What it means
Every customer-acquisition process is a funnel: many people enter, fewer take each successive step, and a small number reach the end. Conversion rate measures the narrowing at each stage.
A retailer's website receives 100,000 visitors, of whom 3,000 add something to a basket, 1,200 start checkout and 900 complete an order. The conversion rates are 3.0% (visit to basket), 40% (basket to checkout), 75% (checkout to order) and 0.9% overall.
Each stage's rate shows where people are lost, and the business works on the stage with the largest loss or the cheapest fix. The measure applies wherever a sequence leads to a result: enquiry to qualified lead to proposal to win in business-to-business sales, visitor to free trial to paid in subscription products, application to approval to drawdown in lending, and application to interview to offer to acceptance in recruitment.
In each, the rates define the efficiency of the process and the cost of the result. Conversion optimisation, which means running continuous experiments on page layouts, offers, forms, pricing presentation and sales scripts, is therefore one of the highest-return activities in marketing.
The financial connection is direct: cost per result = cost of the input / (input volume x conversion rate). A campaign spending $20,000 to bring 10,000 visitors at a 2% conversion rate produces 200 customers at $100 each, and at 3% it produces 300 customers at about $67 each.
Improving conversion is usually cheaper than buying more traffic because the traffic has already been paid for, and in this example a 1-point improvement delivers the same as a 50% increase in traffic. Interpretation needs care because rates vary by channel (search visitors convert better than display visitors because they were looking), by device (mobile typically converts lower), by product, by customer type (returning customers convert far better than new) and by season.
A single overall rate averages these and hides them, so the useful analysis is segmented. Rates also trade off against volume and value: relaxing qualification raises the lead count and lowers the lead-to-sale rate, raising prices lowers conversion but may raise value per conversion, and a discount raises conversion and lowers margin, so the right target is the combination that maximises contribution, not the rate alone.
Definitions must be fixed: is a conversion a completed purchase or a started checkout, is a lead anyone who filled in a form or only those who met qualification criteria, and is the denominator visits or unique visitors? Businesses that measure conversion across teams and periods need a written definition for each stage, or the numbers will not be comparable and improvements will become arguments.
In practice
Real-world examples
Example 1
A software company's free trial converts to paid at 12%, and a change to the onboarding sequence raises it to 17%, adding 40% to new revenue at no acquisition cost.
Example 2
A mortgage broker tracks enquiry-to-application at 35% and application-to-completion at 60%, and finds that a call within one hour of enquiry raises the first rate to 48%.
Example 3
A charity's donation page converts 4% of visitors, and a test showing the impact of a specific gift amount raises it to 6.5%.
Think of it
Conversion rate shows what percentage of potential customers actually become customers-your closing success.
Formula
Calculation
Conversion Rate = Conversions / Opportunities x 100% (at each stage and overall)
Overall funnel conversion = Product of the stage conversion rates
Cost per Conversion = Total cost / Conversions = Cost per opportunity / Conversion rate
Value of a 1-point improvement = Opportunities x 1% x Contribution per conversion
Break-even conversion rate for a campaign = Cost per opportunity / Contribution per conversion
Worked example. An online retailer of outdoor equipment analyses its funnel for a month.
Traffic: 240,000 visits, from search advertising (60,000, cost $48,000), organic search (100,000, no direct cost), email (30,000, cost $2,000), social advertising (50,000, cost $25,000).
Overall: 4,800 orders; conversion rate 2.0%; average order $150; contribution per order (after product cost, shipping and payment fees) $45. Total contribution $216,000; total paid media cost $75,000.
By channel:
- Search advertising: 60,000 visits; 1,800 orders; conversion 3.0%; cost per order $26.67; contribution $81,000 against $48,000: return $1.69 per $1
- Organic: 100,000 visits; 2,100 orders; conversion 2.1%; no cost
- Email: 30,000 visits; 750 orders; conversion 2.5%; cost per order $2.67
- Social: 50,000 visits; 150 orders; conversion 0.3%; cost per order $166.67; contribution $6,750 against $25,000: a loss of $18,250
By stage (all channels, on consistent denominators): visit to product page 70%; product page to basket 9.5%; basket to checkout 50%; checkout to order 60%. Overall = 0.70 x 0.095 x 0.50 x 0.60 = 2.0%, matching the measured rate. The checkout-to-order rate of 60% (40% of people who start checkout abandon it) is the largest loss close to revenue, where each lost sale has already cost the most to attract.
Improvement programme:
- Checkout: remove mandatory account creation, add a guest option, show shipping cost earlier, add two payment methods. Test result: checkout-to-order rises from 60% to 72%. At the same traffic, orders rise by 20%: 4,800 to 5,760; contribution up $43,200 a month. Cost of the changes: $15,000 one-off.
- Social: the 0.3% conversion shows the audience is wrong for direct sales. The budget is cut to $8,000 for brand and retargeting, and the retargeting segment converts at 1.8%: 240 orders, cost per order $33. Saving $17,000 a month of spend and a loss turned into a small profit.
- Mobile: mobile visits (55% of traffic) convert at 1.4% against desktop's 2.7%. A mobile checkout redesign is scheduled; each 0.1 point of mobile conversion is worth 132 orders and $5,940 of contribution a month.
Break-even conversion for search advertising: cost per visit $0.80; contribution per order $45; break-even rate = $0.80 / $45 = 1.8%. At 3.0% the channel is well above break-even; the retailer raises its bid limits on the best-performing keywords, accepting a cost per visit up to $1.20 (break-even 2.7%) where conversion exceeds 3.5%.
After three months: overall conversion 2.5%; orders 6,000 a month on 240,000 visits; contribution $270,000; paid media $63,000. The improvement in conversion delivered $54,000 a month of additional contribution with $12,000 less spend, the equivalent of a 25% increase in traffic (60,000 more visits), which at the search channel's $0.80 a visit would have cost about $48,000 a month to buy.Case study
Seen in the real world
This is a fictional, illustrative scenario. A furniture retailer spent $1,200,000 a year on advertising to bring 1,500,000 visits to its website at a conversion rate of 1.1%, producing 16,500 orders at an average of $600. Its marketing director's plan for growth was more advertising. The finance director analysed the funnel and found that 70% of visitors who added a product to the basket abandoned before completing, that the abandonment concentrated at the delivery-options page (where the customer first saw a $79 delivery charge), and that 40% of the abandoners returned to the site within a week without buying.
The retailer tested including delivery in the price (raising the displayed price by $60 and removing the charge) on half its traffic. Basket-to-order conversion rose from 30% to 44% on the test group, average order value rose because of the built-in delivery, and overall conversion rose from 1.1% to about 1.6%. Applied across all traffic, orders rose to about 24,000 a year on the same advertising spend, an increase equivalent to roughly $545,000 of additional advertising. The finance director's note pointed out that the company had been paying to bring people to a page that sent them away, and that fixing the page had been worth more than any campaign.
Watch out
Common mistakes
- Increasing traffic before fixing conversion, which pays to bring more people to a process that loses most of them.
- Reading one overall conversion rate, which hides the differences by channel, device, product and customer type where the problems and opportunities are.
- Maximising the conversion rate alone. Discounts and relaxed qualification raise the rate and can lower contribution; the target is the combination of rate, volume and value that maximises profit.
Questions
People also ask
What is a good conversion rate?
It depends on the industry, channel and stage: 1% to 3% for e-commerce visit-to-order, 10% to 20% for trial-to-paid in software, 20% to 40% for proposal-to-win in business sales. Compare with the channel benchmark and the business's own history.
How does conversion rate relate to cost per acquisition?
Cost per acquisition is cost per opportunity divided by conversion rate. Improving conversion lowers acquisition cost directly, usually more cheaply than reducing the cost of opportunities.
How should conversion be improved?
By finding the stage with the largest loss, understanding why people leave (analytics, session recordings, surveys), testing changes on a portion of traffic, and adopting what works. Then the next stage.