Pay for the catch. Not the net.

India invented outcome-based business

Before there was a single SaaS dashboard in this country, there was the dabbawala, paid for the tiffin delivered, not the hours walked. There was the broker, who ate only when the deal closed. The thekedaar, paid per project. The karigar, paid per piece. The commission agent in every mandi, every showroom, every export house, whose income was simply a share of the result he produced.

Indian business has a phrase older than any pricing page: no cure, no pay. You do not pay the vaid for the visit. You pay him when the fever breaks.

We did not need a Harvard case study to teach us that you reward the outcome, not the effort. It was in the bones of the bazaar.

India does not need to learn outcome-based pricing. It needs to remember it.

Then we were talked into renting chairs

Somewhere in the last fifteen years, software arrived with a different deal. Pay per seat. Pay per month. Pay whether you use it or not.

So you bought ten licences and used four. You paid for the CRM your sales team quietly went back to managing on WhatsApp. You renewed the tool nobody had opened since the demo. Every Indian promoter knows the small annual shame of the subscription audit, the line items you cannot remember saying yes to.

That model was never built for you. It was built for the vendor, whose revenue was safest when it was disconnected from whether you got any value at all. You were paying rent on a chair, and hoping someone sat in it.

A tool waits to be used. It bills you either way.

An agent is not a tool. It is a worker.

Here is the shift that changes the deal.

The last generation of software gave you a better chair: a faster screen, a cleaner button, a place to type. You still had to do the work. An AI agent does the work. It reads the enquiry, drafts the quotation in your format, chases the payment on the third day, reconciles the order against the ledger, and hands you the one decision that actually needs a human.

You do not log in to an agent the way you log in to a tool. You give it a job, the way you give a job to a person. And the moment software can carry a job end to end, the old question, how many seats do you need, stops making sense. You do not buy your accountant by the seat. You keep him because the books close.

A tool waits to be used. A worker waits to be useful.

The agentic economy, in one line

An agent that can finish a job can also, with your permission, act on it.

Inside limits you set, it can place the reorder before the shelf runs dry, send the GST-ready summary to your CA, book the site visit, raise the invoice, settle the small supplier bill that always slips. Not because it went rogue, but because you told it: up to this amount, in these cases, with this approval, go ahead.

That is the agentic economy. Not chatbots that answer. Workers that act, transact, and report back, within a fence you build and can move whenever you like. The work stops waiting for a free pair of hands.

Outcome-based: the meter moves when the work lands

When software becomes a worker, the honest way to price it changes too.

You begin to think, naturally, the way the mandi always did: pay for the lead that converts, the report that ships, the reconciliation that closes, the night the agent stood guard so no order went unanswered. The meter moves when the work lands, not when the calendar flips.

This is the fairest deal software has ever offered an Indian business, because it puts the vendor on the same side of the table as you. If it does not work, you have not bought a shelf full of nothing. If it works, you are happy to pay, the way you have always been happy to pay the salesman his commission. The incentive and the outcome finally point the same way.

The cheapest worker you will ever hire is the one who bills you only when the job is done.

Why India is built to win this one

Most countries have to unlearn the seat-and-subscription habit to think in outcomes. India barely picked it up. The instinct to pay for results is still the default in most of our firms, which makes the agentic shift feel less like a leap and more like a homecoming.

And the rails are already here. UPI made machine-speed, small-value transactions ordinary in a way most of the world is still catching up to. Our businesses run bilingual by default, English and the mother tongue in the same breath, which is exactly how customers actually talk. MSMEs make up roughly a third of GDP and nearly half of our exports, millions of firms each with one expensive expert and not enough hands. That is not a market that needs convincing. It is a market that needs leverage.

But hold the keys

A worker who can act on your behalf is powerful, which is exactly why ownership has to sit with you, not with a vendor in another country.

The right way to enter the agentic economy is the way you would hand responsibility to a trusted new manager. Clear limits. A spending ceiling. Approvals on anything that matters. A log of every action, in plain sight. And critically, the agent running on infrastructure you control, with your own keys, so your data and your customers' data never become someone else's asset. Under the DPDP Act, that is not just good sense. It is the law catching up to common sense.

The agentic economy should hand you more control, not less. If a deal asks you to give up ownership of your data to get the leverage, it is the old rent trap wearing a new word.

Start with one outcome

You do not adopt the agentic economy with a committee and a two-year roadmap. You adopt it the way you would test a new hand: give it one job and watch.

1. Pick one outcome you can name. Not "use AI." Choose "no enquiry goes unanswered for more than five minutes," or "every quotation out the same day," or "the books reconciled by Monday morning."
2. Set the fence. Decide what the agent may do alone, what needs your nod, and the limit it must never cross. Put it in writing once.
3. Keep the human on the last call. The agent does the ninety minutes of assembly. Your best person keeps the fifteen minutes of judgment. Taste, trust, and the final yes stay with people.
4. Let it compound. Every correction makes the next result better. The worker that never sleeps also never stops learning your way of doing things.

One outcome, proven, beats ten tools, subscribed. Start where your business already feels the gap, and let the result make the case.