The Era of Consumer Customization Is Here
Localization, Machine Learning and Video Converge In Apps
Instagram is turning into a TikTok experience. Well, most consumer experiences are these days. But what does that mean?
For one, machine learning becomes central to a product experience. Secondly, a dynamic interface can change according to the user's interests.
Both are hard to do, especially where videos are involved. And that means a higher cost to build such experiences (servers, expertise, maintenance, and testing), but also forms a "moat" around an existing business.
Why Is This Happening?
Trends Fade, and New Ones Emerge
There are also fads, but they disappear quite quickly. Trends are strong over time and stick around for a decent amount of time. From the perspective of consumer apps, they have to do with behavior changes and demands for certain experiences.
The use of mobile has increased even in lower-penetrated countries, and lockdowns have led people to seek new forms of entertainment. Let’s split the discussion solely by trends first, before circling back on environmental factors leading to increased usage.
Cheaper and Faster Data
Worldwide, mobile 4G is widespread. In Singapore, 20GB deals going for $20 being are quite common.
If you had a slow home internet connection before, you can simply switch to mobile and get a faster, and sometimes more stable experience.
This has increased mobile usage in many respects and will continue to shape the experiences being delivered to your personal devices, internet connectivity is no longer a limiting factor for users to enjoy new apps.
On-demand Entertainment
Hubspot, since 2018, says that customers expect companies to understand their needs. Being as tailored an experience, and customizing each interaction point goes a long way - not many companies can do this.
Companies need to have enough data before they have a machine learning layer to correct provide said experiences, and TikTok is a great example of that.
The “Swipe” is the least-energetic action you can do, and one of the highest dopamine-inducing exchanges for a human. That makes it a strong habit-forming interaction, one that keeps you hooked and coming back for more.
Prior to TikTok, we had Twitter, Facebook, Instagram. Most people are already spending lots of time here. Replies, comments, and shared stories on these platforms give you some pleasure, but it does take some time for responses to occur.
With TikTok, that’s instant. You get to be entertained anytime, anywhere. Once done, switch back to the 3 “old” platforms, and then switch back. Your dopamine level is, on average, always high that way.
Why is TikTok All That?
To elaborate, let’s use an example as a comparison.
If you remember Vine, one of the earlier short-video clip apps since 2012, this was a trend for a while until Twitter discontinued it. It was quickly copied (Instagram, looking at you), and influencers moved over to other platforms such as YouTube.
Could this be the same direction as TikTok?
I don’t think so, and for reasons to do with technology. TikTok has a vast “library” to feed its machine learning model with. Every user’s experience is hyper-personalized and keeps getting smarter which each swipe. Each video is ranked and categorized, again and again, showing only what the user was interested in, in their history, and what appealed most to them would be products/services users highly considered to make a purchase.
That kind of data is effectively a “technological moat” that is difficult for many other companies to achieve. For one, you need the data, of which you need users, of which you need the user experience, of which you need data… you get the point.
Another example would be Tesla’s self-driving capabilities. It has more than 1.8 billion miles driven autonomously since October 2019. Waymo and GM’s Cruise drove just 1.3 million and 447,000 respectively, so Tesla is leap years ahead.
If Tesla wanted to, they could easily have a self-driving taxi service as a two-sided platform to owners and riders, any time.
Who wouldn’t want to own a Tesla now? Get a compelling “user experience” while driving, cost-savings from a solar-charging grid and essentially “free power”, or extra income from loaning out unused cars to get “taxi income”.
This growth mechanism has a strong Flywheel effect, and also serves as a defensive moat. Once started, the sheer momentum drives them to keep taking more and more users, and eventually, revenue. Users expect at least this standard of experience, otherwise, they would not consider an alternative.
TikTok’s kind of data moat combined with the user experience is hard to beat - it will give ByteDance significant room to experiment with other ideas and business models, and that’s exactly what they are doing.
The difference between them and other growth-stage companies? ByteDance is already profitable, at a whopping $3 billion a year since 2020.
Environments Get Overhauled
Would you be spending as much on E-commerce if there wasn’t a lockdown?
Would you be freely accepting $150 entry fees to the nearest casinos if you weren’t traveling as often?
Short-videos? Sure, when we’re traveling and want some bite-sized experiences in-transit.
Stuck at home for days at a time? Definitely want to binge-watch a series, movies.
How about in both situations? Swipe left or right for funny short videos, stop anytime. (Some don’t, as I wrote earlier here - Tiktok vs Snapchat). If you’re bored with recorded content and want real-time engagement, you can easily do a live stream, or view one.
Getting personal validation has never been so quick, and from the comfort of your home, being entertained.
This trend is likely to stay, as data is becoming cheaper, even with post-covid normal, we would likely see people viewing live streams directly, rather than just watching series on-the-go.
Summary
We’re likely to see variants of Tiktok and Clubhouse come out. Be it from start-ups or established tech players, the blatant copycat approach is another land-and-grab strategy to tap on the new social trend.
Social landscapes are changing by the day, and so will the technology architecture.
Increasingly, sophisticated technologies need to be brought together for consumer experiences to reach a ‘passable’ standard. As such, costs are likely to increase, and only well-backed players can enter.
Thanks for the post! The era of consumer customization is definitely already here, and it is amazing that we have technology that makes it possible. But at the same time, it makes me wonder how far should a service be customized and in which services. For example, when it comes to news, it could be quite dangerous to be exposed to only certain types of information based on my taste, as it may lead to information bias.