Start from gross profit per customer
Marketing budgets get set from revenue because revenue is the number everyone knows. It's the wrong input. Revenue doesn't pay for anything — gross profit does, and the gap between the two varies by a factor of ten across ordinary Indian businesses.
Four lines. Do them on paper before you talk to an agency, because these numbers change the conversation entirely.
- Gross profit per customer = average order or contract value × gross margin × purchases you expect in twelve months. Include the repeat purchases you actually get, not the ones you hope for.
- Extra customers per month you want from organic search once it's mature. Be specific. "More" isn't a number.
- Monthly gross profit at that rate = line one × line two. This is what organic search would be worth to you at steady state.
- Allowable monthly SEO spend = 25–50% of line three. Narrower band if you're also funding ads for the same customer; wider if the sales cycle is long and the customer sticks around.
Why percentage-of-revenue rules mislead
The standard advice is to spend 5–10% of revenue on marketing and some fraction of that on SEO. Run it across two real business shapes and watch it break.
A trading business turning over ₹2 crore at a 9% gross margin makes ₹18 lakh of gross profit a year. A 7%-of-revenue marketing budget is ₹14 lakh — seventy-eight percent of every rupee of gross profit the business generates. The same rule applied to a ₹2 crore software business at an 85% margin is ₹14 lakh against ₹1.7 crore of gross profit, which is close to a rounding error. Identical rule, identical revenue, and one of the two businesses would be dead within a year.
The rule fails a second way too. It produces budgets with no reference to what the work costs. If it hands you ₹12,000 a month for a national keyword set, it hasn't given you a small SEO programme — it's given you nothing, spent monthly.
- It ignores margin, which is the only thing that decides what you can afford.
- It ignores existing demand. A business with strong branded search already captures most of its organic ceiling and needs less.
- It ignores the market floor — the price below which the work simply doesn't fit.
- It ignores repeat purchase, which can double or triple the true value of a customer won this month.
Three worked Indian examples
Same method, three very different answers. The numbers below are illustrative shapes, not survey data — swap in your own and the arithmetic still works.
| Business | Gross profit per customer | New customers wanted/month | Monthly gross profit | Allowable SEO budget |
|---|---|---|---|---|
| Dental clinic, Pune | ₹4,800 — ₹8,000 treatment at 60% margin | 15 patients | ₹72,000 | ₹25,000–₹30,000 (35–40%) |
| D2C skincare brand | ₹1,230 — ₹770/order at 55% margin, 1.6 orders a year | 300 orders | ₹3,69,000 | ₹75,000–₹1,10,000 (20–30%) |
| B2B services firm, Bangalore | ₹2,10,000 — ₹6L contract at 35% margin | 0.67 (two a quarter) | ₹1,40,700 | ₹55,000–₹75,000 (40–50%) |
Why the percentages differ
- The clinic takes 35–40% because local SEO reaches its ceiling fast and there's a natural cap on how many patients one chair can see. Also, its keyword set is small — see what local SEO costs.
- The D2C brand takes only 20–30% because it's also paying for performance marketing against the same customer, and the two budgets compete. Its ₹75,000–₹1,10,000 band happens to straddle our ₹99,000 organic bundle, which is why that bundle exists.
- The B2B firm takes 40–50% because the sales cycle is six to nine months and a won client is worth ₹2,10,000 repeatedly. It can tolerate a longer payback than either of the others.
The number that decides everything
Notice the B2B firm needs less than one new client a month to justify ₹75,000. That's the whole argument for SEO in high-value B2B: the bar is absurdly low and everybody sets it too high anyway.
The D2C brand has the opposite problem. Three hundred incremental orders a month from organic is a serious programme, and a founder who budgets ₹40,000 for it has not done this arithmetic.
Where the market sets a floor you don't control
Your allowable budget is one number. What the work costs is another, and it isn't negotiable by wanting it to be lower. Roughly, in India:
- One city, local intent, 20–40 queries: ₹25,000–₹40,000/month. Below ₹25,000 you're funding part-time attention with no measurement layer.
- National, competitive category, 40–100 queries: ₹75,000–₹1,50,000/month. This is where four disciplines have to run at once.
- Ecommerce with thousands of SKUs, or multi-city: ₹1,00,000+ before anything interesting happens, because the technical surface area alone eats a retainer.
- Regulated or YMYL categories — health, finance, legal: add 20–30%, because every page needs review and citations and the ranking bar is higher.
Phasing the same money across twelve months
Most founders divide the annual number by twelve. That's the worst available shape. The audit, the baseline, the technical fixes and the first pillar pages are front-loaded work that gates everything after it — and they're mostly one-time. Spread them thin and you spend six months doing what should take ten weeks.
Here's ₹6,00,000 over a year, flat versus phased.
| Phase | Months | Monthly spend | What it buys |
|---|---|---|---|
| Foundation | 1–3 | ₹75,000 | Audit, frozen baseline, technical fixes, keyword mapping, first pillar pages |
| Build | 4–9 | ₹50,000 | Content on cadence, internal linking, link outreach, monthly review |
| Hold | 10–12 | ₹25,000 | Refresh what's slipping, defend what ranks, monitoring and reporting |
Two conditions before you taper
- The technical foundation is stable and monitored. A taper on a site that still breaks is just a slower failure.
- Content published in months 4–9 is compounding — impressions and positions still rising on their own. If it's flat, tapering locks in the flatness.
What to cut first when the budget shrinks
It will shrink. A quarter goes badly, ad spend gets priority before a festive peak, a receivable goes late. Cut in this order and you lose the least.
- Link spend. Pause it for a quarter. Existing links keep working; you lose compounding, not ground.
- Content volume. Four articles a month becomes two. Keep the cadence — silence is worse than slowness.
- New keyword expansion. Stop opening new fronts. Defend and improve what already ranks between positions 4 and 15.
- Sprint cadence. Move from four-weekly to six-weekly cycles. Same work, same people, fewer of them per year.
- Never: measurement or technical monitoring. Together they're about 10% of the fee and they're the only reason you can tell whether the other 90% worked. The full detail is in how to reduce SEO costs.