How to Measure ROI from Digital Marketing for an Indian Small Business

Ask most agencies about ROI and you'll get a number like fourteen times. Ask how it was calculated and you'll usually find it was revenue divided by ad spend — which is not return on investment, and can describe a business that lost money.

This is a guide to measuring it properly: the arithmetic that tells you whether marketing is actually paying, the full cost most people leave out, and a realistic way to trace customers when half of them arrive on WhatsApp or simply walk through the door.

Use Margin, Not Revenue

The formula that matters is straightforward. Take the gross profit your marketing generated, subtract what the marketing cost, and divide by what it cost.

The word doing the work is gross profit. Revenue is not yours — most of it is the cost of the goods or the service you delivered. Two businesses with identical revenue and identical spend can be at opposite ends of profitable, purely because of margin.

	Retailer at 25% margin	Service business at 60% margin
Ad spend	₹25,000	₹25,000
GST on ad spend (18%)	₹4,500	₹4,500
Agency or freelancer fee	₹15,000	₹15,000
Total cost	₹44,500	₹44,500
Revenue generated	₹1,50,000	₹1,50,000
Gross profit on that revenue	₹37,500	₹90,000
Actual result	Loss of ₹7,000	Profit of ₹45,500

Both businesses could truthfully advertise "six times return on ad spend". One of them went backwards. This single distinction explains most of the gap between what marketing reports claim and what owners see in their bank account.

The one number to memorise

Break-even ROAS is roughly one divided by your gross margin.

At a 25 per cent margin you need about four rupees of revenue per rupee of ad spend just to stand still. At 40 per cent, about 2.5. At 60 per cent, about 1.7. Work yours out once, write it on a card, and you can judge any campaign in ten seconds — which is more than most dashboards will do for you.

Count the Whole Cost

Most ROI calculations only count what the platform billed. The real denominator includes:

• GST on ad spend — eighteen per cent, and not a rounding error on a small budget.
• Agency, freelancer or in-house salary cost.
• Creative production — photography, video, design, even when a family member does it.
• Tools and subscriptions paid annually but used monthly.
• Discounts and free delivery given to win those sales. A twenty per cent festive discount comes straight out of the margin you are measuring.
• Your own time. Not a cash cost, but if you spend two evenings a week on this, it belongs in the decision even if not in the spreadsheet.

The Real Problem: You Can't See Where Customers Came From

Here is what the ROI guides written for e-commerce brands miss. For most Indian small businesses, the sale doesn't happen on a website. Someone sees an ad, messages on WhatsApp, negotiates, and pays at the counter three days later. No analytics tool sees any of that.

You don't need enterprise attribution to fix this. You need five cheap habits.

• A different WhatsApp link per channel. Click-to-chat links can carry a pre-filled message. Make the Instagram version open with a different phrase to the Google version, and your inbox tells you the source without anyone asking.
• A code word or coupon per platform. "Mention DIWALI10" on one channel and a different code elsewhere. Crude, free, and it works even for walk-ins.
• UTM tags on every link you post. Two minutes with a link builder, and traffic stops appearing as an anonymous lump in your analytics.
• A separate phone number in ads, if calls are a meaningful share of your enquiries.
• An enquiry register. One row per enquiry: date, name, source, what they wanted, whether they bought, value. A notebook is fine. This is the single highest-value measurement tool available to a small business, and almost nobody keeps one.

Combine that register with asking every customer how they found you, and within two months you will know more about your marketing than any dashboard could tell you.

Why the Platform Numbers Don't Add Up

If you run ads on two platforms, add up what each claims and you'll find they have jointly sold more than you did.

That isn't dishonesty exactly — it's that each platform counts a conversion whenever it had qualifying contact with the buyer. A customer who saw your Reel on Monday and searched your name on Thursday gets counted twice, once by each. Meta will also count people who never clicked at all, only viewed.

The practical rule: your books are the source of truth. Platform numbers are useful for comparing campaigns within one platform, and misleading for anything else.

For a small business, the cleaner approach is blended: total marketing cost for the month against total gross profit for the month, tracked as a trend. It won't tell you which channel deserves credit, but it will tell you the truth about whether the whole effort is working — which is the question that actually matters. Deciding where that money goes in the first place is a separate exercise, covered in Google Ads vs Meta Ads for Indian small businesses.

What to Track, and What to Ignore

Ignore	Track instead
Followers, likes, impressions	Enquiries per week, by source
Click-through rate on its own	Cost per enquiry
Platform-reported conversions	Enquiries that became customers
Website traffic totals	Gross profit per rupee spent
Reach and views	Repeat purchase rate

Two of those deserve emphasis. Cost per enquiry is the number that tells you whether advertising is working; cost per click tells you almost nothing. And the share of enquiries that become customers is usually where the real problem hides. If a hundred enquiries produce four customers, the fix is in how you follow up, not in your bidding — and if the enquiries are dropping off on your website rather than in conversation, 10 signs your website is costing you customers is the more useful diagnosis.

Give Each Channel Its Right Timeframe

Judging every channel monthly is how good marketing gets cancelled.

• Paid ads: ninety days. The first few weeks are a learning phase and will look worse than the truth.
• Local visibility and SEO: six to nine months, though a properly completed Google Business Profile can produce calls within weeks. Money spent in month one shows up as enquiries in month five, so an in-month ROI calculation on it will always read as a loss.
• Content and organic social: six to twelve months, and the return is partly in cheaper paid performance later, which no report will attribute to it.
• Email and WhatsApp to existing customers: almost immediate, and usually the cheapest revenue in the business.

While you wait on the slow channels, watch leading indicators instead: enquiry volume, the share of people finding you by category rather than by name, and how often customers say they'd already heard of you.

Don't Measure Only the First Sale

A salon, a clinic, a tuition centre, a kirana, a tailor — these businesses live on repeat custom, and first-purchase ROI dramatically understates what a customer is worth.

Work out roughly what an average customer spends with you over a year rather than in one visit. A first sale of ₹800 looks like a poor return against a ₹600 cost per customer; the same customer returning six times a year does not. This is the calculation that decides whether you can afford to keep advertising at all, and it's the one small businesses least often do.

It also reframes where the money should go. If retention is where the value is, budget spent on staying in touch with existing customers usually beats budget spent finding new ones — the argument behind the organic approach in growing a local business on Instagram without paid ads, and the reason a well-run email and WhatsApp programme tends to show the strongest numbers in the whole account.

The One-Page Monthly Review

Everything above fits on a single sheet you can fill in on the first of each month.

• Total marketing cost, including GST, fees and production.
• Enquiries received, split by source from your register.
• Cost per enquiry, overall and by source.
• Customers won, and total value.
• Gross profit on that value.
• Gross profit minus marketing cost — the actual answer.
• One line on what you'll change this month.

Fifteen minutes. Six months of that sheet is a better decision-making tool than any analytics platform, because it contains the offline sales the platforms can't see, and because you'll actually read it.

Be Sceptical of Published Benchmarks

You'll see figures quoted everywhere: email at thirty-six to one, SEO at five to twelve times, average returns of three to five times. Treat them carefully.

Most are recycled between blogs without a traceable source, many are years old, and nearly all are computed on revenue rather than margin — which, as the table above shows, can invert the conclusion. They are also averages across industries with wildly different economics.

Your own three-month history is worth more than any published benchmark, because it reflects your margins, your city, your category and your ability to follow up. Benchmarks are useful for one thing only: noticing when your numbers are wildly out of line and asking why.

When the Numbers Say Stop

Measurement is only useful if you're willing to act on it. Reasonable rules:

• If cost per enquiry has stayed above what a customer is worth for three months, and the conversion rate isn't the problem, stop that channel rather than adding budget to it.
• If enquiries are plentiful but few convert, the marketing is working and the business isn't. Fix the follow-up before touching the campaigns.
• If you cannot tell where enquiries come from, fix the tracking before spending another rupee. Spending blind is the most expensive option available.
• If a channel is genuinely profitable, put more into it before diversifying. Small businesses lose more to spreading thin than to backing one thing.

Revenue divided by ad spend is a marketing metric. Gross profit minus total cost is a business one. Only the second tells you whether to keep going.

Frequently Asked Questions

What is the correct formula for marketing ROI?

Gross profit generated minus total marketing cost, divided by total marketing cost. The two mistakes that make most published figures meaningless are using revenue instead of gross profit, and counting only ad spend instead of the full cost including GST, agency or freelancer fees, tools and discounts given away. Calculated properly the number is usually far smaller than the multiples quoted in advertising.

What ROAS do I need just to break even?

Roughly one divided by your gross margin. At a 25 per cent margin you need about four rupees of revenue for every rupee of ad spend simply to cover the cost of the goods and the advertising, before any fees or overheads. At a 60 per cent margin the break-even point is closer to 1.7. This single number tells you whether a campaign is working better than any dashboard does.

How do I track ROI when customers contact me on WhatsApp or just walk in?

Use separate contact routes for separate channels. A distinct click-to-WhatsApp link per campaign, a different phone number in your ads, a code word or coupon per platform, and a simple enquiry register where every enquiry is logged with its source. Combined with asking each customer how they found you, this covers most of what analytics cannot see for an offline business.

Why do Google and Meta both claim the same sale?

Each platform counts a conversion if it had any qualifying contact with the buyer, so a customer who saw an Instagram ad and later searched your name gets claimed by both. Add up platform-reported sales and the total will exceed what your books show. Trust your own sales record as the single source of truth and use platform numbers only to compare campaigns within the same platform.

How long before I can judge ROI on SEO or content?

Judge paid advertising over about ninety days and channels like SEO, content and organic social over six to nine months. Money spent on those in month one produces enquiries in month five, so an in-month calculation will always look like a loss. Track leading indicators such as enquiry volume, share of unbranded searches and repeat customers while you wait.