The question underneath the question
Every founder who asks whether SEO is worth the cost is really asking one of two different questions, and they have different answers.
The first is: will this produce more leads per rupee than the alternative over the next year? For most businesses in most categories, over twelve months, the honest answer is no. Paid search is faster and more predictable, and anyone telling you otherwise hasn't modelled it.
The second is: what do I own at the end? That's where the comparison inverts, and it inverts hard. So instead of arguing about which channel is better — a question with no general answer — model both to month 24, then switch the money off and watch what happens next. The difference between the two channels is entirely visible in month 25.
Six hours against six months: what a pause actually does
Start with the mechanics, because they're not symmetrical and most comparisons pretend they are.
Paid search is a rental. Your ads sit in an auction that re-runs on every query, and the moment the budget stops, you stop entering. There's no residue. The traffic doesn't taper — it ends, usually inside the same working day.
Organic is different in kind, not degree. There's no subscription connecting your invoice to Google's index. When you stop paying an agency, nothing in Google's systems is notified and nothing is revoked. Your pages stay indexed, your links stay counted, your internal architecture keeps distributing authority. What erodes it is competition: other people keep publishing, your content ages, the SERP changes shape, and a core update lands eventually. That's slow, and it's uneven.
| Time since you stopped paying | The paid side | The organic side |
|---|---|---|
| 6 hours | Impressions end. Your ads leave the auction the moment the budget is paused or exhausted. | Nothing at all. No mechanism links your invoice to the index. |
| 1 month | Zero traffic. Remarketing pools start ageing out of their windows. | Broadly flat. Rankings don't expire and nothing knows you stopped. |
| 3 months | Zero. Restarting means a fresh learning period and current auction prices. | First slide on fast-moving queries where competitors kept shipping. |
| 6 months | Zero. You're buying the same audience again from scratch. | Down on contested commercial terms. Evergreen and technical pages largely intact. |
| 12–24 months | Zero, permanently, unless you pay again. | Materially eroded where the category moves fast, durable where it doesn't. Links and architecture still working. |
The 24-month model, in rupees
Take ₹18,00,000 and spend it two ways over two years. Assumptions are stated so you can swap in your own — that's the only useful way to publish a model like this.
Paid side: the full ₹75,000/mo goes into media with no management fee at all, which flatters paid considerably. Blended cost per click of ₹40, which is a mid-market Indian services number — long-tail informational clicks run in single rupees, while insurance, legal and B2B software routinely clear ₹200–₹300. Two per cent of clicks become leads. That's 1,875 clicks and roughly 37 leads a month, flat, forever, as long as you keep paying.
Organic side: ₹75,000/mo retainer, ex-GST, ramping the way a competently run programme usually ramps — near zero for a quarter, building through months four to twelve, compounding through year two. Ranges rather than points, because the spread between a clean site in a soft market and an old site in a hard one is genuinely that wide.
| Period | Paid leads (₹40 CPC, 2% to lead) | Organic leads (₹75,000/mo retainer) | Spent by end of period |
|---|---|---|---|
| Months 1–3 | 112 | 0–5 | ₹2,25,000 |
| Months 4–6 | 112 | 10–25 | ₹4,50,000 |
| Months 7–9 | 112 | 25–50 | ₹6,75,000 |
| Months 10–12 | 112 | 40–75 | ₹9,00,000 |
| Months 13–18 | 225 | 120–205 | ₹13,50,000 |
| Months 19–24 | 225 | 160–265 | ₹18,00,000 |
| Total at month 24 | ~900 leads · ₹2,000 each | 355–625 leads · ₹2,880–₹5,070 each | ₹18,00,000 each side |
Where it flips: the marginal lead, then month 25
Two crossovers matter, and they happen at different times.
The first is marginal cost per lead. By months 19 to 24 the organic programme in the model is producing roughly 28 to 47 leads a month for a flat ₹75,000, which is ₹1,600–₹2,700 per lead. Paid is stuck at ₹2,000 per lead and will stay there or drift upwards as auction prices rise. So somewhere around months 20 to 24, the next organic lead becomes cheaper than the next paid one — while the cumulative figure is still behind, because year one was expensive and produced almost nothing.
The second crossover is the one that decides the argument. Stop both channels at month 24. Paid goes to zero the same afternoon. Organic keeps producing something like 28 to 47 leads in month 25 for ₹0, then declines gradually across the following year. Even a conservative decay leaves the organic side producing free leads for eighteen to thirty months after the last invoice — and every one of those is pure margin.
That's the entire case. Not that organic is cheaper. Not that it's faster. That it accrues to an asset you keep, and paid accrues to a receipt. If your business is being sold, wound down, or pivoted inside eighteen months, the asset is worthless to you and you should buy ads. That's a real answer and it applies to more companies than founders like to admit.
One clarification that saves arguments later: a reduced retainer is not the same event as a stop. Dropping from ₹75,000 to ₹40,000 keeps the pages current, keeps the technical work from rotting, and keeps a slow publishing cadence alive — which is enough to hold most positions on evergreen topics even if it stops adding new ones. A full stop is what triggers the decay curve above. If cash is the problem, cut the scope and say so out loud, rather than pausing for a quarter and restarting from a worse position than you left.
What actually survives when you stop
Worth being precise about the asset, because 'you own it' is the vaguest sentence in agency sales. Most founders hear it as a reassurance about file ownership. The useful version is a list of six specific things, three of which take years to rebuild if you throw them away.
What's left on the organic side after the last invoice:
- Ranked pages with query history. Not just files — URLs that Google has already assessed, placed, and accumulated behaviour data against. That history is what makes them cheap to revive later.
- Referring domains. Links earned by real coverage don't get handed back when the contract ends. This is the slowest-built and slowest-decaying part of the whole thing.
- Site architecture. URL structure, the internal link graph, canonicals, schema, redirect maps, speed work. Permanent until somebody replatforms without a migration plan — which is how most of it gets destroyed, incidentally.
- The content, as files you own. In our contracts the client keeps every asset, including drafts, briefs and the keyword map. Check yours says the same, because plenty don't.
- Brand demand created along the way. People who found you through a comparison page and later search your name directly. That query doesn't appear in an SEO report and it's frequently the most valuable thing produced.
- A surface AI assistants can retrieve. AI Overviews and assistants cite pages that exist and can be crawled. Your ad has never been cited by anything.
The payback arithmetic to run before you sign anything
The model above uses lead counts. Your board cares about rupees, and converting one to the other needs exactly one number: contribution per lead — gross margin on a closed deal multiplied by the rate at which inbound leads close.
Using the mid-point of the organic ramp above (117 cumulative leads by month 12, 491 by month 24), here's when a ₹75,000/mo programme pays for itself at different contribution levels.
- Cumulative spend against cumulative contribution is the only honest payback measure. Monthly ROI in month three is a meaningless ratio and every agency that reports it knows why.
- Your margin swings this far harder than your agency does. The gap between the ₹1,500 row and the ₹8,000 row is nineteen months of break-even. No amount of execution quality closes that.
- Add 18% GST to the cash plan, recoverable as input credit if you're registered. It doesn't change the return, it changes the cash you need in month one.
- Then decide whether you'd rather own it. Two channels, same rupees, one leaves a balance sheet item. The full price picture is in what SEO costs in India, and the channel-by-channel argument in SEO against paid search.
| Contribution per lead | Contribution by month 12 | Contribution by month 24 | Break-even lands around |
|---|---|---|---|
| ₹1,500 | ₹1,75,500 | ₹7,36,500 | Never, on these assumptions |
| ₹3,000 | ₹3,51,000 | ₹14,73,000 | Month 31 |
| ₹8,000 | ₹9,36,000 | ₹39,28,000 | Month 12 |
| ₹25,000 | ₹29,25,000 | ₹1,22,75,000 | Month 5–6 |
The two places paid wins anyway, and we'd say so
There are categories where the asset argument simply doesn't apply, and pretending otherwise is how agencies sell retainers that shouldn't exist.
Demand with a deadline
Admissions cycles, festive windows, a product launch, an event, a competitor's outage, a funding-driven land grab where being second is the same as being absent. If you need volume in three weeks, there is no organic answer — none, at any price, from anyone.
The same applies where the SERP itself squeezes organic out. On heavily transactional queries the first screen is often ads, shopping units and marketplace listings, and the first organic result sits below the fold on a phone. Ranking there is worth much less than the same position on an informational query, and any model that ignores SERP shape is flattering organic.
Demand you haven't validated yet
If you don't yet know which message, price or audience works, ads buy that answer in a fortnight for ₹30,000–₹50,000. Committing twenty pages to a positioning you haven't tested is ₹6,00,000 of content that gets rewritten, and the rewrite doesn't inherit the rankings.
This is also the cheapest SEO diagnostic in existence. If exact-match paid clicks on your five best keywords don't convert, free versions of those same clicks won't convert either. Run that test before you sign anything — we set out the rest of the qualifying conditions in does every business need an SEO agency.
Why the comparison is artificial, and what we actually recommend
Nobody spends ₹18,00,000 on exactly one channel for two years. We modelled it that way because a forced choice makes the difference visible, not because it's a plan.
In practice the two feed each other, and the feeding is one-directional in a useful way. Paid tells you within weeks which queries produce revenue rather than traffic — that's the most reliable keyword research an SEO programme will ever get, and it costs a fraction of what a research phase costs. Organic then takes the queries paid proved out and stops you renting them forever. Running paid on a term you rank first for is optional; running paid on a term you'll never rank for is just the cost of doing business.
The split we'd suggest to most founders with ₹1,50,000 a month: keep paid running on the queries closest to the transaction, put organic behind the queries people search before they're ready to buy, and review the mix every quarter using cost per qualified lead by channel rather than by campaign.
And if the honest answer for you is that SEO isn't worth the cost — low margin per lead, no category search demand, an eighteen-month horizon — we'd rather tell you that on the first call than nine months into a retainer. We only take three clients a month, and the guarantee we carry is movement against your own frozen trailing-90-day organic lead baseline, not a position on a page. You can't underwrite that for a business the arithmetic already rules out. What actually happens to the numbers after a stop is covered in what happens when you stop doing SEO.