startup ideas and getting funded in this era was to think of something you do on a daily or weekly basis, then prefix it by “online”, and voila, you’ve got a startup idea! What were some consumer ideas that emerged on the internet in the 90s, and the defining companies that leveraged them?

  • Web browser. Netscape.

  • Personalized “PUSH” of media. PointCast.

  • Money transfer. PayPal.

  • Free email. Hotmail.

  • Directories/yellow pages. Yahoo!, WhoWere.

  • Radio. Broadcast.com.

  • Mailing lists. eGroups.com.

  • Instant messaging. ICQ/AOL.

  • Personal websites. Geocities.

  • Shopping. eBay.

  • Search. Google, Altavista.

  • Maps. Mapquest.

  • Classifieds. Craigslist.

  • Online Dating. Match.com.

  • Q&A. AskJeeves.

  • Books. Amazon.

2000s

Now, fast-forward past Y2K and the dot-com bust. The 2000s. The structural shifts of the early 2000s were slightly more nuanced. These shifts were leveraged into startup ideas through a new set of prefixes, such as social, micro, free, unlimited, personalized, cloud, and sometimes, even Ajax. The trick to come up with decent startup ideas in this era was to add one of those prefixes to something people did. Here are some ideas from that era, along with the defining companies that rode the wave (and, occasionally, even created the wave):

  • Blogging. Blogger & LiveJournal.

  • Social bookmarking. del.icio.us.

  • VoIP. Skype.

  • Personalized internet radio. Pandora.

  • Social resume. LinkedIn.

  • Cloud storage. Dropbox.

  • Photo sharing. Flickr.

  • Maps. Google Maps.

  • Location-based social networking. Dodgeball, Gowalla.

  • Unlimited, free e-mail. GMail.

  • Social events. upcoming.org.

  • Social networking. Facebook.

  • Social yellow pages. Yelp.

  • Video streaming. YouTube.

  • Social blogging. Twitter.

  • Mobile messaging. BlackBerry Messenger.

  • Social Q&A. Quora.

  • Forums. Reddit.

  • Movies. Netflix.

  • Recommendations. Hunch.

  • Web roulette. StumbleUpon.

  • RSS. Feedburner.

  • Social gaming. Zynga.

  • Smart home. Nest.

  • Social commerce. Wish.com.

  • Social coupons. Groupon.

  • Social city exploration. Foursquare.

  • Handmade: Etsy.

2010s

After a scintillating launch of the revolutionary iPhone in 2007 and its App Store in 2008, mobile became the big structural shift of the following era. Convert anything from online to mobile, and you’ve got suitors for your seed round. What was even more interesting in 2010s was that the combination of prefixes led to even more startup ideas. Some popular ones other than mobile: social+photo, social+video, curated, share, on-demand, live, ephemeral, and voice-enabled. Some big ideas from the 2010s and the defining companies:

  • Sharing rooms. Airbnb.

  • Social photo bookmarking. Pinterest.

  • Photo sharing. Instagram.

  • Taxi. Uber.

  • Photo messaging. Snapchat.

  • Live videos. Periscope.

  • Dating. Tinder.

  • Groceries. Instacart.

  • Micro-videos. Vine.

  • Classifieds. Letgo, Offerup.

  • On-demand delivery. Postmates.

  • Group video chat. Houseparty.

  • Fiction. Wattpad, Hooked.

  • Astrology. Co-star.

  • Cross-platform messaging. WhatsApp.

  • Music. Spotify.

  • Language learning. Duolingo.

Why even do this?

Certainly, lessons can be learned from the successes and failures of dot-com companies of the late 90s and 2000s. But, can industry trends be predicted and secrets be unlocked by looking at companies of the past? Can these insights then help in making sense of the present and predict the future? The bottom line is that no one can predict the future, but one can certainly make calculated guesses, instead of just plain speculating. I’m starting to believe that fundamental human needs change very occasionally. But every decade or so, new ways emerge to fulfill that need. That small period where it is the right time to fulfill a need is the most ripe for incubating companies.

Despite best efforts (Webvan), you couldn’t make online grocery delivery work in late 90s but you could in 2013 (Instacart). You couldn’t make a social networking website work in 1996 or 2000 but you could in 2003 (Facebook). Airbnb wouldn’t work in 2005 but it did in 2010. “PUSH” didn’t work in 1999 but maybe it is coming together now. Generally, the only good time really to start a valuable company around a wave is that small period. In that small period, you need to act fast and get the execution right. You can make mistakes but you have to make them really fast to get to the right answer. The exciting thing is that in fast-growing economies, you’re more or less always living in that “small period” of a wave that is ripe for the birth of some company or set of companies.

The right time for an idea

The key thing for any entrepreneur who’s still at square one (either starting for the first time or after one or more failures) is to figure out which idea’s small period is happening right now or about to happen in a matter of months. I believe one way to do that is to identify: current trends in consumer behaviour, unmet needs, failing companies, and also, have a basic understanding of all the recent progress going on in a relevant technology. Looking at the past decades of the industry could give you hints.

If you already have an idea, then determining whether you’ve got the right idea or not while you’re in or approaching a wave is tricky. Benedict Evans suggests to think in terms of roadmaps for evaluating whether a technology can become universally accessible and ubiquitous:

In generational shifts, you force the new tool to fit the old workflow, and then the new tool creates a new workflow. … Word replaced typewriters and email replaced Word. … What matters is seeing the value of the capability, not predicting particular applications. … “Everything looks like a toy”, “That is a toy.”, “no one will want that” and “no one wanted want phones either” are both completely true and “not even wrong”. They have no predictive power. … Ask whether there is a technology roadmap or whether this is a superpower — this will give you a roadmap to understanding what might happen.

Yet, you have to remember certain realities, such as the one Walt Mossberg’s final column touched upon:

The big software revolutions, like cloud computing, search engines and social networks are also still growing and improving but have become largely established.

Another observation that M.G. Siegler described well in the context of social networking products was around the concept of “niche networks”. I feel it could apply generally to more categories of ideas, within consumer or enterprise internet. It goes like this: you start by fulfilling a particular niche — for e.g. it could be a website/app that lists gyms around New York and lets them advertise pay-per-click/call. Then, you build a website that lists flower shops. Then, leather jacket stores. You scale this to all kinds of independent but linked websites. $100mn ARR. Then, you build a product that caters to the entire gamut of advertisers on the combined network. $1bn. IPO.

Build something specific

Almost always, successful products first cater to small markets and barely look like products. They just scale rapidly and possibly reinvent their business on the way. All things said, the stupidest thing you could do is “be a visionary” and sit and try to think of an all-encompassing product that could become a billion dollar company. It just doesn’t work like that. The correct way to do it is to build/hack something specific, see if there’s any demand, build a business model, then scale out fast. Not all hacks have lucrative paths out, but that’s where the “vision” thing comes in to play.