Conversion Rate: How to Calculate and Improve It by Channel

An online store celebrated a 1.2% conversion rate the same month a competitor was worried about the exact same 1.2%. Both were right. The first sells electronics, a category where the Brazilian market average sits at 0.5%. The second sells cosmetics, where 2.6% is typical. Same number, different context.

That is the central problem with conversion rate: it is the most quoted metric in marketing meetings and one of the most misread. Calculating it is easy. Knowing what to compare it against, where to measure it, and how to react when it drops is what separates analysis from guesswork.

Conversion rate is the percentage of people who complete a desired action out of everyone who had the chance to complete it. The formula: conversions divided by total visitors (or leads, or opportunities), multiplied by 100. A store with 12,000 sessions and 156 orders has a 1.3% conversion rate.

How to calculate conversion rate in practice

The formula is simple, but every word in it matters:

The numerator is whatever conversion you defined: paid order, captured lead, quote request, signed contract. The denominator is where most people slip. Sessions, unique visitors, clicks, and leads are different bases and produce different rates for the same business.

Here is a funnel with real numbers: a SaaS gets 10,000 visits in a month, captures 400 leads (4% visit-to-lead), qualifies 80 opportunities (20% lead-to-opportunity), and closes 24 contracts (30% opportunity-to-sale). End-to-end conversion is 24 ÷ 10,000 = 0.24%. Those four rates tell four different stories, and each one is "the conversion rate" depending on who is talking. So before debating the number, agree on the numerator and the denominator. For a deeper stage-by-stage view, see our guide on how to build and measure a sales funnel.

What is a good conversion rate?

It depends on the industry and the channel, and the spread is wider than most people assume. The Prax 2025 report, which analyzed over 1,000 Brazilian e-commerce operations with monthly revenue above R$50,000 (about US$9,000), found session-to-order rates ranging from 0.5% in electronics to 5.7% in cleaning products, with beauty at 2.6% and food and beverage at 3.3%. That is a 10x gap between categories, all of them healthy in their own context.

Nuvemshop's NuvemCommerce 2026 report adds a cut that often goes unnoticed: stores billing above R$100,000 a month convert at 1.17% on average, while smaller ones sit at 0.75%. Mature operations convert better because they accumulate qualified traffic, social proof, and reliable logistics.

In paid media the numbers change scale because the denominator changes: the base is the click, not a session from any source. WordStream's 2026 benchmark, covering more than 13,000 Google Ads search campaigns, shows an average conversion rate of 8.18%, ranging from 2.64% in finance and insurance to 16.22% in the pet niche. Compare against those figures only if your definition of conversion matches (usually a lead or on-site action, not a sale).

Context	Average rate	Source
Brazilian e-commerce, electronics (session → order)	0.5%	Prax 2025
Brazilian e-commerce, beauty (session → order)	2.6%	Prax 2025
Brazilian e-commerce, cleaning products (session → order)	5.7%	Prax 2025
Brazilian e-commerce, stores above R$100k/month	1.17%	NuvemCommerce 2026
Google Ads search, overall average (click → conversion)	8.18%	WordStream 2026

In practice, an external benchmark answers exactly one question: am I in a different order of magnitude? If your category converts around 2.6% and you sit at 0.4%, something structural is wrong. If you are at 2.1%, the benchmark has done its job; from there, the comparison that matters is against your own history.

Why comparing conversion rates across channels is a trap

"Instagram converts at 0.8% and Google at 4%, so let's cut Instagram." That logic breaks for two reasons. First, the denominators are rarely comparable: a bottom-of-funnel paid search session is not the same as a discovery session on social. Second, channels play different roles: one generates demand, the other harvests it. Cut the discovery channel and, weeks later, the conversion of the channel that "converted well" tends to drop, because it was harvesting what the other one planted.

The honest use of per-channel conversion is tracking each channel against itself over time and crossing it with cost: CPA and ROAS close the math that conversion rate alone cannot. On that note, see which paid traffic metrics to track and how to calculate your break-even ROAS.

My conversion rate dropped: how do I diagnose it?

A conversion drop triggers panic and, quite often, the wrong reaction: rebuilding the page, firing the agency, changing prices. Before touching anything, follow a checking order that goes from cheapest to most expensive to investigate:

Step 2 deserves an example, because it fools experienced analysts. Suppose your site converts at 3% on desktop and 1% on mobile. With traffic split evenly, the blended rate is 2%. If a new campaign pushes mobile to 80% of visits, the blended rate falls to 1.4% without any page getting worse: 0.2 × 3% + 0.8 × 1% = 1.4%. The drop is real in the aggregate and nonexistent in every segment. You only see this by breaking the rate down by device, channel, and category before drawing conclusions.

How to improve conversion rate in each channel

On your site or store

The gains usually live at the bottom of the funnel, not the top. In the Prax data, sectors with an efficient checkout convert over 70% of shoppers who start payment, while others lose half of their buyers at that stage. Prioritize in this order: payment friction (local payment methods working flawlessly), shipping costs shown early, mobile speed, and social proof near the buy button. Test one change at a time; three simultaneous changes produce a result that teaches nothing.

In paid traffic

The main gain comes from coherence between ad, audience, and page. An ad that promises one thing pointing to a page that delivers another kills conversion before any bid optimization can help. After that, work on negative keywords, intent-based targeting, and campaign-specific landing pages instead of sending everything to the homepage.

In sales (leads and CRM)

If you sell through a sales team, response time is the cheapest variable to attack: a lead answered within minutes converts far better than one answered the next day. Set a first-contact SLA, clear qualification criteria, and measure the rate per stage, not just the final one. Sales conversion connects directly to your marketing KPIs at each funnel stage: if marketing hands over poor leads, the sales rate drops through no fault of the rep.

Frequently asked questions

What is a good conversion rate for e-commerce?

In Brazil, the average hovers between 1% and 1.5%, but the healthy range goes from 0.5% in electronics to over 5% in frequent-repurchase categories, according to the Prax 2025 report. Global averages sit around 1.4% to 2%. Benchmark against your industry first, then against your own history, which is the most reliable reference.

How do I calculate conversion rate in a spreadsheet?

Divide the conversions column by the visitors column and format it as a percentage: =B2/A2. To track by channel, build one row per traffic source and calculate each rate separately, plus the overall rate. Never average percentages across channels: always recalculate from the absolute numbers.

Are conversion rate and CTR the same thing?

No. CTR measures how many people clicked your ad out of those who saw it; conversion rate measures who completed the action after the click. An ad can have high CTR and low conversion, which usually means the promise is misaligned with the landing page.

What is the difference between session and user conversion rate?

Session conversion divides orders by total visits; user conversion divides by unique people. Since a buyer usually visits several times before purchasing, the user rate is always higher. Both are valid, as long as you never mix them in the same comparison.

How often should I monitor conversion rate?

Weekly for operations, monthly for decisions. Day to day, the rate fluctuates on noise (day of the week, weather, a one-off campaign), and reacting to every daily swing produces too much change with too little learning. Drops sustained for two weeks or more deserve the full diagnosis.

Measuring is easy; seeing is the hard part

Calculating a conversion rate takes ten seconds. The real work comes before and after: making sure measurement is intact, segmenting the number before reacting, and resisting the urge to compare rates with different denominators. Most "conversion crises" turn out to be traffic mix or a broken tag, and most real gains live in the checkout and in response time, not in a site redesign.

That work of breaking conversion down by channel, device, and period is exactly the kind of analysis that eats an analyst's afternoon when data lives in five different tools. With Sherlok, you connect Meta Ads, Google Ads, GA4, and your CRM, then simply ask: "which channel dragged my conversion down this month?". The answer comes ready, with the data already cross-referenced, so you spend your time on the decision, not the spreadsheet.