Conversion Rate Calculation: Stop Getting It Wrong

Your website conversion rate is calculated by dividing the number of conversions in a given period by the total number of visits, then multiplying by 100 to get a percentage. If your site received 10,000 visits in a month and generated 200 conversions, your conversion rate is 2%. That is the arithmetic. What makes it complicated is everything that sits around it: which visits you include, what counts as a conversion, and whether the number you end up with actually tells you anything useful.

I have sat in boardrooms where a marketing director has presented a 4.2% conversion rate with obvious pride, and nobody in the room thought to ask what it was calculated on, what the denominator included, or whether it had moved revenue at all. The number had become the goal rather than a signal pointing toward one. That is a habit worth breaking early.

What Is the Correct Formula for Monthly Conversion Rate?

The formula itself is straightforward:

Conversion Rate (%) = (Conversions / Total Visits) x 100

If you had 15,000 visits in March and 375 completed a purchase, your monthly conversion rate is 2.5%. If 450 completed a lead form instead of purchasing, your lead conversion rate is 3%. Same visits, different conversion event, different number. This is where teams start talking past each other, because they are often calculating rates against different events and calling them all “conversion rate.”

Before you run the calculation, you need to be precise about two things: what counts as a visit, and what counts as a conversion. Get either of those wrong and the output is meaningless, however accurately you do the division.

Sessions vs. Users: Which Should Be Your Denominator?

Most analytics platforms give you both sessions and users, and the choice between them changes your conversion rate meaningfully. Sessions count every visit, including multiple visits from the same person. Users count distinct individuals, typically based on cookies or logged-in identity. In practice, users is almost always the smaller number, which means using users as your denominator will produce a higher conversion rate than using sessions.

Neither is wrong. They are measuring different things. If someone visits your site four times before converting, sessions-based calculation captures all four of those visits in the denominator. Users-based calculation counts them once. For e-commerce, sessions often makes more sense because each visit is a genuine purchase opportunity. For B2B lead generation, where someone might research you across several sessions over several weeks, users can give you a cleaner picture of how many distinct prospects converted.

The critical rule is consistency. Pick one and stick to it. When I was running performance programmes across multiple clients simultaneously, the most common source of confusion in monthly reporting was teams switching between sessions and users without flagging it. A conversion rate that appeared to improve 0.8 percentage points month-on-month sometimes turned out to be a denominator change rather than a genuine improvement. That kind of error erodes trust in the data quickly, and once that trust is gone it is very hard to rebuild.

If you are working across conversion rate optimisation as a broader discipline, the CRO hub on The Marketing Juice covers the full programme structure, from measurement foundations through to testing and commercial impact.

Why Your Blended Site-Wide Rate Is Almost Useless

A single conversion rate for your entire website tells you almost nothing actionable. It is an average of wildly different traffic types, intent levels, and page experiences, compressed into one number that obscures all of the interesting variation underneath it.

Consider what is typically inside a blended site-wide rate. You have brand search traffic from people who already know you and are ready to act. You have top-of-funnel content visitors who arrived on a blog post and have no purchase intent whatsoever. You have returning customers logging in. You have people who landed on the wrong page and left in three seconds. Averaging all of that together and calling it your conversion rate is like averaging the temperature of your oven and your freezer and concluding your kitchen is fine.

The segments that actually matter are:

• Traffic source: Paid search, organic, direct, email, and social all convert at different rates because the intent behind each is different. Paid brand terms will almost always outperform generic paid traffic. Organic informational content will convert far below paid commercial terms. Treating them as one pool masks both problems and opportunities.
• Device type: Mobile and desktop conversion rates diverge significantly on most sites, particularly in e-commerce. If your mobile rate is materially lower, that is a product and UX problem worth isolating.
• Landing page: A dedicated campaign landing page optimised for a single action should convert at a different rate than your homepage. If it does not, that is a signal worth investigating.
• New vs. returning visitors: Returning visitors typically convert at a higher rate. A shift in that ratio can explain rate movements that have nothing to do with site changes.

When I took over a performance remit at an agency that had been reporting a flat 1.8% conversion rate for six months, the first thing I did was break it down by source. Paid brand was converting at 6.1%. Generic paid search was at 0.9%. The blended rate had been masking a serious paid search efficiency problem for half a year. The fix was not a CRO programme. It was a paid media restructure. You cannot see that from the blended number.

How to Exclude Traffic That Distorts Your Calculation

Not all visits should be in your conversion rate denominator. Including the wrong traffic inflates your visit count and artificially suppresses your rate, which can lead you to invest in CRO work when the real problem is traffic quality.

Traffic worth filtering out of your core conversion rate calculation includes bot and crawler traffic, internal visits from your own team, visits to pages that have no conversion path (your careers page, your privacy policy, your terms and conditions), and in some cases, visits shorter than a defined threshold that indicate immediate bounces with no real engagement.

Most analytics platforms let you apply filters or create segments that exclude these visit types. GA4 allows you to create custom segments that restrict your denominator to sessions that include a specific page or event, which is useful if you want to measure conversion rate only for visitors who reached a product page or pricing page rather than everyone who touched the site at all.

This matters particularly for content-heavy sites. If you have a large blog generating significant organic traffic from informational queries, those visitors are not in a conversion moment. Including them in your denominator when measuring product conversion rate makes the rate look worse than it is for the audience actually considering a purchase. Separate your content traffic from your commercial traffic and calculate rates for each independently.

Tools like Hotjar’s heatmap suite can help you understand which pages are generating engagement versus which are serving as entry points that lead nowhere, which informs where to draw the line on what to include in your conversion funnel measurement.

Micro-Conversions vs. Macro-Conversions: Calculating Both

Not every conversion is a sale or a lead form submission. Micro-conversions are the smaller actions that indicate progress through your funnel: adding a product to a basket, watching a video, downloading a resource, clicking a pricing page, spending more than two minutes on a key page. Macro-conversions are the end actions that directly drive revenue or pipeline.

You calculate both using the same formula. If 10,000 people visited your site and 800 added something to their basket, your add-to-cart conversion rate is 8%. If 200 of those 800 completed a purchase, your purchase conversion rate is 2% of total visits, or 25% of add-to-cart visitors.

That second calculation, the rate at which basket adds convert to purchases, is often more useful than the headline purchase rate because it isolates the checkout and payment experience from the upstream traffic and product discovery problem. A 25% basket-to-purchase rate on a consumer e-commerce site suggests a checkout friction issue. A 60% rate suggests the checkout is working and the problem is earlier in the funnel.

Understanding how your conversion funnel is structured is the prerequisite for calculating meaningful rates at each stage. Without mapping the funnel first, you end up calculating rates for stages that do not reflect how your customers actually move through your site.

For B2B sites with longer sales cycles, micro-conversions often carry more diagnostic value than macro-conversions in any given month, because the macro-conversion volume is too low to be statistically reliable. A B2B site with 5,000 monthly visits and 12 demo requests has a 0.24% macro-conversion rate. That is a useful number, but 12 conversions is not enough to run meaningful tests against. Micro-conversion rates, measured against larger volumes, give you a more workable signal.

The Conversion Rate Trap: When a Better Number Means Worse Performance

A rising conversion rate is not automatically good news. This is something I have had to explain to clients more times than I can count, and it is one of those points that sounds counterintuitive until you think it through.

If your visit volume drops by 30% because you paused top-of-funnel paid campaigns, and your conversion rate rises from 2.1% to 2.8%, that improvement is almost entirely explained by the removal of low-intent traffic from the denominator. You have not improved your site’s ability to convert. You have changed who is visiting it. The absolute number of conversions has likely fallen even as the rate looks better.

This is why conversion rate should always be read alongside absolute conversion volume and alongside revenue or pipeline value. The metric that matters commercially is not the rate in isolation. It is what the rate multiplied by volume produces in terms of actual business outcome.

I saw this play out in an agency context when a client’s e-commerce team celebrated a conversion rate improvement from 1.9% to 2.4% after a site redesign. When I looked at the full picture, total orders had fallen by around 8% because the redesign had inadvertently slowed page load times on mobile, suppressing visit volume from that segment. The rate looked better. The business was worse. The board only saw the rate.

That experience is part of why I am sceptical of conversion rate as a standalone KPI. It needs context, and that context is always commercial.

How to Build a Monthly Conversion Rate Reporting Framework

A practical monthly conversion rate report should include more than the headline number. Here is the structure I have used across both agency and in-house environments:

• Headline rate with denominator stated: “2.3% conversion rate based on 14,200 sessions, excluding bot traffic and internal visits.” The denominator should always be visible.
• Rate by traffic source: Paid brand, paid non-brand, organic branded, organic non-branded, direct, email, referral. Each should be a separate row.
• Rate by device: Desktop, mobile, tablet as a minimum.
• Rate by key landing page or campaign: If you are running paid campaigns to specific landing pages, each page should have its own rate.
• Month-on-month and year-on-year comparison: Both matter. Month-on-month catches recent changes. Year-on-year controls for seasonality.
• Absolute conversion volume alongside rate: So that a rate improvement accompanied by a volume decline is immediately visible.
• Revenue or pipeline value per conversion: So that a shift in conversion quality (lower average order value, for example) is captured even if the rate holds steady.

This is not a complex framework. It is a disciplined one. The discipline is what most teams lack, not the technical capability to produce it.

Behaviour analytics tools like Hotjar’s heatmap and session recording tools sit alongside this quantitative framework well. They give you the qualitative layer that explains why rates are moving in the direction the numbers show.

What a “Good” Conversion Rate Actually Looks Like

There is no universal benchmark for a good conversion rate, and anyone who gives you one without context is guessing. Conversion rates vary enormously by industry, by traffic source, by device, by price point, by whether you are measuring leads or sales, and by how you have defined and filtered your denominator.

Across the industries I have worked in, the range is genuinely vast. A high-consideration B2B software product might convert at 0.5% to 1.5% on demo requests from generic paid traffic. A well-run e-commerce site selling low-cost consumables might see 4% to 6% from brand search. A financial services lead generation page with strong brand equity and a simple form might reach 8% or higher from direct traffic.

The only benchmark that matters is your own historical performance, segmented properly. Your conversion rate last month versus the same month last year, for the same traffic source on the same device type, is a meaningful comparison. An industry average from a report you found online, based on an undefined methodology and an unknown sample, is not.

What I have found useful when evaluating conversion performance is comparing the rate against your own funnel stages rather than external benchmarks. If your product page to add-to-cart rate is 15% but your add-to-cart to purchase rate is 18%, the problem is clearly in the checkout, not in the product presentation. That kind of internal benchmark is actionable. An external “industry average of 2.35%” is not.

For a broader view of what effective conversion optimisation looks like in practice, these CRO case studies from Crazy Egg show how rate improvements translate into commercial outcomes across different site types and contexts.

Common Calculation Errors and How to Avoid Them

Beyond the denominator question, there are several calculation errors that produce misleading conversion rates in practice.

Double-counting conversions. If a user completes two purchases in a single month, do you count that as two conversions against one user? For revenue calculations, yes. For conversion rate calculations, this depends on whether you are using sessions or users as your denominator. If you use sessions, two purchases in two separate sessions is two conversions in two sessions. If you use users, two purchases from one user is two conversions against one user, which can produce rates above 100% for certain segments, which is a signal that something is misconfigured.

Attribution window mismatches. If someone visits your site on the 28th of March and converts on the 2nd of April, which month does that conversion belong to? Most platforms attribute to the session date, not the conversion date. For monthly calculations, this creates a lag that can make end-of-month and start-of-month numbers look unusual. It is worth understanding how your platform handles this before drawing conclusions from month-boundary data.

Sampling in analytics platforms. High-traffic sites can trigger data sampling in some analytics configurations, which means the numbers you see are estimates rather than exact counts. If your platform is sampling, your conversion rate is an approximation. That is not necessarily a problem, but it is worth knowing, particularly if you are making significant budget decisions on the back of small rate differences.

Goal configuration errors. If your conversion goal is misconfigured and firing on page load rather than on form submission, you will record a conversion for every visitor to a confirmation page, including people who arrived there via back-button navigation or direct URL entry. I have seen this inflate conversion rates by a factor of three on sites where no one had audited the goal setup in over a year. Always verify that your goals are firing correctly before trusting the rate they produce.

The principles behind accurate conversion measurement connect directly to the broader discipline of CRO. This piece from Search Engine Land on core CRO principles is worth reading for the foundational thinking, even if some of the tactical specifics have evolved since it was written.

Turning the Calculation Into a Decision

The point of calculating your conversion rate is not to have a number to report. It is to identify where commercial performance can be improved and to prioritise where to focus effort. A well-constructed conversion rate calculation, segmented properly and read in context, tells you where the gap between traffic and revenue is largest, and therefore where the highest-value work is.

Early in my career, I taught myself to code because I could not get budget to build a website through official channels. The lesson I took from that experience was not about coding. It was about working with what you have and being precise about what you are trying to achieve. Conversion rate calculation is the same discipline applied to analytics: you work with the data you have, you are precise about what it is measuring, and you use it to make a decision rather than to fill a slide.

When I walked into a CEO role and told the board the business would lose around £1 million that year, the credibility came from the precision of the number and the rigour behind it. The same applies to conversion data. A rate of 2.3% calculated on clean, segmented, properly filtered data is worth more than a rate of 4.1% that no one has interrogated. The precision is what makes it actionable.

If you want to understand how conversion rate measurement fits into a full optimisation programme, the CRO section of The Marketing Juice covers the complete picture from audit through to testing and commercial measurement.

There is also a useful distinction to draw between the calculation itself and the optimisation work that follows. Unbounce’s breakdown of the right and wrong approaches to CRO is a good reference for understanding what rigorous optimisation looks like once you have your measurement in order.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions