India Full-Stack Thesis
I think every business offers value in one of three fundamental ways: physical goods, technology, or services. Most businesses just do one. But if you look at the most successful startups in India, it becomes clear that they offer value in at least two of the three ways.
The Three Value Propositions
Technology
A tech product — a marketplace, a SaaS platform, a consumer app — creates value through itself. Gross margins are exceptional, and when the product compounds through data or network effects, it can become genuinely defensible.
But tech-only has real limits in India. With B2B products, most customers operate on thin margins themselves and have limited budgets. Moreover, we’ve always had a huge supply of human capital. Most founders want to build in India but sell to the US. They admit India isn’t a large enough market.
On the consumer side, social networks are dominated by global platforms and we haven’t seen a large Indian business being built here. However, we’ve had a number of successful e-commerce companies, which had to build India-first playbooks.
All businesses use tech implicitly, which is natural. What creates real stickiness is when tech is what the customer directly interacts with, which builds lock-in.
Goods
Physical products are tangible – customers can see and touch what they're buying, which makes trust easier to establish. More importantly, there’s potential to build brands which allow businesses to sell products at a premium and drive repeats.
But there are structural challenges. Most products have limited TAM, and expanding beyond a certain scale means increasing channel complexity and higher customer acquisition costs, driving down operating margins. Usually, startups outsource manufacturing, which saves the capex but makes working capital the constraint. It also means that they become a pure brand or distribution play as production becomes commoditized.
Services
Service businesses are underrated. They require minimal investment but can generate solid cash flows and have regular repeats. Once the playbooks around sales and operations are built, it creates meaningful competitive advantages that are difficult to replicate.
The ceiling is competition and commoditisation. Service businesses in India face customers who negotiate hard on price while expecting top quality. Differentiation is difficult because the service is ultimately delivered by people, and maintaining consistency as headcount scales is challenging. This puts pressure on both pricing and margins.
Combination Compounds
There are three strong reasons why combination models are superior:
When brought together, tech, goods, and services complement each other. Tech makes business models scalable – we’ve seen that play out as distribution for consumer brands and platforms for services. The service angle gives a tech business real-world application.
India is an extremely value-conscious market, a behaviour shaped by economic reality. We expect exceptional quality at affordable prices – two contradictory things at the same time. Platforms that aggregate services bear the risk of customers going away if a third party vendor delivers poor customer experience. Full-stack is admittedly more complex and requires more investment but is the only way to deliver a top-quality experience. More importantly, it opens up larger revenue pools while also leading to better cost control. And when a business takes on more of the stack, it takes more off the customer's plate — they get what they want with less effort. That removal of friction is itself value.
Finally, businesses built along multiple vectors are more defensible. Doing two things well is exponentially harder than doing one, which makes it nearly impossible for others to copy the model. It takes time to perfect the playbook but once in place, the execution compounding creates extremely high barriers to entry. Difficulty is the moat. As others try to catch up, you’re already many years ahead.
Let’s Look at the Proof
Now, I want you to look at any successful Indian startup through this framework. Which one is a pure play? Nearly every single one is a combination, and that’s what made them work. Tech is present in all of them — but tech alone was never the business.
Tech + Services
Eternal and Blinkit, Swiggy and Instamart, and Zepto – they are all tech platforms that own the fulfillment layer. What about Flipkart and Amazon? Marketplaces that own the entire stack from logistics to warehousing to even payments. Meesho scaled significantly with outsourced logistics but Valmo became the centerpiece of their business model. Urban Company is perhaps the biggest success of all – nobody ever imagined that a service business could reach that scale, but it did, thanks to tech. There’s no shortage of 3PL operators in India, but Delhivery is what it is today because of technology.
Tech + Goods
Nykaa built a content and commerce platform, then extended into own-brand products — using data from the platform to inform what goods to create. Ola Electric and Ather are essentially vehicle manufacturers but they innovated around EV technology and are now adding software to differentiate versus legacy manufacturers. FirstCry combined e-commerce tech with own-label baby products. Same with Lenskart – own-brand eyewear, sold online (e-commerce) and now, offline (retail).
The AI Disruption Paradox
For the longest time, technology mattered above all else, and whoever had the better tech product had a better chance of success. But as tech talent grew, it meant more people could make equally good tech products. As a result, sales and distribution became more important. Even so, tech continued to be the strongest of the three. Building good tech products was still hard but it was more scalable versus goods and services businesses.
But, things have already started to change with AI because it is further sinking the time and cost to build technology. People are building things in days that took months just a year ago, and it’s spreading across all kinds of use cases. This means that having technology in the business will become the baseline, not the differentiator. And that bar will only keep rising.
The implication is significant – you might no longer have a competitive advantage purely on the back of a tech product. You will need more, so will have to extend towards the other two axes: the quality of the goods, the reliability of the service, the depth of operational execution, how you’re able to offer a full solution instead of just one part of it. Or maybe, the differentiator will be how AI-native the entire business is — not just the product, but the operations, the service delivery, the supply chain. The full-stack model, in that world, becomes even more important.
The Hypothesis
The rationale behind the full-stack model is simple: it allows you to offer more value to the customer. The model demands more patience to get right, probably even more capital, definitely an evolving mindset, carries more risk, but that’s exactly what creates moats.
That said, this doesn’t mean full-stack makes the business better automatically. You still need to have a strong value proposition. You’ll still make trade-offs around whether to strengthen the single value prop further or expand across the other axes. You’ll probably have to deal with cultural shifts as you reorganise priorities. None of this is easy.
You might think this makes sense only when looking back at large-scale startups. However, we have several portfolio companies that are building forward using this approach from a Seed stage. And this has made their value proposition much stronger and business model more defensible.