Conversion rate

The formula, and the denominator nobody agrees on

Conversion rate = conversions ÷ some population, times 100. The arithmetic is trivial. The argument is about the population.

All three denominators below are legitimate. Mixing them inside one report is not, and it's the single most common way a monthly deck flatters itself.

• Sessions — the GA4 default. Best for comparing landing pages and channels. A visitor who returns three times before enquiring drags this down.
• Users — best for long consideration cycles, which is most B2B and most high-ticket services in India. Always higher than the session figure.
• Clicks or ad impressions — used in paid reporting, where the denominator is what you paid for.

The Indian measurement gap: calls and WhatsApp

Here's the thing nobody documents properly. On a typical Indian service or D2C site, the contact form is the third-choice channel. People tap the WhatsApp button. People tap the phone number. Both leave your website instantly and neither shows up in a default analytics setup.

The result is a founder looking at a 0.4% conversion rate, concluding SEO doesn't work, while the sales team is drowning in WhatsApp enquiries nobody attributed to anything.

Three fixes, in order of effort.

1. Track the outbound clicks. Every wa.me or api.whatsapp.com link click and every tel: link click should fire a GA4 event. This is a tag manager trigger and half an hour of work. It won't tell you what the conversation produced, but it tells you which pages start conversations.
2. Put the page in the message. Pre-fill the WhatsApp text with the page or product name — wa.me/91XXXXXXXXXX?text=Hi, I'm enquiring about the Bangalore audit. Now sales can see the source in the chat itself, no attribution model required.
3. Use a call tracking number if calls are the main channel. A separate number for organic visitors closes the loop properly. It costs money and it's worth it above roughly 50 calls a month.

Micro versus macro conversions

A macro conversion is the thing that pays salaries: a purchase, a qualified enquiry, a booked call. A micro conversion is the step before it.

You need both. Macro conversions are too rare on a young site to steer by — at 20 leads a month, one good week looks like a 40% improvement and means nothing. Micro conversions give you enough volume to see whether a page is working before the revenue arrives.

• Macro — order placed, form submitted, call connected, demo booked, quote requested.
• Micro — pricing page viewed, WhatsApp click, phone number tapped, PDF downloaded, size guide opened, 75% scroll on a service page.
• Neither — newsletter pop-up dismissed, 10-second timer, any event you invented to make a dashboard look busy.

What good looks like, and why benchmarks lie

Every published benchmark averages across countries, industries, traffic mixes and definitions you can't see. Use the bands below as planning ranges to sanity-check a model, never as a target to be held to.

The ranges are wide for a real reason: conversion rate is mostly a function of intent, and intent is a function of which queries you rank for. A page ranking for buy X in Bangalore and a page ranking for what is X convert an order of magnitude apart while sitting on the same site.

Why untracked traffic is unmanageable

Without conversion tracking, every decision becomes an opinion. You can't tell whether the article that brought 4,000 visits brought a single customer, so you optimise for the visits — and organic traffic is the easiest metric on earth to inflate with queries nobody buys from.

It also makes an agency unaccountable, which is why so few insist on it. We won't start work without it: the guarantee runs against your trailing-90-day qualified leads from organic, frozen on day one, so if that number doesn't exist there is nothing to guarantee. Miss the baseline in 90 days and we keep working free until we beat it. That contract only functions if the counting is honest — which is why conversion tracking setup happens in week one, not month four.