I Built an Affirmations App Because Affirmations Made Me Feel Worse

Three years ago I stood in my bathroom, looked at my own reflection, and said "I am successful" out loud, alone, like the internet told me to.

I was not successful. I'd been laid off four months earlier, I was burning savings, and the project I'd left my job to build was quietly failing. And in the half-second after I said the words, something in my chest did the exact opposite of what the YouTube video promised. It didn't swell. It flinched. A dry little internal voice said no you're not, and then — this is the part nobody warns you about — it started listing reasons. The bank balance. The unanswered emails. The demo that had gone badly enough that I'd stopped sending it.

I did this every morning for about three weeks, because the advice everywhere was consistency. By the end of it I could recite my own failures faster and more fluently than before I started. The affirmations hadn't just failed to help. They'd given my inner critic a daily rehearsal slot.

I assumed I was doing it wrong. I was half right.

Why did affirmations make me feel worse instead of better?

The thing that changed everything for me wasn't a mindset shift. It was a footnote.

I went looking for evidence that affirmations worked — partly out of stubbornness, partly because I wanted permission to quit — and found a 2009 study out of the University of Waterloo, published in Psychological Science by Joanne Wood and colleagues. Participants repeated the phrase "I'm a lovable person." For people who already felt good about themselves, it helped a little. For people with low self-esteem, repeating the phrase left them feeling worse than doing nothing at all.

Read that again, because it's the whole story: the people affirmations are marketed to hardest are the people the classic version can actively backfire on.

The mechanism, as best I understand it, is that your mind doesn't passively receive statements — it audits them. Tell yourself something close to what you already believe and it gets absorbed. Tell yourself something far outside it and your mind treats it like a false claim from a stranger: it assembles counterevidence. "I am successful," said by a guy four months into a failed launch, isn't an affirmation. It's an invitation to cross-examine yourself. And I accepted, daily, for three weeks.

That reframe mattered more to me than any technique: the flinch wasn't a character flaw. It was an accurate measurement of the gap between the claim and my life.

If you want the fuller, less autobiographical version of this mechanism, I wrote it up separately in why affirmations feel fake. But the short version is: my affirmations weren't too negative or too positive. They were too far.

So do affirmations actually work, or not?

Here's where I landed after a couple of months of reading, and it's more interesting than either camp's answer.

The pop-culture version — repeat a grand claim until reality complies — isn't what the supportive research describes. But the research that does hold up (a lot of it under the name self-affirmation theory) describes something quieter: people affirming things that are already true. Values they actually hold. Hard things they've actually survived. And in those studies, that practice measurably changed how people handled stress and criticism. Not magic. Steering attention back onto real ground.

So both camps are wrong in an instructive way. The affirmation-poster crowd is wrong that belief is optional — the Waterloo study says unbelievable claims can backfire. The cynics are wrong that the whole practice is empty — deliberately rehearsing what's true about you, instead of letting your critic run the highlight reel, is a real and useful skill.

The version that works lives in a middle that almost nobody sells, because it doesn't demo well:

An affirmation that works is less impressive than the one that doesn't.

That sentence became the thesis of everything I built afterward.

What I actually changed

I didn't quit affirmations. I shrank them until they stopped triggering the audit. Concretely:

• I am successful became I'm still building, and I haven't quit. True. Verifiably, boringly true. My brain had nothing to object to, which felt anticlimactic — and then, weirdly, it felt like relief. Being leveled with instead of sold to.
• I am confident became I'm learning to send the email before I've perfected it. "I'm learning to" turned out to be nearly refutation-proof — you demonstrably are, just by trying.
• Everything works out for me became I've gotten through every bad month so far. Not a prediction. A summary my critic kept leaving out of the record.
• On genuinely bad days: I can handle the next five minutes. Never once did that phrase flinch back at me. It claims almost nothing, which is exactly why I could say it at my worst.

I started calling these bridge statements — phrases that stand on what's true now and lean toward where you're going, instead of teleporting to the destination and daring your brain to object. The test I used was crude and effective: say the phrase, watch the reaction. Settle or shrug, keep it. Eye-roll or chest-tighten, shrink it. (I later formalized this into the believability test, which is still the first thing I'd hand anyone starting out.)

Two honest caveats from that period, because they matter.

First: it was slow. Weeks of small phrases before I noticed my default self-talk shifting — before the voice narrating my day started sounding more like the phrases and less like the prosecutor. No breakthrough moment. More like a room getting gradually less cold.

Second: some of what I was carrying that year wasn't affirmation-sized. Part of what made "I am successful" so corrosive was that it papered over a low stretch that deserved actual attention. Better phrases helped my daily self-talk; they were not a substitute for the therapist I eventually talked to, and I'd never suggest otherwise. If your worst days are heavier than a phrase can hold — a mood that hasn't lifted in weeks, a worthlessness that doesn't budge — the honest move is professional support first, phrases alongside. That's not failing at self-talk. That's the same principle, applied at scale: match the tool to what's true.

Why that turned into an app

Once the small-phrase approach was working, I did the obvious thing: I went looking for an app to keep me consistent. And every one I tried had the same flaw — the flaw, the one the Waterloo study documented in 2009 and the industry has spent the better part of two decades ignoring.

They all pushed the same phrases at everyone. "I am abundant." "I am limitless." "I attract everything I desire." Notifications shipping grand claims to strangers, calibrated to nobody's actual evidence, with no mechanism anywhere for the one question that determines whether the phrase helps or backfires: do you believe this?

It's not hard to see why. "I am limitless" screenshots beautifully. "I'm learning to send the email before I've perfected it" does not. The incentives of the affirmation industry point directly away from what the research says works — and the cost lands on the person standing at the mirror concluding that they're broken, when what's actually broken is the calibration. There's a name for the failure mode, too: it's the same mechanism behind toxic positivity, just pointed inward.

So I started building the thing I couldn't find. Affirm Away is built from the opposite premise: it will ask what's actually true for you before it ever suggests a phrase, keep every affirmation inside the range you can say without flinching, and treat the flinch as data — a signal to shrink the claim, never to push through. Bridge statements, not poster language. If you'd rather write your own from scratch, everything I learned about the drafting side is in how to write affirmations that actually feel true — the app will just automate the calibration and the consistency.

The version of this I wish someone had told me

If you've tried affirmations and felt worse — that flinch, that eye-roll, that quiet yeah, right — you're not too broken for this and you're not too cynical. You're likely just holding a phrase that's too far from your life, and your brain did its job by rejecting it.

Claim less. Shrink the statement until some part of you nods instead of arguing. "I'm still here" will outwork "I am unstoppable" every week of the year, for the unglamorous reason that you can actually say it on a Tuesday and mean it.

That nod — the small, honest one — is the entire practice. It's what I couldn't find three years ago, standing in my bathroom losing arguments with my own reflection. It's why the app exists. And if affirmations ever made you feel worse, it's the specific thing I'm building to fix.