A/B testing your email signature: simple frameworks that work
Every week, thousands of emails get sent with a broken image, a missing link, or a CTA that nobody clicks. The culprit is usually the signature — the one piece of email infrastructure most teams set once and forget. But your signature sits below every outbound email, every support reply, and every cold outreach. If it's not optimised, you're leaving clicks and conversions on the table.
Email signature A/B testing is the systematic way to fix that. By running controlled experiments on your sign-off block, you can measure which layout, link placement, or CTA actually drives engagement. Here are four frameworks that work with real email clients and deliver actionable data.
Why your signature deserves its own test plan
Most marketers treat the signature as a static afterthought — a logo, a phone number, a LinkedIn icon, done. But your signature has a unique property: it appears on every email from your domain, regardless of the campaign or recipient. That means even a small improvement in click-through rate compounds across thousands of sends.
The problem is that email clients render signatures inconsistently. Gmail strips images by default. Outlook uses Word's rendering engine, which mangles CSS. Apple Mail respects line-height but ignores max-width on images. A/B testing your signature across these environments isn't vanity — it's the only way to know what actually survives delivery and gets clicked.
A proper test isolates one variable at a time: the number of links, the position of the CTA, the presence of a banner image, or the font stack. Without that discipline, you'll attribute a lift to the wrong change and waste the next quarter optimising a ghost.
Framework 1: The link density test
The most common mistake in email signatures is link overload. Salespeople want their calendar link, LinkedIn, Twitter, phone number, and a "book a demo" button — all in one block. The result is a visual mess that gets ignored or, worse, flagged as spam.
The hypothesis: Fewer links drive higher click-through on the primary CTA.
Set up two signature variants:
- Variant A (control): Three links — website, LinkedIn, and a primary CTA (e.g., "Book a meeting").
- Variant B (treatment): One link — the primary CTA only. Remove all secondary social icons and text links.
Run the test for two weeks or until each variant has been seen in at least 500 sent emails. Track click-through on the primary CTA link using a redirect or a tracking service that can distinguish between the two variants.
What you'll likely find: Variant B often wins by 20-40% on primary CTA clicks. The reason is decision fatigue — when you give recipients one clear action, they take it. When you give them five, they take none. This effect is amplified on mobile, where tiny tap targets make multi-link signatures unusable.
If you need social links for credibility, move them below the CTA in a smaller, monochrome format. The primary action stays above the fold.
When to break this rule
If your signature is primarily for internal or transactional emails (support tickets, order confirmations), link density matters less because the recipient already has a specific task. In those cases, include the links that reduce support load — knowledge base, account page, status page — even if it means six links. Test for reduced reply rate rather than click-through.
Framework 2: The banner vs. text-only test
Banner images in signatures are polarising. A well-designed banner can boost event registration or product awareness. A broken banner — blocked by Gmail, clipped by Outlook, or simply too wide — hurts your brand every time it fails to load.
The hypothesis: A text-only signature with a single link outperforms a banner-based signature on overall engagement.
Build two variants:
- Variant A (control): Text-only signature — name, title, company, phone, one link. No images.
- Variant B (treatment): Same text block, but with a 600x100 pixel banner image below it linking to a landing page.
Track two metrics: click-through on any link in the signature, and the image load rate (how many recipients actually saw the banner). Most email clients block images by default, so the banner will only render for recipients who enable image loading. That's typically 30-50% of opens, depending on your audience.
What you'll often see: The text-only variant drives more total clicks because every recipient sees the link. The banner variant might have a higher click rate among those who see it, but the absolute number of clicks is lower due to image blocking. The exception is internal communications where you control the email client and can enforce image loading.
If you must use a banner, keep it under 600px wide (safe for all clients), use a simple background colour instead of a photo (reduces file size and improves rendering), and always include a text fallback link directly below the image. Never rely on the image alone for your CTA.
Framework 3: The CTA position test
Where you place your primary link matters more than the link text. Recipients scan signatures in a predictable pattern: name, title, then the first clickable element below that. If your CTA is buried after three social icons, it's invisible.
The hypothesis: Moving the primary CTA to the first position in the link stack increases click-through.
Test these two layouts:
- Variant A (control): Name, title, phone, LinkedIn, Twitter, CTA link.
- Variant B (treatment): Name, title, CTA link, phone, LinkedIn, Twitter.
Everything else stays identical — same link text, same font, same spacing. Run the test for two weeks.
The results are usually stark. Variant B can double the click-through rate on the primary CTA. The reason is serial position effect — items at the beginning of a list get more attention and better recall. Your CTA competes with every other element in the signature, so put it first.
This framework works especially well for sales teams where the signature is part of a cold email sequence. The prospect has already read your message; the signature is the last thing they see. If the CTA is prominent and immediate, they're more likely to act before switching tabs.
One caveat: if your signature includes a mandatory legal disclaimer (common in regulated industries), place it below the signature block, separated by a thin horizontal rule. The disclaimer is not part of the test and should not interfere with the CTA position.
Framework 4: The mobile-first test
Over 60% of email opens happen on mobile devices. Yet most signatures are designed on a 24-inch monitor with full CSS support. On a 375px-wide iPhone screen, a signature with three columns of icons collapses into a jumbled stack that requires horizontal scrolling or accidental taps.
The hypothesis: A single-column, left-aligned signature outperforms a multi-column, centred signature on mobile click-through.
Create two variants:
- Variant A (control): Multi-column layout — logo on the left, contact info in the middle, social icons on the right. Centred alignment.
- Variant B (treatment): Single-column layout — logo on top, contact info below, social icons at the bottom. All left-aligned. No floats, no tables with fixed widths.
Test on mobile click-through specifically. If your tracking tool can segment by device, isolate the mobile data. If not, assume the majority of your audience is mobile and optimise for that.
What you'll find: Variant B almost always wins on mobile. The single-column layout adapts to any screen width without breaking. Left-alignment matches the reading flow of most languages. And the lack of nested tables means Outlook on Windows doesn't mangle the spacing.
Desktop users won't notice the difference — the signature still looks clean and professional. But mobile users will have a significantly better experience, which translates to higher click-through and lower frustration.
How newmail.click handles this
Newmail.click lets you build and deploy multiple signature variants without touching HTML or CSS. The free editor handles responsive layouts, font stacks that survive Outlook, and image hosting with automatic resizing. If you upgrade to Pro, you get click and open tracking per variant, so you can run the frameworks above without setting up redirects or writing tracking scripts. Pro also sends Slack or email alerts when a variant hits a statistically significant lift, and it appends UTM parameters to your links by default — no manual tagging. You don't need a developer to run these tests; you just need a hypothesis and two weeks of email volume.
Putting it all together
Pick one framework and run it for two weeks. Don't test multiple variables at once — you won't know which change caused the lift. Track clicks, not opens (opens are inflated by preview panes and don't correlate with action). And always test against a control that represents your current signature, not a stripped-down version you'd never actually use.
Once you have a winner, deploy it to your whole team and start the next test. Over three months, you can optimise link density, CTA position, mobile layout, and image strategy. The compound effect of those incremental wins will show up in your meeting booking rate, your demo request volume, and your support ticket deflection.
Your signature sends itself a thousand times a day. Make every one of those sends count.