How to advertise your website: the channels that pay back

This is one of the most common questions we get from founders, and it is almost always the wrong first question. Advertising is not something you do to a website. It is something you do to a specific offer, aimed at a specific buyer, measured against a specific payback period. Get those three right and almost any channel works. Get them wrong and no channel does.

What follows is the sequence we use when a brand asks us where to spend. It is written for Indian businesses selling something, not as a generic global checklist.

Before you spend a rupee, four things have to be true

Most wasted ad budget in India is not wasted on the wrong channel. It is spent before the business was ready to absorb traffic. Check these first.

• You know what a customer is worth. Not revenue. Contribution after product cost, shipping, payment charges, returns and the discount you actually gave. If you cannot state that number, you cannot judge any campaign.
• Tracking exists and you trust it. Conversion events firing correctly, orders attributable to a source, and a way to see repeat purchase. Advertising into a broken measurement setup buys opinions, not data.
• The destination converts. Sending paid traffic to a page that converts poorly is the most expensive way to discover it converts poorly. Fix the page first. The specifics are in landing page conversion optimisation.
• You can fulfil the demand. Stock, serviceable pin codes and support capacity. Advertising into a stockout trains customers to leave.

If any of those four is missing, the highest-return advertising decision available to you is to not advertise this week.

The channel map, and what each one is actually for

Indian brands have more channels than budget. The useful way to think about them is by the job each does, because they are not substitutes for one another.

Search advertising: harvesting demand that already exists

Someone is already looking for what you sell, and search captures that intent. It is the fastest way to learn whether your offer converts, because the traffic arrives pre-qualified. It is also capped. You cannot buy more demand than the category has. If your search volume is thin, search will look efficient and stay small.

Social advertising: creating demand that did not exist

Nobody woke up looking for your product. Social interrupts, so creative carries almost all of the performance. The biggest lever is not targeting, it is how many genuinely distinct creative concepts you can test each month. The testing discipline is covered in Meta ads creative testing.

Marketplace advertising: buying position where the transaction happens

If you sell on Amazon, Flipkart, Myntra, Nykaa or Meesho, the ad auction sits inches from the buy button. Intent is highest and the path to purchase is shortest. It is also the channel brands most often underestimate structurally, because campaign architecture sets your ceiling long before bidding does. See Amazon PPC campaign structure, marketplace bid strategy, and for Meesho, promoted listings ROI.

Quick commerce retail media: the fastest-inflating auction

Blinkit, Zepto, Instamart and Flipkart Minutes now sell placement much as marketplaces do, and this is the channel changing fastest. Our own reporting this month found brands raising festive quick commerce ad budgets 25 to 35 percent year on year, festive ad rates up 30 to 50 percent, and cost per click moving from roughly Rs 5 to 10 into the Rs 7 to 15 range in competitive categories. The mechanics are in the Blinkit ads guide, the pacing problem in daypart ad pacing, and the category background in what is retail media.

Search visibility you do not pay per click for

Organic search is not advertising, but it competes for the same budget line and changes what advertising has to carry. It is slower, it compounds, and it is increasingly complicated by AI answers. If you are weighing it against paid, what SEO actually costs in India and how to choose an SEO company are the honest starting points.

Owned channels: the cheapest media you already have

Email, WhatsApp and your own app cost almost nothing per send, and they are the only audience a platform cannot take away from you. Most brands advertise hard to acquire a customer and then never speak to them again. That is a discipline problem, not a media problem.

How to choose your first channel

Do not run four channels at once on a small budget. You will learn nothing from any of them. Choose using the shape of your business.

• People already search for your category and you sell direct: start with search. Prove conversion, then widen.
• Your product is visual, impulse-led or new to the market: start with social, and treat creative volume as the real budget line.
• Most of your sales already happen on a marketplace: start with marketplace ads. You are closest to the transaction and the data is cleanest.
• You sell food, beverages, beauty or daily essentials: quick commerce retail media is where your category is being decided right now.
• You sell a considered, high-value service: search plus owned channels, and expect a longer sales cycle than any dashboard will show you.

Run one channel properly for long enough to reach a decision, then add a second. At small budgets, sequential beats simultaneous every time.

What to spend

There is no correct percentage, and anyone quoting you one has not seen your numbers. Budget is an output of two things: what a customer is worth to you, and how long you can wait to get that money back.

Work backwards. Decide the most you can pay to acquire a customer while staying viable at your payback period. That is your ceiling, not your target. Then size the first test so it produces a decision within a few weeks, and so it does not hurt if the answer is no. A test that runs too small runs forever and never concludes.

The payback question is the one founders skip and the one that kills brands. It is covered properly in CAC payback period.

Measuring it without fooling yourself

This is where advertising money is genuinely lost, and none of it is glamorous.

Blended numbers hide the truth. A healthy blended cost of acquisition can conceal one channel quietly subsidising another, and usually does. The failure mode is set out in why blended CAC is lying to you.

Platform-reported returns are not incremental. Every platform claims the conversions it touched, and some of those customers would have bought anyway. The only way to know what your advertising caused is to withhold it somewhere and compare. The method is in geo holdout testing.

Judge on contribution, not revenue. A campaign showing a strong return on ad spend can still lose money once returns, shipping and discounting are counted. Put the ad cost into the same profit and loss as everything else.

The mistakes that waste a first budget

• Changing everything at once. New creative, new audience and a new bid on the same day means you learn nothing from the outcome.
• Stopping too early. Most accounts are killed before they have enough conversions to be judged. Decide the evaluation window before launch and hold to it.
• Discounting to make the numbers work. If the offer only converts at a discount, you have found a pricing answer, not a media one.
• Buying awareness before you can convert. Awareness is a real objective. It is the wrong first objective for a brand that has not proven it can sell.
• Optimising to the wrong event. Optimise to clicks or add-to-cart and you will buy exactly those. Optimise to the event that pays you.

Do it yourself, hire, or both

Run it yourself while the budget is small and the learning matters more than the efficiency. You will come to understand your own numbers in a way no report gives you.

Bring in help when the constraint stops being knowledge and becomes execution volume: more creative than you can produce, more channels than you can watch daily, or a marketplace account where structure and daily management decide the result. When you do hire, buy against a defined scope and a measurement standard you set yourself. We wrote about doing that well in scoping an agency engagement.

The short version

Advertising your website is not a channel problem. Know what a customer is worth, make sure the destination converts, pick one channel that matches how your category is actually bought, spend enough to reach a decision, and measure incrementally rather than believing the dashboard. Do that and the channel question mostly answers itself.