What is a website conversion:
definition, types and how to measure it
“Conversion” is a word you hear in every marketing meeting — and everyone uses it slightly differently. For a shop owner it’s a completed order. For a B2B company a submitted enquiry. For SaaS a trial signup. Without a clear definition, though, there’s nothing to measure and nothing to improve.
This article explains what a conversion actually is, what types exist, and how to set it up and measure it on your own site.
The simple definition
A conversion is the moment a visitor takes the action you want from them. Nothing more, nothing less. A visitor arrives as an anonymous number in your analytics, and the conversion turns them into a lead, a customer, or at least someone who took a step in the right direction.
What “the action” is, each site defines for itself — based on how it makes money:
- E-commerce: a completed and paid order
- B2B services: a submitted enquiry or a booked call
- SaaS: a trial signup or an activated paid subscription
- Local services: a phone call or a submitted booking form
Without that definition written down, conversion optimization happens blind — everyone on the team pictures something different by “better conversion”.
Conversion rate: turning the conversion into a number
A conversion on its own is a single event. To work with it, you relate it to traffic — which gives you the conversion rate:
conversions ÷ visits × 100 = conversion rate (%)
If 2,000 people visit in a month and 40 submit an enquiry, the rate is 2 %. As a rough frame, B2B sites tend to land between 1–3 %, e-commerce usually lower, local services with clear intent higher. Those numbers are orientation only — they matter mostly compared with your own site over time, not with someone else’s industry.
The work of making that number grow is called conversion rate optimization (CRO) — typically incremental changes to the site (headlines, forms, CTAs) verified with an A/B test.
Micro and macro conversions — why one number isn’t enough
Not every conversion carries the same weight. There are two types:
Macro conversion — the site’s main goal
The one the business actually cares about, directly tied to revenue: a paid order, a signed contract, an activated subscription. A site usually has one, at most two.
Micro conversions — steps toward the main goal
Smaller actions showing the visitor is moving toward the macro conversion even though they haven’t reached it: adding to cart, downloading a price list, subscribing to a newsletter, clicking through to pricing, opening the chat.
Why they matter: if you track only the final conversion, a site with modest traffic waits months for a sample large enough to conclude anything. Micro conversions react faster and show where the journey breaks — before it shows up in revenue.
The whole sequence from first visit to macro conversion is the conversion funnel. Micro conversions are its floors — and that’s usually where the answer to “where and why do people drop out” is hiding.
Quick glossary
| Term | What it means | Example |
|---|---|---|
| Conversion | A single visitor action you want | A submitted enquiry |
| Conversion rate | Share of conversions in total traffic | 2 % |
| Macro conversion | The site’s main business goal | A paid order |
| Micro conversion | A step toward the macro conversion | Opening chat, downloading pricing |
| Conversion funnel | The whole sequence from visit to macro conversion | Homepage → offer → form → confirmation |
| CRO | Systematically raising the conversion rate | Rewriting a CTA and verifying with an A/B test |
How to measure it properly
Definitions and formulas are useless if the conversion doesn’t exist in your analytics as a measurable event. Three steps:
1. Pin down the exact moment of conversion. Not “someone orders”, but a specific page or event — typically a thank-you page after form submission, or an order confirmation.
2. Set up the measurement. In GA4 that means creating a key event tied to that page view or form submission. Without it, the conversion rate lives in your head, not in your data.
3. Verify the event actually fires. Submit your own form in test mode and check in GA4 DebugView that the event was recorded. Broken measurement is the single most common reason companies see a distorted or zero conversion rate — the problem isn’t the site, it’s that conversions aren’t being measured at all.
What a conversion is not — three common mistakes
Traffic isn’t conversion. A site with hundreds of thousands of pageviews can have zero business effect if nobody takes the step that brings money. Traffic is the funnel’s input, not its output.
Engagement isn’t conversion. Long time on page or clicks between subpages look good in analytics, but on their own they sign nobody up and sell nothing. Engagement can precede a conversion; it doesn’t replace it.
A higher conversion rate doesn’t automatically mean more money. If the site brings cheap, badly targeted leads who never pay, the rate rises but revenue doesn’t. Conversion rate belongs next to the quality of what it brings in — not on its own.
Why bother with conversion at all
Raising the conversion rate is the cheapest way to get more customers out of a website — it needs no extra traffic and no extra ad spend, just better work with the traffic you already have. A site with 2,000 monthly visits at 1 % gives you 20 customers. The same site at 2 % gives you 40 — without another cent of acquisition.
Once conversion is clearly defined and measured, the next step is finding where visitors drop out and why. That’s covered in detail in how to increase your website conversion rate.
The takeaway
A conversion is a concrete, measurable action you want from a visitor — not a vague sense that “the site works”. Before optimizing anything, define your macro conversion, add a few micro conversions, and verify both are genuinely measured.
Only with that in place does it make sense to ask how to raise the rate. If you don’t know where your conversions actually leak, Kaen finds those places continuously — it reads your data, spots where the funnel drops off, and proposes the test.