Email conversion rate is the percentage of recipients who completed the action a campaign was asking for — a purchase, a signup, a booking, a reply. The formula is simple. The arguments are all about the denominator, and the denominator is what decides whether the number is useful or decorative.
The three denominators, and which to use
Conversions ÷ delivered. This is the honest one. It tells you what your email produced relative to the effort of sending it, and it is the only version that stays comparable when your list or your deliverability changes. Use this as your headline number.
Conversions ÷ opens. This tells you how persuasive the email was once someone read it. It is a useful diagnostic and a terrible headline, because Apple Mail Privacy Protection inflates opens for a large and unstable share of your list. Any metric with opens in the denominator now measures Apple's proxy fetching as much as it measures your audience.
Conversions ÷ clicks. This measures the landing page, not the email. Worth watching, because a high click rate with a low click-to-conversion rate is a specific and fixable diagnosis: the email sold something the page did not deliver.
Report the first, diagnose with the third, and be careful with the second.
Calculating it without lying to yourself
The mechanical part is easy: tag your links, attribute conversions to the tagged source, divide by delivered. The part that goes wrong is the attribution window and the attribution model.
Window. A 30-day window will attribute to email a great many purchases that would have happened anyway. A same-session window will miss the genuine considered purchases that email is often best at starting. Seven days is a defensible default for most businesses; whatever you pick, keep it fixed, because changing the window changes the number more than any campaign improvement ever will.
Model. Last-click gives email credit only when it was the final touch, which systematically undercounts email's role in long consideration cycles and overcounts it for the abandoned-cart flow that fired ninety seconds after someone left. There is no correct answer here, only a consistent one. We work through the options in email attribution.
Deduplication. If someone clicks two links in the same email and converts once, that is one conversion. If someone receives both a newsletter and an automated flow in the same week and converts, both will claim it unless you decide which wins. Write down the rule and apply it everywhere.
What a good rate looks like
Published benchmarks put ecommerce email conversion somewhere around 1–3% of delivered for broadcast campaigns, with triggered and transactional flows several times higher — abandoned cart in particular routinely runs an order of magnitude above broadcast. B2B lead generation typically sits lower on conversion and much higher on value per conversion.
Treat all of that as orientation rather than a target. Benchmarks are averaged across wildly different products, prices and list qualities, and your own trailing six months is a far more useful comparison. The broader numbers are collected in email marketing benchmarks.
Why your conversion rate is lower than you want
In descending order of how often it is the actual cause:
- The email reached the wrong people. A campaign sent to your whole list converts worse than the same campaign sent to the segment it was written for. This is a segmentation problem wearing a conversion costume.
- It did not reach the inbox at all. Delivered is not the same as inboxed, and your conversion rate quietly includes everyone who received it in the spam folder. Check inbox placement before you rewrite anything.
- One email asked for several things. Every additional call to action divides attention. The highest-converting campaigns most people ever send are the ones with a single link repeated three times.
- The landing page broke the promise. Click-to-conversion below about 10% usually means the page is not the thing the email described.
- The offer is not compelling. Last on the list because it is the first thing everyone blames and rarely the actual problem.
Improving it, in order of leverage
Segment first — matching message to audience moves conversion more than any copy change. Then reduce the number of asks per email to one. Then fix the click-to-conversion gap by making the landing page continue the email's sentence rather than starting a new one. Then, and only then, test subject lines and copy, which mostly move opens and clicks rather than conversion.
Triggered flows deserve a separate mention because they are the highest-converting email most businesses can build and many still have not: abandoned cart, post-purchase, and welcome. If your conversion rate is disappointing and you do not have those three running, build them before you optimise anything.
You can put rough numbers on the upside with our ROI calculator before committing the work.
Frequently asked questions
What is a good email conversion rate?
For ecommerce broadcast campaigns, roughly 1–3% of delivered is typical, with triggered flows substantially higher. For B2B, conversion is usually lower and worth more per event. The useful benchmark is your own trailing average, because the published ranges average across businesses nothing like yours.
Should I calculate conversion rate on opens or on delivered?
Delivered, for anything you report. Open-based conversion rates are distorted by privacy-protection proxy opens, which vary by audience and change without warning, so the denominator moves for reasons unrelated to your performance.
Why is my click rate high but my conversion rate low?
Almost always a mismatch between the email and the landing page. The email created a specific expectation and the page did not immediately confirm it. Check that the page headline echoes the email's promise and that the action is above the fold.
How do I attribute a conversion that happened days after the click?
Pick an attribution window, apply it consistently, and record which model you used. Seven days is a reasonable default. The exact choice matters far less than never changing it mid-comparison, because window changes swamp real performance differences.
