Breakdown: Network Effect
A phone is useless if you are the only person who owns one. That simple idea explains why Amazon, TikTok, and Meesho are so hard to beat, and why Zerodha and Groww became giants without it. This is not investment advice. If you like blogs that simplify topics like this, consider liking & Sharing. Your feedback is welcome & would help me get better.
Think about the first person in your colony to buy a telephone. Who could they call? Nobody. The phone was a box on the table. Then a neighbour bought one, then a relative in another city, and slowly the same box became more and more useful, even though nothing about the box had changed. What changed was the number of other people holding one.
That is a network effect. A product has a network effect when it becomes more useful to you each time someone else starts using it. The investment firm NFX, which studies this closely, puts it simply: a product has network effects when it “becomes more valuable as usage increases” (NFX).
I wanted to understand which of the companies I use every day really have one, and which only look like they do. So I picked four global names, Amazon, Google, Canva and TikTok, and four Indian ones, Zerodha, Groww, Nykaa and Meesho, and tested each of them.
Why the crowd matters so much
Here is the maths that makes network effects powerful. Every new person can connect with every person already there. With 2 people, there is 1 possible connection. With 5 people, there are 10. With 10 people, there are 45.
The rule is n × (n − 1) ÷ 2. So 1,000 people can make 499,500 connections. This idea is often linked to Robert Metcalfe, who co-invented Ethernet (Wikipedia), and it explains why a network’s value can grow much faster than its size. Real life is messier, because you will never talk to most of the other users, but the direction holds. Each new member adds more value than the one before.
Four kinds of network effect, and three look-alikes
Not every network effect works the same way. Here are the four kinds that matter for our eight companies.
- Direct: more people means more people to reach. WhatsApp is useful because your family is on it.
- Two-sided: more buyers pull in more sellers, and more sellers pull in more buyers. Each side makes the platform better for the other.
- Data: every search or swipe teaches the system to serve the next person better.
- Creator: more users attract more creators, whose work attracts more users.
The part I found most useful is the bottom row. Three things often get mistaken for network effects:
- Scale: being big and cheap helps you, but not because someone else joined.
- Virality: friends inviting friends is how an app spreads. It does not make the app itself better.
- Brand: people choose a company because they trust it.
All three can build a great business. They are just not network effects.
A network effect you can draw: Meesho
The clearest way to see a two-sided network effect is as a wheel that turns faster each time it goes round.
In July 2021, Meesho became, in its own words, “the first e-commerce company in the country to introduce zero percent commission” for sellers. Sellers and manufacturers joining the platform rose 25% soon after (MediaBrief). More sellers meant more choice and lower prices. That pulled in more shoppers, especially from smaller towns, which in turn pulled in more sellers. By the 12 months to September 2025, Meesho had 23.42 crore annual transacting users and 7.06 lakh active sellers (IndMoney).
Each turn of the wheel is easier than the one before. That is what a strong network effect feels like from the inside.
Amazon: strong. Amazon opened its site to outside sellers in 2000. Shoppers attracted sellers, and sellers brought more choice and lower prices, which attracted more shoppers. Every review a buyer leaves helps the next buyer decide. In his 2018 letter to shareholders, Jeff Bezos wrote that third-party sales had grown from 3% to 58% of Amazon’s physical sales since 1999 (Amazon). The crowd of sellers is now a bigger part of Amazon than Amazon itself.
Google: medium. Google has a real data network effect: billions of searches teach it which results people find useful. It also runs a two-sided market, where searchers attract advertisers. But switching search engines takes one click. In 2024, a US court ruled that Google is a monopolist, and found it had paid more than $26 billion in 2021 to be the default search engine on phones and browsers (NPR). So part of Google’s lead comes from the network, and part comes from being the default. Today Google handles about 90% of the world’s searches (Search Engine Journal, citing StatCounter).
Canva: medium. Much of Canva is a tool you use on your own. The network shows up in two places. First, creators publish templates and earn royalties from a pool that grows with Canva’s subscription revenue (Canva). More users means a bigger pool, which attracts more creators, which means more templates. Second, teams share designs, so once your team is on Canva, it is harder to leave. Canva had more than 265 million monthly users by the end of 2025 (Wikipedia).
Get Amar’s stories in your inbox
Join Medium for free to get updates from this writer.
TikTok: strong. TikTok is unusual. You do not need any friends on it for it to work. The feed works because of everyone else: more viewers attract more creators, and more creators give the algorithm more videos to choose from, so every feed gets better. TikTok went from 55 million monthly users in January 2018 to 1 billion in September 2021 (Tubefilter).
The Indian four
This is where the honest test gets interesting. Two of the four Indian companies became giants without a strong network effect.
Zerodha: weak. My trades on Zerodha do not get any better because someone else joined. Zerodha won by cutting fees, building a clean app and teaching investors for free. In January 2019, NSE data showed it had become India’s largest broker, with 8.47 lakh clients against ICICI Securities’ 8.44 lakh (YourStory). It is a great business. Its edge is cost and trust, not the crowd.
Groww: weak. Groww built a simple app for first-time investors, many of them from smaller cities (Contrary Research). Friends telling friends helped it spread fast. That is viral growth, not a network effect: the app spreads, but it does not get better for you as more people join. Even so, Groww passed Zerodha in active NSE clients in September 2023 (Business Standard), and by June 2026 it had about 1.31 crore active clients, roughly twice Zerodha’s 68 lakh (StartupTalky).
Nykaa: medium. Nykaa hosts more than 4,000 brands, and shoppers attract brands just as brands attract shoppers. Reviews and more than 28,000 creators and affiliates help the next buyer choose (BeautyMatter). But Nykaa buys and stocks much of what it sells, so trust in Nykaa itself matters as much as the network.
Meesho: strong. The clearest network effect of the four, as the flywheel above shows.
The scorecard
To keep my grading consistent, I asked the same four questions of every company.
A company gets STRONG only if it is a clear Yes to the first two questions: does one more user make it better for me, and do two sides pull each other in? Amazon, TikTok and Meesho pass. Google, Canva and Nykaa have real but partial network effects. Zerodha and Groww are mostly No.
That last point is the one I keep coming back to. A weak network effect does not mean a weak company. Zerodha and Groww are two of India’s most successful fintech businesses. It only means that if a better or cheaper rival arrived tomorrow, their crowd alone would not protect them. Amazon’s and Meesho’s crowd would.
My Observations:
Before I did this, I used “network effect” for any company that grew fast. Now I ask one question: does this product get better for me when a stranger joins? If the answer is yes, the company has something rivals find very hard to copy. If the answer is no, its success rests on something else, like price, product or trust, and those need defending every single day.
In the next Breakdown, I look at the moment when a network effect actually kicks in: the tipping point.
This is part of Breakdown, a series where I take one business idea and pull it apart until the numbers make sense. The network effect grades are my own judgement, based on the sources linked here. This is not investment advice. If you like blogs that simplify topics like this, consider Liking, Sharing & Commenting.
Sources
- NFX: The Network Effects Manual
- Wikipedia: Metcalfe’s law
- Amazon: 2018 Letter to Shareholders
- Wikipedia: Amazon Marketplace
- NPR: Google loses antitrust case over search dominance (Aug 2024)
- Search Engine Journal: Search engine market share (StatCounter data)
- Canva: About the Canva Creators program
- Wikipedia: Canva
- Tubefilter: TikTok reaches 1 billion monthly active users
- YourStory: Zerodha becomes India’s largest broker (Jan 2019)
- Contrary Research: Groww
- Business Standard: Groww topples Zerodha (Oct 2023)
- StartupTalky: India’s stock broking market in June 2026
- BeautyMatter: How Nykaa built India’s beauty market
- MediaBrief: Meesho zero percent commission (Jul 2021)
- IndMoney: Meesho IPO review
