Is Your Website Working? Count Enquiries, Not Visits

Most owners judge a website the way they judge weather: a general feeling, occasionally checked against a number nobody fully trusts. The monthly report says visits are up eleven per cent. The phone has not rung any more than usual. Both things are true, and neither tells you whether the money you spent on the site was well spent.

The fix is not a better dashboard. It is deciding, in one sentence, what the site is supposed to produce — and then measuring that one thing per visitor rather than in total.

Decide what the site is for first

Before you open any analytics tool, finish this sentence: this website has done its job when a stranger ______.

For most Indian small and mid-sized businesses the answer is one of a handful of things, and they are not interchangeable.

Business	The site has worked when a stranger…
Interior contractor	sends a WhatsApp message or fills the site-visit form
Chartered accountant	books a call, or phones the number on the page
Manufacturer selling to dealers	downloads the product catalogue and leaves a company name
Local clinic	books an appointment, or taps the Google Maps link
Online store	completes a paid order
B2B services firm with a long cycle	joins the mailing list or asks for a proposal

Write yours down. If you cannot, that is your real problem, and no amount of traffic will solve it. A site with no defined action is a brochure, and a brochure's only measurable success is that someone read it — which you can neither verify nor bank.

One caution: pick a primary action and treat everything else as secondary. Sites that ask equally for a call, a form, a brochure download, a newsletter signup and a WhatsApp chat tend to get fewer of all five. Choose the one that actually starts a sale in your business, and let the rest sit quietly below it.

Enquiries per hundred visitors

Here is the number that replaces traffic: how many enquiries do you get for every hundred people who land on the site?

Total enquiries move for boring reasons — a festival, a competitor's ad budget, a holiday week. Enquiries per hundred visitors moves for interesting ones. It tells you whether the site is persuading people, independent of how many people turned up.

Two businesses illustrate why the ratio beats the total:

• Site A: 4,000 visits a month, 12 enquiries. That is 0.3 per hundred.
• Site B: 600 visits a month, 24 enquiries. That is 4 per hundred.

Site A's owner is being congratulated on traffic growth. Site B's owner has the better website by a wide margin, and the obvious next move — get more of the same kind of visitor. Site A's owner who buys more traffic is buying more of a problem.

Do not chase a benchmark figure here. The honest position is that "good" varies enormously by trade, by price point, by whether the visitor arrived from a search for your brand name or from a general query, and anyone quoting you an industry-standard conversion rate for Indian SMEs is quoting a number they cannot support. What matters is your ratio, measured the same way each month, moving in the right direction.

Counting enquiries honestly

This is where most measurement falls apart, because a large share of Indian enquiries never touch a form. They arrive as a phone call, a WhatsApp message, or a walk-in from someone who checked your address online.

Practical ways to catch them:

1. Make every click on your phone number and WhatsApp link a tracked event. In Google Analytics 4 — Google's current free analytics tool — outbound taps on tel: and wa.me links can be recorded as events. That gives you intent-to-contact, which is not the same as a contact but is far better than nothing.
2. Use a separate number on the website if you can manage it operationally. A dedicated line, or even a second WhatsApp Business number, makes attribution unambiguous. If that is too much, train whoever answers to ask one question: "May I know where you found us?" Log the answer.
3. Filter the junk. Many contact forms collect bot submissions and SEO spam. Count only enquiries a human would call back. If you have a CRM where enquiries land, mark them at the point of entry, not at the end of the month from memory.
4. Record the source. A hidden field that captures the referring channel, or tagged campaign links, means you can later separate "the website converted" from "Google Ads converted and the website got out of the way".

Then take it one step further. Of those enquiries, how many were worth answering? Ten enquiries where eight are students asking for internships is worse than three from people with budgets. Track enquiries and qualified enquiries separately, even if qualification is just a tick in a spreadsheet column. This is the point at which measurement stops being a reporting exercise and starts changing what you put on the site.

What a bad number is telling you

Say your ratio is poor. Three distinct things could be wrong, and they need different fixes.

Wrong people are arriving

Open Google Search Console — free, from Google — and look at the actual queries bringing impressions and clicks. If a company selling industrial pumps is getting clicks from people searching "pump price" meaning a cycle pump, the site is fine and the traffic is noise. The answer is sharper pages for the queries that matter, not a redesign. That is a search visibility job.

The right people arrive and leave

Check the page they land on against the thing they searched for. A visitor searching for a specific service who lands on your homepage has to do work to find the relevant page, and many will not. Also check speed honestly: Google publishes its Core Web Vitals thresholds — Largest Contentful Paint should be under 2.5 seconds, Interaction to Next Paint under 200 milliseconds, Cumulative Layout Shift under 0.1. A page that fails these on a mid-range Android phone over mobile data is losing people before they read a word. That is build and performance territory.

They read everything and still do not contact you

Usually the page never makes a specific promise, or it asks for too much. A form with eleven fields including GST number, at the bottom of a page, with no indication of what happens next, is a form designed to be abandoned. Name, phone, one line of context. Tell them when you will reply.

What not to buy

Some things sold under the heading of measurement are not worth your money, including things an agency like ours could happily bill for.

• A twenty-page monthly PDF report. If it does not contain enquiries, qualified enquiries, and a comparison against last month, it is a document about activity, not results. One page is enough.
• Heatmap and session-recording retainers on a low-traffic site. Heatmaps are genuinely useful at volume. On a site getting a few hundred visits a month they mostly produce anecdotes you will over-interpret. Buy them when you have enough traffic to see a pattern, not before.
• Bounce rate as a headline metric. It has always been easy to misread, and in GA4 the engagement definitions changed. A visitor who reads your pricing page fully and then phones you can be recorded unhelpfully. Do not let anyone build a story around it.
• Keyword rank reports as proof of value. Rankings vary by location, device and personalisation. They are an input, not an outcome.
• A redesign bought to fix conversions you have never measured. You may redesign away the only thing that was working.

A reasonable scoreboard

For most businesses, five numbers reviewed monthly are plenty:

1. Enquiries (forms, calls, WhatsApp, combined)
2. Qualified enquiries
3. Visitors
4. Enquiries per hundred visitors
5. Which three pages produced the enquiries

Number five is the one owners skip and then regret. It is very common for a single service page or a well-written FAQ to generate most of the contact, while the homepage carousel everyone argued about contributes nothing. Once you know which pages earn, you know where to spend.

Be patient with the data. If you receive a handful of enquiries a month, a single month's ratio is close to meaningless — one unusual week moves it. Three months is a trend. Judge accordingly, and resist the urge to change five things at once, because then you will not know which one worked.

What to do this week

Write down the one action that means your site has worked. Count last month's enquiries from your inbox, call log and WhatsApp — roughly is fine for the first pass. Divide by visitors, multiply by a hundred. You now have a baseline you did not have yesterday.

Then fix the counting so next month's figure is solid: tracked phone and WhatsApp taps, a shorter form, enquiries logged where you can see the source. If you want help setting that up properly, or you would like a second opinion on which part of the chain is leaking, tell us what your site is meant to produce and we will look at it against that definition rather than against traffic.

Questions we get asked