Every founder eventually asks us the same thing. Do these performance models actually work? Or is it just a pitch to win the deal, before you get moved onto a retainer? Fair question. After six years of pay-per-lead work across insurance, healthcare, property, education and B2B in India, here is the honest answer.
It works. But only when three things are true at once. The lead definition is tight. The path from click to sale actually works. And the numbers in your category can be measured. Miss one and the model breaks, sometimes badly. Below: where it works, where it fails, what fair rates look like, and what to demand from anyone selling you a pay-per-lead deal.
If an agency has quoted you a retainer, at least understand the alternative before you sign. We run B2C lead generation on pure performance for most clients. But only because we have learned where it fails, and we turn those jobs down.
Performance lead gen doesn't fail because the model is bad. It fails because the lead definition is loose, or the sales follow-up is broken, or the vertical doesn't fit. The model itself is robust.
What "performance-based lead gen" actually means
The agency charges for each qualified lead it delivers, at a rate you both agree first. You pay for the ads yourself, through your own ad accounts. The agency gets paid on leads, not on hours worked or campaigns launched.
Simple enough on paper. It gets complicated over two words: "qualified lead." That phrase decides whether this deal is fair, or whether one side is about to game the other.
The lead-quality trap
This is what kills most of these deals in the first 60 days. The contract says Rs. 400 per qualified lead. You think qualified means a real prospect. The agency thinks qualified means the form passed validation. Same word. Two different deals.
Then the leads arrive. The agency hits its target easily. 250 a week. But your sales team cannot reach most of them, and the ones they do reach are not who you were promised. Cost per lead looks brilliant. Cost per customer is a disaster. You want to stop. The agency points at the contract.
This is often not bad faith. It is the algorithm doing exactly what it was told. Tell Meta the goal is form submissions, and it hunts for the cheapest people who submit forms. There are plenty of those. They are just not your customers.
The cheapest lead is almost never the most valuable lead. Tight lead definitions are how you protect against that gap.
What a tight lead definition looks like
The fairest lead definitions are action-based and falsifiable. Here are three examples we've used in real contracts:
Real estate — high-ticket residential
Qualified lead: A person who confirmed budget range and intended timeline on a 10-minute discovery call with the in-house sales team. Form submissions that don't pick up the phone within 7 days are not qualified. Leads outside the agreed budget band (e.g., looking at sub-Rs. 50 lakh apartments when the project is Rs. 1.5 crore+) are not qualified.
B2B SaaS — mid-market
Qualified lead: A demo booked through Calendly by someone with a verified work email matching the ICP company list (10-500 employees, B2B SaaS). The demo must be attended; no-shows count as half-leads, billable at 50% of the agreed rate.
Healthcare — high-value procedures
Qualified lead: A person who confirmed the procedure they were enquiring about and confirmed they were within the catchment area on a 5-minute screening call. Leads outside the catchment area, or enquiring about procedures the clinic doesn't offer, are not qualified.
Notice the pattern. Every one of these has a human step in it — a call, a demo, a screening. That is the part no algorithm can fake. The agency can generate form fills all day. If nobody will take a five-minute call, the lead does not count.
What good performance lead gen rates look like in India
| Vertical | Fair CPL Range (INR) | Min Monthly Spend |
|---|---|---|
| Insurance (term, health) | Rs. 200-450 | Rs. 50,000 |
| Edtech (test prep, K-12) | Rs. 250-500 | Rs. 50,000 |
| D2C ecommerce (lead-then-sell) | Rs. 300-600 | Rs. 75,000 |
| Healthcare (clinic, IVF, dental) | Rs. 600-1,800 | Rs. 60,000 |
| Real estate (residential) | Rs. 1,200-3,000 | Rs. 1,00,000 |
| B2B services / SaaS demos | Rs. 1,500-5,000 | Rs. 75,000 |
| Coaching / high-ticket info products | Rs. 400-1,200 | Rs. 50,000 |
These ranges assume tightly defined leads with a human check. Loose leads — anyone who typed an email — will be quoted at 30 to 50 percent of these prices. Then your conversion collapses. The saving is not real.
When performance lead gen doesn't work
Be honest with yourself. There are situations where performance lead gen is the wrong model regardless of how well it's structured:
- Sales cycles longer than 6 months — when the gap between lead and revenue is too long to attribute, both sides end up arguing about whether leads from January are responsible for May's pipeline
- Verticals with extreme seasonality — wedding planning, tax preparation, school admissions — where the baseline shifts so much month-to-month that any per-lead rate is either hugely profitable for one side or unsustainable for the other
- Brand-new product launches with no conversion data — the agency can't price per-lead until they know what the lead-to-customer rate is, and the client can't tell them until they've tested
- Broken sales follow-up — leads not called within 24 hours die, however good they were. The agency cannot control what happens after the form, but it will get blamed for close rates that were really a sales problem
What goes wrong — and whose fault it usually is
- Loose lead definitions that reward volume over quality.Every contract should specify a falsifiable qualification step the agency cannot game. Without one, the algorithm chases the cheapest form submissions, which are almost never your real customers.
- Slow follow-up blamed on the agency.If your team takes two days to call, even perfect leads look like rubbish. Most "the leads are bad" complaints are really "we called too late" complaints. See our WhatsApp follow-up system for the fix.
- Agencies setting baselines they know are easy.If the agency proposes a target CPL that you can hit on day one with no optimisation, the agency is hedging. Targets should be ambitious but achievable, not soft.
- Attribution fights that belonged in the contract.Did that lead come from the agency's Meta campaign, or from your own email list? Settle it before launch. Not in month three, when the invoices get awkward.
- Cash-flow gaps.You fund the ads all month. The agency invoices for May's leads afterwards. That leaves you 30 to 45 days out of pocket. Plan for it.
What to negotiate in a performance lead gen contract
Six clauses to put in writing before any pay-per-lead engagement:
- Lead definition — exact, falsifiable, with a human qualification step that the algorithm cannot game
- Disqualification process — how you flag leads that don't meet the definition, the timeframe to dispute (typically 7 days from delivery), and the dispute resolution mechanism
- Per-lead rate — flat or tiered (rates often drop above volume thresholds; e.g., first 100 leads at Rs. 500, leads 101-300 at Rs. 425)
- Minimum spend commitment — both the agency's minimum (so they can run real campaigns) and the client's monthly cap (so spend doesn't spiral)
- Attribution methodology — last-click on Meta, GA4 last-non-direct, your CRM's source tracking — pick one and stick with it
- Initial term — usually 90 days minimum, with monthly renewal after; longer minimums protect the agency, not you
Frequently asked questions
In closing
Performance-based lead generation isn't a magic bullet — it's a contract structure. It works when both sides understand exactly what they're agreeing to, and it fails when either side is hoping the fine print will go their way. The single most important thing you can do as a buyer is define "qualified lead" tightly enough that the model can't be gamed.
If you're spending Rs. 50,000+ per month on lead generation through an agency on retainer, ask them to model what the engagement would look like as pure pay-per-lead. The numbers will be uncomfortable for them, which is exactly why it's worth asking.
At GUROB, we run pay-per-lead engagements for B2C verticals, hybrid commission-on-revenue for D2C and ecommerce, and per-install pricing for app marketing. The structure depends on what's measurable in your business — the audit (free) starts there.