CRO: raising conversion without buying more traffic

Conversion rate optimisation means getting more of the people already visiting you to do what you want them to do. It is the most profitable work in digital marketing, because it does not require buying more traffic: it requires stopping the loss of the traffic you already get.

The arithmetic is easy to see. Going from 1% to 1.5% conversion equals 50% more customers on the same advertising budget. Buying that same 50% costs half of what you already spend.

What is a good conversion rate?

The honest answer is that it depends so much that sector averages are useless. What is useful is comparing yourself against yourself.

Still, an order of magnitude so you do not despair:

Type of site	Usual range
Online shop	1% to 3%
B2B lead generation	2% to 5%
Campaign landing page	5% to 15%
Professional service with a form	1% to 4%

If you are far below, it is almost never the colour of the button. It is usually one of the four things in the next section.

Where do people actually drop off?

In four places, in this order of frequency:

1. They do not understand what you sell in the first five seconds. The number one cause and the cheapest to fix. If your headline says “We transform your business”, it says nothing.
2. They do not trust you. No address, no phone, no faces, no reviews. In Colombia this weighs heavily.
3. You ask for too much. A nine-field form for “more information” is a barrier, not a filter.
4. It does not work on a phone. Small buttons, impossible forms, endless loading on mobile data.

Before touching anything, open your own site on a phone using mobile data and try to buy something. The problem usually surfaces on its own.

Which changes actually move the number?

Ordered by impact, based on what I have seen:

Change	Typical impact	Effort
Headline that says what you do and for whom	High	Low
Removing form fields	High	Low
Adding WhatsApp as a contact option	High in Colombia	Low
Trust signals visible at the top	Medium high	Low
Publishing a price or price range	Medium high	Business decision
Mobile loading speed	Medium	Medium
Full redesign	Unpredictable	High

That last row is the one most often requested and the one with the worst risk-to-result ratio. A redesign changes fifty things at once and afterwards nobody knows which one worked.

Why does WhatsApp change the equation here?

Because in Colombia people prefer writing to filling in a form. A form is an act of faith: you do not know whether anyone will read it or when you will hear back. WhatsApp is immediate and visible.

With one important caveat: if nobody answers quickly, WhatsApp converts worse than the form, because the expectation is different. If you are going to add it, either someone is attending it or an assistant answers the basics and flags a human.

How do you test without fooling yourself?

Here is the most expensive technical error: most small-business A/B tests do not have enough traffic to conclude anything.

To detect a 20% improvement on a 2% base rate you need thousands of visits per variant. With 300 visits a month, any difference you see is chance. Declaring a winner there is worse than not testing, because you make a wrong decision confidently.

What to do without volume:

• Change one thing at a time and compare month over month, knowing it is a signal and not proof.
• Prioritise big changes, not nuances. A different headline, not a different shade of blue.
• Use qualitative methods: session recordings, heatmaps and above all watching five people use your site. Five users surface most of the serious problems.

What do you measure?

Not the site-wide rate, which mixes incomparable things. Break it down:

• By source. Search traffic and paid traffic convert differently. Averaging them hides the problem.
• By device. There is almost always a large mobile-versus-desktop gap, and that is where the opportunity sits.
• By entry page. A single bad page can drag the average down.
• Through to the sale, not to the form. A form bringing leads who never buy is not converting: it is generating work.

That last point has settled more arguments than any other in the companies I support.

Frequently asked questions

What is a conversion rate? The percentage of visitors who take the action you want: buying, leaving their details or writing to you. If 20 out of 1,000 visits write to you, your rate is 2%.

What is a good conversion rate? It depends so much on sector and traffic type that general averages are not useful. As a reference, an online shop usually sits between 1% and 3%. What helps is comparing yourself month over month.

How many visits do I need for an A/B test? Thousands per variant to detect moderate differences. With a few hundred visits a month, an A/B test concludes nothing and it is better to make large changes and observe.

Where do I start if I know nothing about this? Open your site on a phone using mobile data and try to buy from yourself or contact yourself. Note every point where you hesitate. That list is your plan.

Is adding WhatsApp to the site worth it? In Colombia, very much so, as long as someone answers quickly. If a message sits unanswered for two hours, it converts worse than a form.

How many fields should a form have? The minimum needed to reply. Usually a name and one way to reach them. Every extra field costs leads, and missing details can be asked for later in the conversation.

Does publishing prices help or scare people off? It reduces the number of enquiries and improves their quality. For a professional service it is almost always worth it, because it filters out people who were never going to buy.

Does a full redesign improve conversion? Not necessarily, and often it makes things worse at first. Changing everything at once destroys your ability to know what worked. A sequence of measured changes is better.