Platform strategy for product managers, network effects to AI agents

Platform strategy covers the choices a product manager makes when a product connects two or more groups, such as banks and users or sellers and buyers, and value grows as each side grows. IBEF reports 713 banks live on UPI in April 2026, and the Zinnov-Nasscom 2026 report classes about 39% of India's 2,117 GCCs as portfolio hubs that own product, platform and IP.

What changes in strategy when a product becomes a platform?

A product serves one group of users with one set of features. A platform serves at least two groups that depend on each other, so the manager's decisions shift from "what should users get" to "who should be on each side, what do they pay and what may they build". Feature roadmaps still exist, but they sit under a rules-and-incentives layer.

UPI is a useful example because its scale is published. IBEF reports that in April 2026 it processed 22.35 billion transactions worth Rs 29.03 lakh crore, with 713 banks live on the platform. Banks and apps sit on one side and users and merchants on the other, and every new participant raises the value of the network for the rest.

The same logic applies at smaller scale. A marketplace needs sellers to attract buyers and buyers to attract sellers, an internal data platform needs product teams that rely on it, and an AI agent platform needs tools and data that agents can call. The strategy question is always which side to grow first and what to give it.

Hiring shows the same pattern. Meesho's Senior Product Manager posting for its Data and Analytics Platform (Bengaluru, listed 23 September 2026) asks for 7+ years overall with 4+ in product management and includes managing a small team of product analysts and associate product managers. The one platform posting read sits at the experienced end of the ladder (7+ years); the ladder is set out in the product manager career path in India.

How do network effects and multi-sided markets show up in Indian examples?

Network effects arise when each added participant makes the platform more valuable to the others. The Indian evidence is in market size and growth, and each figure below belongs to a different kind of platform.

IBEF projects India's e-commerce industry growing from US$ 125 billion in 2024 to US$ 345 billion by 2030, a compound annual growth rate of 18.4%. Redseer's report of 17 September 2026 said online retail grew about 25% year on year in the first half of 2026 and is projected to cross USD 90 billion in 2026. Redseer put quick commerce at about USD 9 billion in that half, with about 60 million monthly transacting users.

Multi-sided markets add a trade-off. A marketplace that subsidises one side to reach scale has to show that the other side will pay later. Meesho's PM II posting (listed 18 August 2026) reflects this by asking product managers to plan go-to-market with marketing stakeholders, so growth work and product work share one owner.

Redseer's figures are projections and half-year readings from a consultancy, and IBEF's are projections. They show a market that rewards scale, not that any single marketplace will win it.

How should a platform handle monetisation, openness and governance?

Monetisation on a platform is a choice of who pays and for what, for example a fee per transaction on one side or advertising sold around the activity.

Ad platforms are the clearest live example. InMobi's Programmatic Monetization posting (Bangalore, first published 7 October 2026) asks for 8-10 years of product management experience with significant ad tech or programmatic experience, and lists building AI-driven products for demand optimisation, pricing, forecasting and anomaly detection among its responsibilities. Razorpay's Product Manager II - Ads posting (first published 1 September 2026) asks for 4-7 years and ownership of business metrics and customer outcomes.

Openness and governance are the harder strategy questions. Opening a platform to outside builders grows what it can do, but it also forces rules on quality, data access and dispute handling. A payments rail needs clear rules because money moves across 713 banks, and an agent platform needs them because software acts on a user's behalf.

Where do AI agents sit on a platform?

Agents can be users of a platform, calling its services through interfaces, or products built on it. Paytm's "Product Management - Pi Agents - SMB" posting (Noida, listed 27 August 2026, team Inference and Agentic AI) asks for 5-9 years of product management experience and prefers hands-on experience building AI-native or agentic products, including prompt design, evals and agent orchestration.

Several questions follow from that. Who gets to publish an agent, how an agent is evaluated before it acts, what an agent is allowed to see and how its usage is priced can fall to the product manager. A fuller treatment of the hands-on side is in agentic AI for product managers.

LinkedIn's Skills on the Rise 2026 for India lists API, Prompt Engineering, Workflow Automation and LLMOps among the fastest-growing "AI & Automation" skills. The list is cross-role, not PM-only. Paytm's AI Payments posting (Noida, listed 21 October 2026, products for the UAE market) adds a boundary: it expects an AI-first mindset but states that building models is not part of the role.

Atlassian's 2025 survey of product teams in the US, France and Germany found AI saving about two hours a day, mostly on routine work, and not yet helping with prioritisation, planning and advanced analytics.

Which platform roles appear in live Indian postings?

The table pairs each platform type with an Indian scale signal and a live posting reviewed on 30 October 2026. Each scale signal comes from a different publisher and measures something different, so the rows cannot be added together.

Entry-level platform postings were absent from the 12 postings sampled. Naukri's JobSpeak for September 2026 showed hiring up 5% to 7% at 8-15 years of experience and down 2% at 4-7 years, across all roles in India. LinkedIn's Jobs on the Rise 2026 list for India names Product Manager among the top roles people moved from into the AI manager title, whose median prior experience is 6 years.

GCC-owned platforms are the hardest to see. Zinnov-Nasscom reports 2,117 GCCs across 3,728 units and a 2.36 million workforce, with data as of March 2026. Zinnov sells GCC set-up services, and its 39% is its own classification, so it signals that platform ownership from India exists without counting PM jobs.

Working as a PM on an internal platform, what metrics and skills matter most?

A PM who has spent two years on an internal platform has a particular advantage and a particular risk. The advantage is understanding of dependencies across teams. The risk is that internal users cannot leave, so weak adoption can hide behind a captive audience.

Platform metrics differ from feature metrics. Feature teams watch conversion and retention, while platform teams watch adoption by the teams or partners they serve, time to integrate, reliability and the value created for the other side. These measures are a framing, not a standard.

When engineering leads a modernisation, product management's role is to hold the outcome. That means tying the migration to a business metric and to the experience of the platform's users, not to the architecture alone. Experimentation habits help, though Atlassian's survey of 1,000+ product professionals in the US and Europe found that only 60% of teams make experimentation a regular part of their process, and 84% of product teams worried their current products would not succeed in the market.

Stakeholder management also counts. LinkedIn's 2026 India skills list names Stakeholder Management among the fastest-growing skills, again not PM-specific. For the metrics side in depth, see product metrics and experimentation skills, and for the wider scope of AI in the role, the existing guide to the scope of product management with AI in 2026.

Where does a product management programme fit?

For managers who want structured grounding in this area, the Certificate Programme in Product Management in the Agentic AI Era is one route. It is a live online programme from IIM Tiruchirappalli (IIM Trichy), run with Masai as the implementation partner, covering product foundations, platform strategy, AI for product managers, agentic AI and a team capstone to design and pitch an agentic-AI-powered product. The program is open to graduates and above.

Judgement about platforms comes mostly from working on one, so the programme is best treated as a framework to apply to real work, not a replacement for it.

Frequently asked questions

What is a platform product manager?

A platform product manager owns a product that other teams, partners or users build on or transact through, such as a payments rail, a marketplace or a data platform. Meesho's Senior Product Manager posting for its Data and Analytics Platform is one example, asking for 7+ years overall.

Do you need technical skills to be a product manager on a platform?

It varies by employer. Zeta's PM II makes 3+ years of hands-on development mandatory, while Paytm's AI Payments posting says building models is not part of the role. Paytm's UPI posting asks for SQL and analytics tools.

Do product managers have a future as AI agents spread?

Demand data says yes for now. Lenny's Newsletter, using TrueUp data, found open PM roles at tech companies worldwide at a three-year high in March 2026, and Paytm's Pi Agents posting seeks PMs for agent products. Neither figure is a forecast.

What is the job demand for platform product managers in India?

Naukri's JobSpeak covers all roles and has no platform-PM line. Its September 2026 edition showed GCC hiring up 4% and AI and ML roles up 20%, and Zinnov-Nasscom counts 2,117 GCCs. These describe the market around platform roles, not the roles themselves.

Next step

The Certificate Programme in Product Management in the Agentic AI Era, IIM Trichy page sets out the curriculum and how to apply.