DotComSecrets Labs 108 Proven Split Test Winners is a swipe file of A/B tests that lifted conversions. Here's what the report covers and how to use it.
DotComSecrets Labs 108 Proven Split Test Winners is a swipe file of A/B tests that the DotComSecrets Labs team ran on opt-in pages, sales pages, order forms, and email subject lines, along with the variation that produced the higher conversion rate in each test. It is a research shortcut for marketers who want documented starting points instead of guesses, but it is not a guarantee that the same change will win in your funnel. The 108 entries work best as a bank of hypotheses that you re-test on your own traffic.
What the 108 Proven Split Test Winners Actually Are
DotComSecrets Labs grew out of the testing culture around Russell Brunson's DotComSecrets and ClickFunnels ecosystem, where funnel owners are pushed to test one element at a time. The report collects the tests that reached a clear, measurable result rather than the ones that ended in a tie.
Most entries show three things:
- The element tested — a headline, an opt-in form, a button, a video, a price presentation, or a subject line.
- The control and the challenger — the original version and the variation that beat it.
- The outcome — which version won and, in many entries, roughly by how much.
The material is typically delivered as a downloadable PDF or a members-area swipe file with screenshots, so you can see the actual pages rather than just a summary. That visual format is a large part of why the file gets passed around in marketing communities.
The Types of Tests Usually Covered
Most of the winners fall into a handful of funnel stages. The table below maps those categories to the patterns that tend to appear in the report.
| Funnel element | What gets tested | Pattern that often wins | How to apply it |
|---|---|---|---|
| Headline | Benefit vs. curiosity vs. specificity | Concrete promises with a number or timeframe | Rewrite your headline as one challenger and test |
| Opt-in form | Field count, placement, button copy | Fewer fields and action-based button text | Cut one field at a time, not five |
| Hero media | Video vs. static image vs. text only | Short video on cold traffic, plain text on warm traffic | Test the media and hold the copy constant |
| Order form | Layout, trust badges, guarantees, payment options | Visible guarantee near the payment button | Add the guarantee block as a single variable |
| Email subject line | Length, curiosity, personalization | Short, plain, curiosity-driven subjects | Test on your most engaged segment first |
The winners are usually small, single-variable changes rather than full page redesigns, and that is exactly what makes them reusable. Single-variable tests are the only kind where you can cleanly attribute a lift to one thing.
Why a Winner in One Funnel Isn't a Winner Everywhere
A split test result is only valid for the audience, traffic source, and offer it was tested on. DotComSecrets Labs ran its tests on its own traffic, which skews toward online marketers buying information products.
Several factors can flip a result when you move it to your own pages:
- Audience — a cold social audience and a warm email list respond to different promises.
- Offer strength — a weak offer caps conversion no matter how good the page is.
- Traffic intent — search visitors and interrupt-traffic visitors arrive with different levels of awareness.
- Sample size — a 40% lift measured on 30 conversions is mostly noise.
A split test winner is a documented result for one audience and one offer, not a universal conversion rule.
How to Use the 108 as a Hypothesis Bank
The practical value of the file is that it shortens the list of things worth testing. A workable process looks like this:
- Pick three to five entries that match the funnel stage you are working on right now.
- Duplicate your current page so the original stays live as the control.
- Change exactly one element per challenger.
- Set a sample target before launch — a common rule of thumb is about 100 conversions per variant.
- Run the test for at least one full week and through a complete sales cycle.
- Keep the winner only if the gap holds at roughly 95% confidence; otherwise call it a tie.
- Log every result so your own swipe file eventually beats anyone else's.
Testing tools have changed since the report was written. Google Optimize shut down on September 30, 2023, so most funnel builders now rely on native testing features inside ClickFunnels or on third-party platforms such as VWO, Optimizely, and Convert.com.
Mistakes That Waste a Swipe File
Marketers who get little from the report tend to make the same handful of errors.
- Copying the whole page. You then have no idea which change produced the lift.
- Calling a winner after two days. Early leads often regress toward the average.
- Ignoring mobile. A desktop winner can lose on phones, and most US traffic is mobile.
- Testing during a promotion. A sale period distorts normal buyer behavior.
- Skipping the offer. No headline fixes an offer nobody wants.
DotComSecrets Labs vs. Other "Labs" Names Online
The word "labs" appears in many unrelated niches, which is why this page draws some confused visitors. If you landed here from Proven labs peptides reddit threads or while trying to work out whether is true research labs legit, you are looking for a research-chemical supplier — vendors such as Heritage Labs operate in that space, and it has nothing to do with funnel testing. This page covers only the marketing split test compilation, and peptide products sold for human use sit outside FDA approval in the United States.
Frequently Asked Questions
What is DotComSecrets Labs 108 Proven Split Test Winners?
It is a compiled swipe file of 108 A/B tests that the DotComSecrets Labs team ran on opt-in pages, sales pages, order forms, and email subject lines, showing which variation produced the higher conversion rate. Each entry generally includes the element tested, the control version, the challenger version, and the outcome. It is a starting point for your own tests, not a guaranteed result.
Is the 108 Proven Split Test Winners report still relevant today?
Yes, the underlying testing principles still hold, because headline specificity, form length, and button copy continue to move conversion rates. What has changed is the tooling — Google Optimize shut down on September 30, 2023, so most marketers now test inside their funnel builder or with platforms like VWO and Optimizely. Treat the report as a source of hypotheses and validate everything on current traffic.
How many conversions do I need before trusting a split test result?
A common rule of thumb is roughly 100 conversions per variant before a result is worth acting on, though the number depends on your baseline conversion rate and the size of the lift you expect. Small lifts need far more data than large ones. If you cannot reach that volume, run a bigger change instead of a minor tweak.
This page provides educational research information and does not replace medical advice, diagnosis, or treatment.