Conversion rate is the percentage of visitors who take a desired action. That action could be a purchase, a form submission, a phone call, or anything else you define as a conversion. The formula is simple: conversions divided by total visitors, multiplied by 100. What makes it complicated is deciding what counts as a conversion, which visitors to include in the denominator, and over what time period. Those decisions shape the number more than most people realise.
Conversion rate is the multiplier that sits between your traffic and your revenue. Double your conversion rate and you double the value of every pound spent on acquiring traffic, without spending a penny more on ads. It is also one of the clearest indicators of whether your marketing and your product are aligned. A low conversion rate on high-intent traffic usually means your offer, your page, or your buying experience is broken. That is a problem worth fixing before you spend more on driving traffic to it.
You measure conversion rate by dividing the number of completed actions by the number of sessions or users, depending on your preferred method. Session-based conversion rate counts each visit as a separate opportunity. User-based conversion rate counts unique people, regardless of how many times they visited. Neither is wrong, but mixing them will give you misleading trends. For most businesses, tracking conversion rate by traffic source, device type, and landing page gives you enough granularity to make informed decisions about where to focus your optimisation efforts.
The most common mistake is reporting one blended conversion rate for your entire site and treating it as meaningful. A 2% average conversion rate tells you almost nothing if your branded traffic converts at 8% and your prospecting traffic converts at 0.5%. The aggregate hides the signal. Another frequent error is running conversion rate optimisation tests without enough traffic to reach statistical significance. We have seen businesses overhaul their entire checkout flow based on an A/B test that ran for three days on 200 visitors. That is not data; it is noise.
Direct answers to the conversion rate questions we hear most in discovery calls.
The question is too broad to answer usefully without context. Conversion rates vary dramatically by industry, traffic source, device, and what you define as a conversion. A 1% purchase rate on cold paid social traffic can be excellent. A 5% form submission rate on branded search might be poor. The only benchmark that matters is your own historical performance, segmented by the variables that affect it.
Start with your highest-traffic, lowest-converting pages and look for friction. Slow load times, unclear calls to action, too many form fields, and poor mobile experience are the usual suspects. Fixing these costs far less than a redesign and often delivers faster results. Copy changes alone can move conversion rates significantly if the current messaging does not match the intent of the traffic arriving on the page.
Fix your conversion rate first if it is below your break-even threshold. Driving more traffic to a page that does not convert is just spending more money to lose more money faster. Once your conversion rate is at a level where your unit economics work, then scaling traffic becomes the growth lever. The right answer is usually both, but in sequence, not simultaneously.
We start with your data, not assumptions. That means a proper analytics audit to make sure conversion tracking is accurate, followed by a segmented analysis to find where the real problems are. From there, we build a testing roadmap prioritised by potential revenue impact and traffic volume. Our goal is to transfer the entire CRO process to your team, including the analytical framework, not just the test results.
It depends entirely on your traffic volume. High-traffic sites can run meaningful tests and see validated results within two to four weeks. Lower-traffic sites might need six to eight weeks per test to reach statistical significance. Beware of anyone promising CRO results in a week. Either they have enormous traffic to work with, or they are reading results before the data justifies it.
