Evidence-based relationship advice means a specific claim traces back to controlled research that measured a real outcome and reported it honestly, limits and all. It is not proven by testimonials, confidence, a big following, or one coach story about one couple who got the ring. To judge any piece of advice, put its proof on the Source Ladder, then ask whether the person selling it will show you the evidence and name their conflict. Everything sold on before-and-after screenshots alone is marketing wearing the word science.

Here is the uncomfortable part, and I am the wrong person to be telling you, which is exactly why you should hear it. The entire industry you are trying to evaluate, the coaches and the books and the courses and the accounts posting a screenshot of one text that "changed everything," almost never shows you its evidence. It shows you a result. A woman who got the ring. A guy who came back. A before and after with the boring middle deleted. And it lands on you because a result feels like proof.

It is not.

A result is one data point with the failures cropped out, handed to you by the same person asking for your money. You do not need a psychology degree to see through that. You need a way to rank where a claim's evidence actually sits, and a short test for whether the person making it is teaching you or selling you.

That is this whole page.

Start with what evidence-based actually means

Evidence-based is not a vibe, and it is not a badge you earn for sounding smart. It means something exact. Somebody made a claim. They tested that claim against a comparison group. They measured an outcome you actually care about. Then they reported what happened, including the parts that did not work.

Claim. Test. Measure. Report the misses too.

Most advice dies at step one. It never makes a claim specific enough to test. "Be high value" cannot be tested. "Raise your standards" cannot be tested. But "cutting your reply length in the first three weeks raises how often he initiates" can be. The moment advice gets specific enough to measure, it also gets specific enough to be proven wrong, and that is precisely why so much of the industry stays vague. Vague can never be caught.

The phrase evidence-based gets stapled onto marketing all day long. Anyone can type it. So stop treating the label as the evidence and start looking at what is sitting underneath it.

The Source Ladder

The bottom rung is the anecdote. "This worked for me." "My client sent this exact text and he proposed a month later." One story. No comparison group. No idea what happened to the people it failed. The Federal Trade Commission, which polices advertising claims in the United States, is blunt about this in its guidance on substantiating claims: anecdotal evidence based solely on the experiences of individual consumers is generally insufficient, and objective claims about what a product or program does need competent and reliable scientific evidence behind them before anyone sells them to you. A testimonial is not automatically a lie. It is just the weakest possible evidence, dressed up to look like the strongest.

The next rung up is informed opinion. A coach who has watched a thousand relationships develops real pattern recognition, and that is worth something. It is worth more when the coach says "this is what I have seen," and worth nothing the second the coach says "studies show" and then never names one. Opinion is honest right up until it borrows the authority of research it does not actually have.

Above opinion sits professional and clinical guidance. Licensed therapists, established professional bodies, official consumer protection. Not infallible, but accountable to a standard, a license, and a complaints process.

The top rung is controlled research. Not one study you can cherry-pick, but the weight of many, measured and pooled. And that rung has a surprise on it that most marketers pray you never notice.

What top-rung evidence actually looks like

Real evidence is humble. That is the part nobody selling you a course wants you to internalize.

When researchers pooled the studies on marriage and relationship education, one large meta-analytic study worked through 86 reports and more than 500 effect sizes and found that the programs did help. The effects on relationship quality landed around d = .30 to .36, and on communication skills around .43 to .45. Small to moderate. Measured. Replicated across many samples. And then the same study did the thing marketing never does. It named what it could not conclude, flagging that the samples lacked racial, ethnic, and economic diversity, so it refused to claim the results held for everyone.

Read that again. The strongest evidence in the room reported modest effects and then listed its own blind spots.

That is the tell you can carry everywhere. Honest evidence is specific, modest, and openly limited. Marketing is universal, dramatic, and never wrong. So when a coach promises a system that works on any man, every time, no exceptions, the certainty itself is the red flag. The people who actually have the data are more careful than the people who only have a funnel.

Run the disclosure test

Once you know where a claim sits on the ladder, run one more check on the human being making it. Three questions, in order.

First, is there a specific claim, or just a feeling? "He will chase you" is a feeling. "Do this and he commits within ninety days" is a claim. Vague promises cannot be checked, which is the entire reason they stay vague.

Second, what is the evidence, and will they show it before you buy? The same standard the Federal Trade Commission uses on advertisers is the one you should use as a buyer. A claim is supposed to have real substantiation behind it before it is ever made, not after your card clears. So if someone will not show you the basis for their promise until you have paid, treat the promise as unproven. Because it is.

Third, do they disclose the conflict? Everyone selling advice has one. I have one. The honest ones say it out loud. The ones hiding it behind "I just want to help you find love" are managing you, and the softness of the pitch is doing the hiding.

The tells that you are being sold, not taught

A few signals show up almost every time. A guarantee that a specific person will feel a specific way, which cannot exist, because you do not control another human being, and anyone promising it is either lying or does not understand what they are selling. Secrets and hidden triggers, as if adult attraction ran on a cheat code you can buy. Urgency and scarcity bolted onto the checkout so you purchase before you think. Testimonials carrying the entire weight of the proof. And the loudest one of all, no stated limitations, ever.

Teaching tells you where it fails. Selling pretends it never does.

I run an operation where my team has thousands of conversations with men every single week, and even with that much live data feeding me, I can point to exactly where my read breaks down and which men it does not describe. Anyone claiming their advice works on everyone has less data than that and more confidence than the actual researchers. That combination, maximum certainty with minimum evidence, is the whole scam in one line.

A message that tests any advice source

You do not have to guess. Make them answer. Before you buy a course, join a program, or take a coach seriously, send one message and then read the response instead of the pitch.

SEND THIS BEFORE YOU BUY

Before I sign up, what is the main claim you are making, and what is it based on? I am not asking for your best result. I want your typical one. What happens for the people it does not work for, and what is your refund policy if I turn out to be one of them?

A source with real evidence answers this calmly. It names the claim, points to something outside its own testimonials, and gives you an honest hit rate. A source running on marketing gets defensive, pivots to your mindset, or tells you that asking questions like this is the reason you are still single.

The dodge is the answer.

Where every kind of advice actually sits

Be fair about the whole shelf, this page included. Most dating books, most coaches, most courses, and most accounts, mine among them, live in the middle of the ladder. Experience-based synthesis. Pattern recognition built from a lot of exposure. That is genuinely useful, and it is not the same thing as controlled research, and the honest move is to say so rather than borrow a lab coat for the sales page.

So place each option where it belongs. A licensed therapist works from an accountable, regulated standard. A researcher's meta-analysis sits at the top and will also disappoint you with how careful and modest it is. A book or a coach sits in the middle, worth exactly as much as the author's honesty about that. If you want to compare those three routes head to head, dating coach versus book versus therapy lays them side by side. To pressure-test a specific title, use how to evaluate dating advice books. To screen a coach before you pay, dating coach red flags and how to choose a dating coach do the vetting for you.

None of this asks you to become a scientist. It asks you to stop letting a good result stand in for real evidence, and to make the person selling you certainty show their work.

The advice that survives those questions is the advice worth keeping. The rest was always just a screenshot with the failures cropped out.