How to Allocate a Marketing Budget in India: A Working Planning Guide

A marketing budget is a set of decisions about risk, not a universal percentage of revenue. A growing Indian D2C store, a local professional service and a B2B software business can spend the same amount and need entirely different plans. The useful question is not “What does everyone spend?” It is “What can this business afford to learn, and what result would justify the next investment?”

This guide walks through that decision using a clearly labelled example. The figures are invented planning inputs, not Adservex client results or Indian market benchmarks. Replace them with your own order, margin and cash data. You will finish with an operating budget that separates advertising, people, creative, tools and learning instead of hiding everything inside a single ROAS target.

Start with a decision and a time horizon

Choose the decision your budget must support. A business with repeat purchases might want to establish a viable first-order acquisition cost. A software company might need to find whether a narrow customer segment books qualified demonstrations. A services firm might want more enquiries that its existing team can actually fulfil. These are different jobs, even if all three use Google Ads.

Write a sentence that names the outcome, time window and constraint. For example: “Over the next twelve weeks, assess whether we can acquire customers for this product while preserving cash for fulfilment.” This is more useful than “grow ten times” because it can be connected to a spreadsheet, a campaign and a stop decision. Include what will not be funded: a new geography, an untested product line or channels that need creative you cannot produce.

The time horizon must fit the business. Paid campaigns can produce observations quickly, but refunds and repeat purchases take longer. Search content may need a longer observation period. Do not judge all channels on the same seven-day sales report just because that report is convenient to download.

Calculate contribution before choosing a channel

Revenue is not money available for advertising. Start with net realised revenue and subtract costs that change with each order: product cost, payment fees, packaging, fulfilment, shipping support and expected returns. Decide how discounts and taxes are handled so that two people can reproduce the same number. An accountant can help make the definition consistent with your business records.

For a D2C store, cash-on-delivery creates an important distinction between an order placed and an order delivered and collected. A campaign dashboard may report the placed order before you know whether it will generate cash. Compare placed-order reporting with delivered-order economics rather than silently treating the two as equivalent. Returns, replacements and failed deliveries should appear in the planning assumptions.

For a service business, contribution might be contract revenue less the delivery labour and subcontracting required to serve the client. For software, include relevant onboarding and support costs when those materially change with each customer. There is no single definition that works across every model. The requirement is a written definition that your marketing and finance teams both understand.

A worked acquisition example

Imagine a store with these hypothetical inputs for one delivered order. The values are illustrative and do not describe a typical Indian brand.

Input	Illustrative amount	Why it matters
Net realised revenue	₹2,000	After the adjustments used by finance
Product cost	₹850	Cost of the item sold
Payment, packaging and fulfilment	₹180	Costs required to complete the sale
Shipping and expected returns allowance	₹220	Estimate based on your own history
Contribution before acquisition	₹750	Revenue less the costs above

If the business wants to retain ₹250 per acquired order for overheads and risk, its provisional first-order acquisition ceiling is ₹500. That does not mean every campaign should bid exactly ₹500. It means the planning model has a starting constraint that needs testing. If delivered-order costs rise, the ceiling falls. If your average basket includes multiple products, calculate the contribution at the order level rather than applying a single-product margin to the entire basket.

Suppose the test goal is 120 new delivered orders. At the illustrative ceiling, media spend would be up to ₹60,000. Creative, measurement and agency fees sit outside that calculation unless your definition explicitly includes them. If those costs add another ₹30,000, the total programme cost is ₹90,000, not ₹60,000. Reporting both media-only and fully loaded acquisition cost prevents an apparently profitable campaign from concealing an unprofitable operating plan.

Treat repeat purchases cautiously

Repeat purchasing can support a higher acquisition cost, but only if the evidence and cash timing support it. Do not use an optimistic lifetime-value slide to justify losses today. Segment customers by their first purchase month, then track what those groups actually buy again, when the revenue arrives and how much contribution remains after retention costs.

A repeat purchase made nine months later does not pay next month's salary. Finance should decide the acceptable payback window from available cash, working-capital commitments and the uncertainty of repeat demand. A six-month payback is not a universal rule for Indian businesses. Some can sustain longer cycles; others cannot carry even a short cash deficit.

When cohort data is limited, plan a conservative case using first-order economics and show a separate scenario for repeat contribution. Keep that scenario out of the primary acquisition ceiling until the evidence is credible. This makes uncertainty visible rather than pretending it has disappeared because the spreadsheet contains a precise decimal.

Separate the budget into operating jobs

Useful budgets show what the money buys. Advertising buys distribution. Creative buys assets that explain the offer. Measurement buys reliable observation. Retention buys communication with consenting customers. People and agency fees buy execution capacity. These jobs overlap, but listing them separately makes omissions obvious.

Here is a hypothetical ₹1,20,000 monthly pilot, offered as a worksheet rather than a recommended benchmark:

Job	Example allocation	Deliverable to agree
Acquisition media	₹60,000	One defined acquisition test
Creative and landing-page work	₹25,000	Named assets and a usable destination
Measurement and reporting	₹10,000	Tested events and a decision report
Execution support	₹15,000	Agreed responsibilities and review time
Contingency	₹10,000	Approval required before use

The table totals ₹1,20,000. A common planning error is to allocate the full amount and then add a test pot on top. If experiments need protection, fund them inside the envelope. Another error is to double-count creative that is already included in an agency retainer. Record whether each cost is incremental, recurring or included elsewhere.

Pick a channel based on the unanswered question

Channel selection follows the question. Search advertising can help test demand expressed through queries. Paid social can test whether a product explanation resonates with a defined audience, provided the team has credible creative. Email can help existing subscribers understand products or return to a useful offer; it cannot manufacture an opted-in customer list from nothing.

An early-stage business with little evidence should resist running every channel at once. Five small campaigns can generate five ambiguous reports while producing no decisive learning. One focused experiment with reliable tracking may be more informative. A mature business may need a portfolio of acquisition and retention investments because its decision is about incremental contribution, not initial viability.

Read our Google Ads versus Meta Ads comparison if channel roles are unclear. For a broader plan, use the small-business marketing strategy guide. Neither replaces checking whether your offer, fulfilment and sales process are ready for the demand you hope to create.

Write the experiment before spending

An experiment should include a hypothesis, audience, offer, observation window, spending limit and decision rule. “Try Instagram” is not an experiment. “Test whether a product demonstration produces relevant enquiries from this audience, with this landing page and these qualification questions” is closer to one. The decision rule might concern enquiry quality, customer contribution or whether the tracking is reliable enough to continue.

Define failure as well as success. You might stop because tracking duplicates purchases, leads are outside the service area or the product cannot be delivered at the promised price. These are operational failures, not reasons to keep buying traffic until a dashboard improves. Put someone in charge of each stop condition so it is not ignored between marketing and sales.

Use a test log. Each row should contain the change, date, spend, observation, uncertainty and next action. Keep the scope stable enough to interpret the result. If you change price, creative, targeting and checkout simultaneously, a better outcome may be welcome, but you will not know which change caused it. Sometimes emergency fixes require that trade-off; record it honestly.

Review weekly, but do not confuse noise with evidence

Weekly checks should catch broken tracking, overspending, disapproved ads, unavailable inventory and leads that cannot be handled. These are important even before the campaign has enough observations for a performance judgment. A campaign can be statistically inconclusive and operationally unsafe at the same time.

For a performance review, align the date ranges. Advertising reports, CRM records and finance data may recognise an outcome at different times. Include conversion lag, cancellation lag and sales follow-up lag. Compare equivalent periods and customer definitions. Do not announce that acquisition improved merely because today's platform report includes more attributed purchases than last week's incomplete export.

Write down the confidence level of each decision. “Sales improved, but delivered orders are not yet reconciled” is a useful conclusion. “ROAS increased, therefore double the budget” hides a missing check. Your review should describe what is known, what remains uncertain and the smallest next investment that could resolve the uncertainty.

Reallocate by marginal return and capacity

The best historical average is not automatically the best destination for additional spend. A campaign's next rupee may perform differently from its previous rupee. Consider audience saturation, creative fatigue, inventory availability and the sales team's ability to respond. A business can waste spend by generating more demand than it can fulfil even when its click metrics look healthy.

Before moving money, ask whether the apparent difference survives a change in attribution or customer quality. Brand search may show a stronger reported return while mostly collecting demand created elsewhere. A prospecting campaign may show weaker immediate revenue but produce genuinely new customers. The answer is not to favour one automatically; it is to distinguish their jobs and investigate incrementality where the budget and data allow.

Reallocation can also mean investing in a destination rather than a channel. If visitors cannot find delivery information, or enquiries wait days for a reply, another ad variation may be less useful than fixing those obstacles. Record these investments as part of marketing's operating plan so the media team is not held responsible for problems it cannot control.

Questions to ask an agency or internal team

Ask what costs are excluded from the proposal. Confirm ownership of ad accounts, landing pages, creative files and reporting access. Request the definition of a qualified lead or customer, the basis for any benchmark and the person responsible for validating revenue. A useful proposal explains both execution and measurement rather than listing a large quantity of deliverables.

If someone quotes a guaranteed return, ask what happens to the guarantee when inventory, discounts or conversion tracking change. A forecast can be a scenario with transparent assumptions; it is not a promise merely because it appears in a professionally designed deck. Compare proposals on scope and evidence instead of treating the lowest fee or highest projection as proof of value.

Our agency pricing guide covers proposal structure. If you need support, contact Adservex with your goal, current channels and budget constraint. A useful conversation starts with those inputs, not an invented estimate of what your business should spend.

Your one-page budget worksheet

Record the outcome, twelve-week cash envelope, net revenue definition, contribution calculation and provisional acquisition ceiling. List every cost category, including taxes and fees where applicable. Name the experiment, its owner and the observation period. Add stop conditions for tracking, fulfilment, lead quality and cash exposure. Finally, state which records will be reconciled before any increase is approved.

The worksheet is complete when another person can explain why each line exists and what evidence would change it. It does not need to predict the future perfectly. It needs to make assumptions visible, stop avoidable losses and let the team learn without confusing more spending with more progress.

Further reading and measurement references

Consult Google Ads documentation on conversion value when configuring value-based measurement, and Google's attribution documentation when interpreting reported credit. Platform attribution is not the same as a finance-approved contribution calculation. Keep both views, explain their differences and use the business decision to determine which view matters.