There was an interesting article in Insead Knowledge (Jason Davis and Vikas Aggarwal) about how innovation gets copied by imitators. I liked the question it posed:
“In the game of digital innovation, imitators often outrace the original innovators…at the heart of this race is what we call the innovator’s imitation dilemma. This dilemma arises because an innovator’s attempts to innovate by mobilising in-house knowledge can often end up benefitting rivals more than the innovator, particularly when rivals share and propagate knowledge among each other”
“So how can an innovator succeed in the face of imitation by rivals?”
This is not a new question. And it is not limited to just innovation. Innovation, R&D, and a company’s learning curve can all be undercut by leaks to competitors, also called the spill-over effect. And this can happen even faster in digital businesses, which are often just information. So here is my question:
Can YouTube, Facebook, and others successfully copy TikTok’s success in short videos?
Because they are all trying right now.
I’ll give you my answer. But first a bit of theory.
Innovation and Learning Curve vs. Leaks (i.e., Spill-Over Effect)
I have recently written about how learning and experience curves are an important part of improving operating performance. Especially in manufacturing-intensive industries where there are complicated but repeated activities. Basically, these are about taking advantage of the fact that humans get better the more they do something.
Learning curve benefits also tend to happen more at larger companies, which have greater volume and experience to learn from. So learning and experience effects often get tied together with (or at least coincide with) economies of scale as a source of advantage. This topic was covered in the below podcast and is in the Concept Library.
- Amazon, Tencent and When Rate of Learning Becomes a Competitive Advantage (Tech Strategy – Podcast 131)
Learning and experience curves as concepts also naturally combine with the innovation. Experience by individuals, teams and companies naturally leads to improvements in processes and products by incremental innovation.
The problem, as the article points out, is that innovations and experience-based improvements get copied by competitors. People switch companies and take knowledge with them. Secretes get told. Innovations and other improvements become industry “best practices”, which pretty much every consultant studies. So there is a tension between learning and innovation and leaks, also called the spill-over effect. This all sort of goes together as a topic under operational performance.
Fortunately, strategists like NYU Professor Pankaj Ghemawat have been writing about this subject for decades. Ghemawat says leakages are more likely to happen when:
- Improvements are in products, not processes. These can be more easily reverse-engineered.
- Learning is vested in a small number of employees (such as design teams). They can be more easily hired away.
- “Internal process learnings” are much more well defended against leaks. These are similar to what Hamilton Helmer (7 Powers) calls Process Power. These are lots of small process improvements within a company like Toyota. They are hard to document and replicate.
Note: Another example of hard to replicate improvements are in machine tools and highly specializing machining techniques. Warren Buffett has invested in a couple of these types of companies over the years.
Which brings me to TikTok vs. YouTube (and others).
As I am writing this, TikTok is surging globally. It is in +150 countries and has recently passed 1B monthly active users. That is less than YouTube (2.2B) and Facebook (2.9B), but TikTok is growing fast. And it commands fantastic attention from its users, who watch its short videos an average of 90 minutes per day. Algorithmic-chosen short videos have turned out to be the new killer app. In response, Facebook has launched Reels. YouTube has launched Shorts.
But, per the innovation vs. imitation question, can TikTok actually be copied?
What Does It Mean to Copy TikTok?
Two years ago, the US government almost forced TikTok to sell its US business to Microsoft (and Oracle). But what exactly would they have been selling?
- The users and their viewing history?
- The content creators and their uploaded videos?
- The US and other non-China staff?
- The core software for the app and servers?
- The algorithms doing the matching?
- The training data for the algorithms?
What exactly do you need to buy (or copy) to have a version of TikTok?
- What about all the tech in China?
- The brainpower in Beijing that creates and improves the algorithms?
- What about the thousands of China-based software engineers?
- What about the vast operations teams that labels videos for the algorithms?
Is this the Ghemawat situation? Where:
- Improvements are in products, not processes.
- Learning is vested in a small number of employees (such as design teams).
Or is TikTok’s highly addictive video app mostly about “internal process learnings”, which cannot really be sold or copied?
It was a strange political discussion. The US government was arguing that TikTok needed to be US-based, but nobody could actually say what that actually meant. And it’s relevant to the competition with YouTube and others. To catch TikTok, what exactly do they have to copy and is it doable?
My approach is just to use my 6 levels of competition. And I find that gets me a pretty solid answer.
How the Other Tech Giants Compare to TikTok in Short Video
We can start with competitive advantages between the current video giants YouTube and TikTok.
- They are both audience builder platforms. They have similar business models.
- They have similar services. They both now offer short and long video. TikTok is going into long-form video. YouTube is going after short-form.
- They have similar user groups, consumers and content creators globally.
We could also consider internationally-expanding Kuaishou (Kwai) in this group.
Looking at competitive advantages, we can basically see fairly impressive competitive advantages versus everyone else.
- Both have powerful, global network effects, based on a long-tail of user generated content.
- Both have economies of scale in fixed costs (mostly IT spending and marketing).
- Both have good share of the consumer mind and habit formation. TikTok, in particular, is really habit forming. However, consumer behavior and technology does change fast in media.
So far, no major structural differences. The first big difference we see between YouTube and TikTok is in the Barrier to Entry.
YouTube benefits from having a massive and growing library of user generated content. And much of this content remains valuable. This is very hard to replicate. TikTok, in contrast, has content that decays rapidly. A new entrant would need much less content from day one to offer a viable short video service. And while we have seen no new entrants against YouTube, we have seen many companies launch viable short video services (WeChat, Facebook, Snap, etc.). The barrier to enter is much lower.
Moving on, we can look at the digital operating basics, which are pretty simple except for the machine learning capabilities I mentioned.
And then we get to what I think is the key question, which is whether TikTok has successfully run a Machine Learning / AI marathon (SMILE) that YouTube and others simply cannot catch.
Recall my list for digital marathons.
This is where the question of whether TikTok can be copied will play out.
Going back to the article on innovation vs. imitation (in “The Innovator’s Imitation Dilemma: TikTok and Facebook in Context”), the authors argue that the solution to leakage for TIkTok and other digital companies is “complex continuous innovation“.
“So how can an innovator succeed in the face of imitation by rivals? The answer is an approach we call complex continuous innovation, where individuals in the firm repeatedly reconfigure elements of their existing knowledge, fusing them together to deliver new product solutions in markets with many interdependent parts.”
Doesn’t their phrase “complex continuous innovation” sound like a digital marathon?
They basically argue that to avoid imitation, a company must continually innovate in complex areas that are difficult replicate. However, they point to innovation and I argue there are actually five different types of digital capabilities that continuous advance and pull ahead. For TikTok, I think the digital marathon that matters is Machine Learning, AI and Zero Human Capabilities (the M in SMILE).
And I think it’s all about the training data.
That is the key digital capability that must be copied to match TikTok. It’s not the algorithms. They are important but not unique. TikTok as a product is designed to capture data about videos and individual viewing preferences. That’s why you can only watch one video at a time. It captures a unique corpus of training data that continually updates its algorithms. And it does this at massive scale. Billions of users around the world are watching for hours each day. There is a walled garden of incoming training that is continuously updating individual matching as people change their viewing preferences. No other companies can access their data. This accumulated training data and the system that captures it is what the other companies must replicate to match their operating performance.
So how are the competitors doing? Are they catching TikTok in the machine learning marathon?
- WeChat has placed the highest priority on short video (and live streaming). Both are now pretty good in terms of operating performance. Again, it’s easier to be a new entrant in short video. It’s not awesome but this is TikTok’s most serious competitor. Fortunately, they are only in China.
- Kuaishou is a reasonable competitor. The matching is ok, not great. And they are going international and gaining scale. Worth watching.
- YouTube has been working on this for a long time for long form video. But their shorts feature is pretty bad.
- Facebook’s Reels sucks.
OK. Final question.
Who Is Going to Win in Short Video Internationally?
Internationally (i.e., not WeChat), I think this is mostly about YouTube vs. TikTok, with TikTok way out front in short video.
However, they are both serious video entertainment specialized companies. And they both have users in the billions watching their videos every day, generating training data. They have similar business models and competitive advantages, TikTok has better operating performance in Machine Learning. But YouTube has a higher barrier to entry for long videos.
I think TikTok is going to win. They are doing exactly what I would recommend they do.
- They are continuing to build demand and supply side economies of scale. Basically, network effects and economies of scale in IT and marketing.
- They are expanding into long videos.
- They are are pure-breed AI company. They are running their marathon in machine learning.
- And, importantly, they are launching more and more apps for their users, all of which rely on algorithmic matching. They are expanding their ML capabilities horizontally across multiple product types (music streaming, gaming, education, long video, etc.). YouTube is not doing this. This will get TikTok / ByteDance greater and greater scale and expertise in algorithmic matching.
I like ByteDance. They are a cheetah, hyperspecialized to do just two things.
- Creating mobile apps with very enjoyable user interfaces and user experiences.
- Doing AI and machine learning for the matching of text, music, video and other content.
Thanks for reading, jeff
- Why I Really Like Amazon’s Strategy, Despite the Crap Consumer Experience (US-Asia Tech Strategy – Daily Article)
- 3 Big Questions for GoTo (Gojek + Tokopedia) Going Forward (2 of 2)(Winning Tech Strategy – Daily Article)
From the Concept Library, concepts for this article are:
- Learning Curve and Experience Effect
- SMILE Marathon: Rate of Learning and Adaptation
- SMILE Marathon: Sustained Innovation
- CA7: Rate of Learning and Process Cost Advantages
From the Company Library, companies for this article are:
- TikTok / ByteDance
- YouTube / Shorts
- Facebook / Reels
I write, speak and consult about how to win (and not lose) in digital strategy and transformation.
I am the founder of TechMoat Consulting, a boutique consulting firm that helps retailers, brands, and technology companies exploit digital change to grow faster, innovate better and build digital moats. Get in touch here.
My book series Moats and Marathons is one-of-a-kind framework for building and measuring competitive advantages in digital businesses.