*15 March 2025*

One data point is worth a dollar. Two are worth two dollars. But ten million? That can build a company. Data’s value scales non-linearly - when aggregated and leveraged effectively, it fuels exponential growth.

Companies harness this power through the Virtuous Data Cycle, where data collection, analysis, and application continuously enhance products, improve user experience, and attract more users, generating even more valuable data.

In the simplest terms, the Virtuous Data Cycle works as follows:

Collect user data.
Store & organize it efficiently.
Analyze for patterns and insights.
Apply insights to improve products and services.
Enhance user experience, driving engagement and retention.
Repeat, compounding value at every iteration.

Unlocking Exponential Growth with the Network Effect

Network Effects emerge when a product or service becomes more valuable as more people use it. Cities illustrate this principle: as populations grow, infrastructure, businesses, and opportunities expand, making them even more attractive. This self-reinforcing loop follows Zipf’s Law, where the largest city dominates, and the second-largest is about half its size, the third a third, and so on.

The distribution of population sizes of 276 metropolitan areas in the USA in 2000 on a log-log scale, which clearly demonstrates a Power Law distribution.

The same applies to companies. As they grow, they attract more users, talent, and investment, strengthening their market position. The Virtuous Data Cycle ties in closely with the theory of Network Effects. In data-driven businesses, this relationship becomes even more powerful because data not only enhances the user experience but also fuels monetisation and competitive advantage.

How Network Effects Strengthen the Virtuous Data Cycle

More users → More data: As a platform grows, it collects richer insights on user behaviour, refining its services and improving the experience.
More data → Better algorithms & personalisation: Large datasets power smarter AI and recommendation engines - think Facebook’s News Feed, YouTube’s recommendations, or TikTok’s For You Page. Better personalisation boosts engagement, reinforcing the cycle.
Better experience → Higher retention & growth: Improved experiences keep users engaged, drive word-of-mouth growth, and strengthen network effects, fuelling exponential expansion.
More users → Stronger market position: A massive user base creates a competitive moat - attracting top talent, greater investment opportunities, increasing efficiency, and even influencing industry policies. New entrants struggle to compete without comparable data.
More engagement & data → Higher revenue & infrastructure investment: Increased engagement unlocks monetisation (ads, subscriptions, commerce). Higher revenue funds R&D and infrastructure improvements, further enhancing the experience.
The cycle repeats, compounding dominance: Each loop strengthens the platform’s edge, making it harder for competitors to catch up.

Not all companies master this cycle. Twitter had network effects but never fully leveraged its data to drive advertising revenue. OpenAI capitalised on first-mover advantage to amass user feedback, but its long-term profitability remains uncertain. Facebook and Google, however, perfected both the Virtuous Data Cycle and Network Effects - turning data into dominance.

Case Study: OpenAI – From Hallucination Station to the AI Powerhouse

When OpenAI released GPT-2 in 2019, it was impressive but far from revolutionary. OpenAI had taken the research published by Google’s DeepMind on transformers, and turned it into a web app where users could ask it questions. The model was prone to hallucinations (see example below) and novelty wore off as early adopters grew frustrated with the lack of functionality. But OpenAI kept iterating.

OpenAI’s virtuous data cycle improved their family of GPT models by leveraging research breakthroughs and through increased user interactions, using feedback (thumbs-up/down) and conversations to refine responses. They leveraged their industry partnership with Microsoft, amongst others, to fuel growth and adoption.

By making their platform free to use and appealing to consumers directly, OpenAI's ChatGPT reached 100 million users in just two months after its launch in November, making it the fastest-growing consumer application in history. They began to appeal to developers and businesses with their APIs, who embedded them into diverse applications and created a broader ecosystem. They continued moving into the B2B space by leveraging their partnership with Microsoft to integrate GPT-4 into Bing and Microsoft 365, which reinforced their position as the default B2B AI provider. Competitors like Google faced delays in launching alternatives, allowing OpenAI to capture market share before rivals could respond. Their massive user base and collection of data make their models increasingly difficult to rival, although some competitors are now seemingly making headway in this regard.

While they mastered the network effect, their road to profitability remains uncertain. In 2024, OpenAI reportedly spent $9 billion to make $4 billion. They spent an estimated $3-4 billion on training, another $2 billion on inference (running models to answer users’ questions), $1.5 billion on salaries and employee benefits, $500 million on data-related expenses and the remainder on various other operating expenses. Their future profitability hinges on the appetite of users and companies to fork out hundreds or even thousands of dollars for a tool that some consider only marginally better than open-source competitors.

Data and Networks as a Business Model

To grow a B2C platform exponentially, you have to eliminate bottlenecks:

Leverage word-of-mouth: users’ testimonials and organic network effects should be your primary marketing strategy. OpenAI never had to advertise heavily to acquire users - virality did the work.
Prioritise feedback from super users. A few engaged users will provide the most valuable insights, guiding product development.
Build clean, high-quality data pipelines from the get-go: early adopters will provide the most valuable insights into your product’s strengths and weaknesses, and set the direction for the next stage of evolution. A caveat to this - don’t optimise too early. Use off-the shelf tools like Google Sheets while your user base is small enough. It’s not dumb if it works.
Reduce onboarding friction. If sign-up takes more than a minute, you risk losing users before they even experience your product.
Embed data privacy compliance from day one. Regulations like GDPR and CCPA can become major roadblocks. Retrofitting compliance later is costly and erodes trust.

Network Effects: B2C vs. B2B

In B2C, network effects are straightforward - users want to be where their friends are. FOMO fuels adoption.

B2B is different. The decision-maker isn’t always the end user, meaning you need to convince multiple stakeholders - often their boss’s boss - to invest in your platform. Unlike B2C, where shared adoption creates value, B2B companies sometimes benefit when competitors don’t use the same tools they do.

However, network effects still apply in B2B. Once a tool reaches critical mass, not using it becomes a competitive disadvantage. Employees switch jobs and introduce their favorite tools to new workplaces. Over time, widely adopted products, like SEMRush for SEO or Cloudflare for cybersecurity, become industry standards.

How B2B Can Mimic B2C Growth Strategies

Some B2B markets (e.g., enterprise SaaS, healthcare, government contracts) move slower due to long sales cycles, procurement processes, and compliance requirements. However, many B2B businesses have successfully scaled by adopting B2C-style viral tactics.

1. Freemium Model → ChatGPT

The base product is free, making it easy for individuals and businesses to adopt. However, OpenAI didn’t bake in privacy from the start. Conversations may be used to train future models (part of their Virtuous Data Cycle). To unlock enterprise-grade security and compliance, businesses must upgrade.

2. Pay-to-Play Model → Instagram

Instagram is free for businesses, but organic reach is restricted. Barring going viral with a clever reel or partnering with influencers, to fully benefit from the platform’s network effects, companies must pay transaction fees (2.9% for Instagram Checkout) and invest in ads to reach a broader audience.

3. Viral Adoption + Enterprise Lock-in → Figma

Figma started as a free, collaborative design tool, making it easy for designers to work together. As its adoption grew, it became an industry standard (network effect). Eventually, businesses had no choice but to integrate Figma, and pay for enterprise features like security, admin controls, and private cloud hosting.

Key Takeaway

B2B companies no longer have to rely solely on long sales cycles and enterprise deals to scale. By leveraging freemium models, network effects, and viral adoption strategies, they can accelerate growth and become indispensable in their industries, just like successful B2C platforms.

But growth fueled by network effects is only as strong as the data foundation behind it. Without structured, high-quality data, companies risk losing insights, stalling product evolution, and missing key opportunities.

To fully unlock the potential of your Virtuous Data Cycle, ask yourself:

Is your data structured for success? Do you have well-organized databases that enable seamless analysis and decision-making?
Is your company truly data-first? Does data literacy extend across teams, or is it siloed within a few roles?
What’s the non-monetary value of a new user? Beyond revenue, what insights do you gain from each customer, and what do you lose when they churn?
What drives freemium-to-paid conversion? Are you tracking the key incentives and friction points that push users to upgrade?
How well do you track user behaviour? Are you consistently analyzing engagement patterns and using those insights to refine your product?
Are you gathering direct user feedback? Users tolerate mediocre products, until a competitor better meets their needs. How often do you survey your users?
How sticky is your product? How difficult or easy would it be for a user to switch if a better alternative emerged?

B2B companies that master both network effects and data-driven strategy create products that are not just widely adopted but deeply embedded in their industries. The companies that fail to do so leave the door open for someone else to become the next industry standard.

If you enjoyed this post, please consider supporting my writing by buying me a coffee.

Buy me a coffee

Notes on Virtuous Data Cycles and Network Effects