Flywheels - The Unstoppable Force of Growth

Last week, we discussed growth loops and where to find them. The post generated good discussions, and I am grateful for folks providing inputs and making it better. Please keep them coming :)

This week I am going to touch a topic often confused with growth loops — Flywheels. The reason for this confusion is simple; they look and feel the same. But as it happens with frameworks, the real value lies in nuances of these frameworks. So if you want to apply the Flywheel concept, it’s better to understand it well.

Jim Collins even wrote a book about this, Turning The Flywheel. Below is an excerpt from his book about Flywheel where he explains the Flywheels concept.

Picture a huge, heavy flywheel—a massive metal disk mounted horizontally on an axle, about 30 feet in diameter, 2 feet thick, and weighing about 5,000 pounds. Now imagine that your task is to get the flywheel rotating on the axle as fast and long as possible. 

Pushing with great effort, you get the flywheel to inch forward, moving almost imperceptibly at first. You keep pushing and, after two or three hours of persistent effort, you get the flywheel to complete one entire turn.

You keep pushing, and the flywheel begins to move a bit faster, and with continued great effort, you move it around a second rotation. You keep pushing in a consistent direction.  Three turns ... four ... five ... six ... the flywheel builds up speed ... seven ... eight ... you keep pushing ... nine ... ten ... it builds momentum ... eleven ... twelve ... moving faster with each turn … twenty … thirty … fifty … a hundred.

Then, at some point—breakthrough!  The momentum of the thing kicks in your favor, hurling the flywheel forward, turn after turn ... whoosh! ... its own heavy weight working for you. You're pushing no harder than during the first rotation, but the flywheel goes faster and faster.  Each turn of the flywheel builds upon work done earlier, compounding your investment of effort. A thousand times faster, then ten thousand, then a hundred thousand. The huge heavy disk flies forward, with almost unstoppable momentum.

Remember the words ‘unstoppable momentum’ when you think of Flywheels.

Limitation of Loops and Birth of Flywheels

‘Yo’ will forever be remembered as the app that was too simple. It lets users do exactly one thing: say "yo" to each other.

![Yo App Goes From Seriously Dumb to Surprisingly Useful]({"src":"https://bucketeer-e05bbc84-baa3-437e-9518-adb32be77984.s3.amazonaws.com/public/images/45040028-262b-492d-969d-56490347af06_610x424.jpeg","height":424,"width":610,"resizeWidth":474,"bytes":null,"alt":"Yo App Goes From Seriously Dumb to Surprisingly Useful","title":null,"type":null,"href":null})

Despite its simpler premise, the app had strong growth. The app quietly launched in April of 2014 and grew. In July, the app got $1.5 million in funding from investors. By September, the app had 2.7 million registered users and 1.2 million monthly active users.

Fast forward today, the app still lives. But ‘Yo’ is asking users to donate to its Patreon to keep the app alive. From the Play Store description of the app

Does ‘Yo’ have a growth loop? It does. Here is its growth loop.

A growth loop brings new users to the platform as it did for the ‘Yo’ app. But it does nothing about retaining them. So even if a product has growth loops, it may or may not have sustainable growth.

This is why we needed a new tool to think about sustainable growth — Flywheels.

Flywheels

The term comes from physics — a flywheel is a mechanical device specifically designed to efficiently store rotational (kinetic) energy. The key idea here is that Flywheels resist changes in rotational speed (ω) and momentum by their moment of inertia (I), meaning once they are in motion, they are difficult to slow. Here is a 1- min short video explainer on this.

In the startup world, this implies flywheel-like products gain enough momentum and it becomes their moat over time. It’s hard to beat/slow their flywheel of growth.

Back in 2014, when Uber was valued at tens of billions of dollars, there was a lot of criticism from valuation experts. Famous investor Ashwath Damodaran wrote

Earlier this month, investors poured $1.2 billion into Uber, a tech company whose smartphone app connects taxi drivers to passengers. The share of the business these investors received suggests that Uber is worth $17 billion, a mind-boggling sum for a young company with only a few hundred million dollars in revenue.

An equally compelling response came from one of Uber’s earliest investors, Bill Gurley. This tweet exchange between David Sacks and Bill sums his response.

This tweet explains why Investors got bullish on Uber. I will take a minute to elaborate on the drawing that David Sacks shared.

1. As more users join Uber, more drivers will come looking for them.

2. The geographic density of Uber increases, i.e. more uber drivers and riders in an area. More driver density would mean faster pickup —> even more users will start coming for faster pickups

3. More users will lead to more # of rides for a driver. More rides —> better income. So they can afford to lower the prices.

All of it becomes a virtuous growth cycle. Lower prices and low waiting times will keep users on the platform. As the users are on Uber, drivers will also be on the platform. The lockin will keep Uber’s flywheel in motion over a long period of time. Long enough to weed out any competition, and build a sustainable company.

Today Uber is valued at $65 B in public markets. Whatever one’s feeling is about Uber's valuation, it definitely changed the world. Flywheels often do that as we will see next.

Superimposition of Loops = Flywheel Effect

In the last post about growth loops, we discussed the superimposition of loops. Flywheels often have multiple loops/wheels and can be seen as a superimposition of loops.

Let’s take the best Flywheel built to date to understand this — Amazon. An excerpt from the Amazon website which explains how Amazon grows.

Amazon is built on the concept of a virtuous cycle focused on the customer. The idea was constructed on a napkin by CEO Jeff Bezos, and still remains a living, breathing part of Amazon..

Jeff Wilke explains the meaning of the virtuous cycle in this video.

To break various loops/wheels in motion

1st loop is around customers and seller growth

Customer Experience (CX) —> More Customers —> More Sellers —> More Selection —> CX

2nd loop is around prices and cost structure

More Customers —> Lower Cost Structure —> Lower Prices —> Better CX —> More Customers

3rd important loop not listed above is around delivery time :

More Customers —> Better Logistic Network Utilisation —> Lower Delivery cost +Faster Delivery —> More Customers

The flywheel also has multiple actors like sellers and buyers dependent on each other and creating lockin with each other. And it’s not just Amazon. Best marketplace models are often flywheels with multiple actors. Examples - Uber (riders and drivers), AirBnB (host and guest), AliExpress (buyer and seller), Swiggy (restaurant, buyers), UrbanClap (service providers, buyers), Angellist (investors, startups), etc.

Key to Building Flywheels

Flywheels are all about building advantages. A good list of advantages always comes handy when thinking about Flywheels. Here are the 6 most common advantages -

1. Direct Network Effect examples are telephone networks, the Internet, Facebook, WhatsApp, etc.

2. Economies of Scale can be seen in many software companies. This is because most of their costs are fixed. For example, Google costs come mainly from servers and software development, and these do not increase with the number of users. Every additional search lowers the average cost of a search. Examples of companies profiting extensively from Economies of Scale are Google, Facebook, Salesforce, Amazon, Netflix, etc.

3. 2-sided Network Effect can be seen in marketplaces like Uber, Airbnb, AliExpress, Amazon, Swiggy, UrbanClap, Opentable, Angellist, etc. Recently, it has also happened on social network apps like Instagram with the advent of micro-influencers. Users and micro-influencers create 2-sided network effects.

4. Switching Costs examples include SaaS companies like Salesforce, Intuit, Microsoft, etc. These are mostly unique to products where you can use only one of them at a time. Though B2C companies build switching costs by creating loyalty programs. Amazon has done a fabulous job with Prime, which has an annual churn of 9% only.

5. A brand habit example is Apple. Other companies with strong brand habits are Disney, Tesla, and Nike.

6. Proprietary tech in software companies is mostly in the form of data-based AI/ML algorithms. Google’s search, Netflix’s recommendations, Amazon product search and recommendations, Alexa Skills, etc. are some of the examples. In absence of big data, it’s hard to build this advantage. Though hardware products like Apple and Samsung also rely on hardware or design patents to build an advantage.

You may have noticed a few names appearing multiple times. This is because of the more the advantages the better the flywheel. Let’s see a few examples

1. Uber checks 3 of 6 advantages — 2-sided network, brand habit, proprietary tech like surge pricing

2. Amazon checks 5 of 6 advantages — 2-sided network, economies of scale, switching cost through prime, brand habit, proprietary tech like the recommendation and product search, AWS, etc.

3. Youtube checks 5 of 6 advantages — 2-sided network, economies of scale in CDN costs, switching cost for creators as one has to move subscriber base, brand habit, proprietary tech like a recommendation, and product search.

4. Apple checks all 6 of 6 advantages — 2 sided-network of app developers and iPhone users, economies of scale in manufacturing, direct network effect to some extent, high switching cost due to familiarity of the interface, strong brand habit, proprietary tech on hardware and design.

5. Tripadvisor checks 3 of 6 advantages — 2 sided-network of users and travel commodities, economies of scale high as the cost is limited to software development and the content is user generated increasing the marginal ROI of TripAdvisor manifold, strong brand presence as it keeps appearing on the first page of your search results when it comes to travel.

Differences: Loops and Flywheels

At the start of this chapter, we said that Loops and Flywheels are different. Now we have seen that flywheels are often a superimposition of growth loops. So how are flywheels and loops they different? Why do we need two different terms when thinking about the growth of the product?

To understand that, let’s look at the products which grew due to growth loops but fizzled out in lack of flywheel

1. We have already discussed the Yo app. Yo had 1 of 6 advantages — direct network effects. But a lot of users don’t find Yo useful, so the direct network effect is also limited. Is it a wonder that it failed?

2. HQ Trivia is another example. It was a popular live quiz show where you could earn a life by inviting friends. So it built a viral growth loop and grew to > 2 million DAUs. But because of the lack of any of the advantages (0 of 6), it fell fast and fell hard.

3. Vine, a popular video loop app that offered creators the ability to make and share six-second videos similar to TikTok, had 1 of 6 advantages — direct network effects. For a short period, it also had another advantage, i.e. 2-sided Network Effect. In 2016, Twitter announced that it was shutting down Vine. The app couldn't compete with the rising popularity of Instagram, and users who were once considered as Vine celebrities started posting videos on Instagram. Why did Tiktok succeed, whereas Vine failed?

Vine versus Tiktok

Failure to make money for Vine and Creators, Scale, UX — From TheVerge

Years of executive churn likely contributed to Vine’s failure to make money. For a while, brands were happy to pay Vine stars directly to make ads and share them to their millions of followers. But after Snapchat and Instagram grew into hundreds of millions of daily users, marketers’ interest in Vine dropped significantly. They had once longed for ways to grow their own followings on the app — through paid placement offerings similar to Twitter’s promoted tweets and promoted accounts.But Vine never came through with any options, in part because the founders resisted monetization from the start, sources said. It never took a cut of stars’ deals with brands

TikTok is the overseas version of Douyin by Bytedance.

There were about 400M users of Douyin in China. More importantly, China had a developed business ecosystem formed around KOLs (key opinion leaders) who made their livings online by creating short videos for brands, etc. TikTok allows creators to monetize their own content, and it's got more tools for creators to customize or collaborate on content.

The tools built by Tiktok for creators to make good content and collaborate are pretty good. This means delightful content for the end-users.

The scale allows Tiktok to built a much better content recommendation engine, which is a key strength of Bytedance. The scale of Tiktok also creates a 2-sided network effect.

All-in-all, Tiktok checks 5 of 6 advantages

1. The stronger 2-sided network effect between users and creators

2. Direct network effect built by users sharing these videos with each other

3. Economies of scale due to user-generated content (UGC)

4. Switching costs for creators in the form of losing money if they leave the platform

5. Proprietary technology like recommendation engine which will get better with more users and more usage

Loops vs Flywheels

Looking at these examples and what we discussed before, we can identify a few key differences in Loops and Flywheels:

1. Flywheels are superimposed loops/ multiple wheels interacting with each other. Loops can be seen as a subset of Flywheels.

2. Loops with shorter cycles can make a product viral and it grows very fast like HQ Trivia and Vine. But if it gains none of the advantages of Flywheels over time, it has a high chance of failure. A key difference between loops and flywheels is the cycle time. Flywheels have a longer cycle time because of the completion of a cycle is dependent on multiple actors in the system. As the old adage goes, “Anything worthwhile takes time to build” ?

3. All Flywheels build sustainable businesses with high momentum. The momentum can serve as the moat for the business as it takes a long time to slow down. Few loops build a sustainable business and the ones that do have long cycles in themselves. UGC loops translating to SEO benefits have a long cycle, same is the case of paid marketing where the LTV exceedeeng CAC will have a long cycle.

4. The feedback cycles are also longer in Flywheels. Even if something breaks, it’s barely noticeable initially because of high momentum. So such products should always define leading indicators of business and growth like NPS, rating, etc. Lagging indicators like revenue and profits start showing signs of worry when it’s too late.

Please note that flywheels and loops do overlap. It’s hard to establish clear distinction at times. But that’s the beauty of frameworks and tools, you should have as many in your toolbox. So get a good grasp over both of these tools while crafting a growth strategy ?

So what’s stopping your product from creating a Flywheel? DMs open on both LinkedIn and Twitter for any discussions.

See you next week. We will start the topic of Onboarding and Retention from the next week.

Sincerely,

Deepak