In short: Conversion rate optimization increases the share of visitors who sign up, apply, fund, buy or book, through hypothesis, testing and measurement. It needs enough traffic to test, the conversion the business actually runs on, and the right visitors. If the visitors were never going to buy, the problem is upstream.
Key points
- Redesigning a dated page, or declaring victory on a hundred visitors, is not CRO.
- At a three percent conversion rate, detecting a meaningful improvement can need thousands of visitors per variation.
- Optimising sign-ups when the business runs on funded accounts can raise sign-ups and lower funding rates.
- Gains are rarely in the button; usually in message match, next-step clarity, unanswered doubts, form length and mobile speed.
- For restricted-category companies, each percentage point of conversion is a channel that does not have to be built.
Conversion rate optimization, CRO, is the practice of increasing the share of visitors who do the thing a page or product asks them to do: sign up, apply, fund, buy, book. It is one of the few marketing disciplines where the same traffic can produce more customers, which is why it is attractive, and why it is oversold.
What it involves
Understanding why visitors do not convert: analytics to see where they leave, session recordings and heatmaps to see how they behave, surveys and interviews to hear what stopped them, and support tickets to see what confused them. Forming a specific hypothesis: visitors abandon the application at the identity step because they do not know why the information is needed. Changing one thing. Testing the change against the original with enough visitors to tell the difference. Keeping what wins, discarding what does not, and recording both.
The discipline is in the hypothesis and the measurement. Redesigning a page because it looks dated is not CRO. Changing a button color and declaring victory on a hundred visitors is not either.
What it needs to work
Traffic. A test needs enough visitors to detect a real difference. For a page converting at three percent, detecting a meaningful improvement can require thousands of visitors per variation. Pages with a few hundred visitors a month cannot be tested statistically; they can be improved by judgment and observation, which is a different, honest activity.
A defined conversion. The event being optimized has to be the one that matters. Optimizing sign-ups when the business runs on funded accounts can raise sign-ups and lower funding rates.
Patience. Tests run for full weeks to include weekday and weekend behavior, and they run to their planned sample size, not until the result looks good.
Where the gains usually are
Rarely in the button. Usually in: the match between what the visitor was promised and what the page says; the clarity of what happens next; the doubts left unanswered; the length and order of forms; the speed of the page on a phone; and, for regulated products, the moment the visitor is asked for identity or money without being told why.
The largest single improvement most companies can make is answering the question the visitor actually has, at the point they have it.
When it is worth doing
When there is enough traffic to test, when the conversion being optimized is the one the business runs on, and when the traffic is the right traffic. A company with qualified visitors and a leaking funnel gets more from CRO than from any new channel, because every improvement compounds across all future traffic.
When it is a distraction
When the problem is traffic volume or traffic quality. A page converting badly because the visitors were never going to buy cannot be optimized into converting; the fix is upstream, in the channel and the audience. And when the company has so little traffic that testing is impossible; there, the work is observation, judgment, and copying what already works elsewhere, not a testing program.
For companies with rented traffic
For a company in a restricted category, where every visitor arrived through a channel that took work to open, conversion rate is worth more than it is to a company that can buy traffic freely. Each percentage point is a channel that does not have to be built. CRO is one of the cheapest investments available to these companies and one of the least made.
lowob takeaway: CRO turns the same traffic into more customers through hypothesis, testing, and measurement. It needs enough traffic to test, the right conversion to optimize, and the right visitors; otherwise the problem is upstream.
Common questions
What is conversion rate optimization?
The practice of increasing the share of visitors who do what a page or product asks, such as sign up, apply, fund, buy or book, by forming a hypothesis, changing one thing and testing it against the original.
How much traffic do you need for A/B testing?
Enough to detect a real difference. For a page converting at three percent, that can mean thousands of visitors per variation; pages with a few hundred visitors a month cannot be tested statistically.
When is CRO a waste of time?
When the problem is traffic volume or quality, or when traffic is too low to test. Then the work is fixing the channel and audience, or improving by observation and judgment.
Where do the biggest conversion gains usually come from?
Answering the visitor's actual question at the point they have it: matching the promise, clarifying the next step, addressing doubts, fixing forms and mobile speed, and explaining why identity or money is requested.