Three stories this week span AI economics, market structure, and payments infrastructure — and they share a pattern worth noticing. Companies quietly stopped buying software once cheap agentic coding made "build it yourself" the default instead of the fallback. India's stock market found a floor that has nothing to do with the buyers everyone still keys their FII commentary around. And WhatsApp Pay became India's largest UPI app not by shipping anything new, but because a regulator lifted a limit that had nothing to do with product quality.
Buying software just became optional for a third of companies
McKinsey's "State of AI in 2026" survey found that 32% of organizations have skipped buying at least one software product or feature this year because agentic coding tools let them build it in-house instead. That's the clearest build-vs-buy signal the SaaS industry has gotten yet — not a benchmark or a demo, but actual purchasing decisions changing.
If you're building developer tools or SaaS right now, the products most exposed are the ones without a real moat: internal dashboards, simple CRUD apps, workflow glue — the kind of thing a competent team can now scaffold with an agent in an afternoon. What still holds up is anything with a genuine data advantage or a compliance surface a competitor can't replicate in a weekend of vibe coding. Worth actually auditing your own roadmap against that test rather than assuming it doesn't apply to you.
The market floor moved from Wall Street to systematic investment plans
This week: FIIs sold ₹5,612 crore in the cash segment while DIIs bought ₹23,156 crore, a net institutional inflow of roughly ₹17,544 crore. FIIs sold hardest on Monday (₹7,986 crore); DIIs bought hardest on Friday (₹8,930 crore). The Nifty spent the week failing to hold above 24,000, with crude near $95/bbl and Fed rate uncertainty weighing on sentiment.
The old heuristic — watch FII flows, trade the direction — keeps losing predictive power because domestic flows, powered by SIP contributions north of ₹31,000 crore a month, now absorb foreign selling more often than not. That's structurally different from a market where sentiment swings on whatever headline hit Bloomberg overnight. It also means retail capital, moving on autopilot through mutual funds, is increasingly the marginal buyer holding the market up — a different risk profile than one driven by foreign capital, and worth understanding rather than assuming away.
WhatsApp just became India's biggest UPI app, and it's not about the product
WhatsApp Pay processed 167.89 million UPI transactions worth ₹12,957 crore in July alone, overtaking both CRED and Amazon Pay. The trigger wasn't a feature release — it was NPCI removing the onboarding limits that had been capping WhatsApp Pay's growth all along.
For anyone building on top of a platform-dependent distribution channel, this is worth sitting with. WhatsApp didn't out-build CRED's UX or Amazon Pay's rewards program; it already had default-app gravity in a billion pockets, and the only thing holding it back was a regulatory ceiling. The moment that ceiling lifted, product quality stopped being the deciding variable. If your growth strategy depends on a platform's rules staying where they are today, that's a risk worth pricing in explicitly, not an assumption to leave unstated.
The thread connecting all three: in each case, the constraint that mattered wasn't technology maturing or a product getting better, it was a threshold quietly moving. Agentic coding removed the cost threshold that made buying software the default. Systematic retail flows removed FII sentiment as the swing factor in market direction. And NPCI removed a regulatory cap that had nothing to do with merit. The narratives people reach for to explain "why X won" — better product, better model, better timing — often lag behind the actual threshold that moved. Worth checking which of your own assumptions about your market are still resting on a threshold that already shifted.
Top comments (0)