There's a specific kind of stuck that happens the night before an ad goes live. You've read it eight times. You've swapped one word for a slightly better word, then swapped it back. You're not improving the ad anymore. You're just handling it.
Most owners assume the fix is more scrutiny: one more look, one more friend's opinion, one more version. But the research on how people actually make good-enough decisions, and separate research on what actually makes an ad work, both point somewhere else. The problem isn't that you haven't found the best ad. It's that you're still looking for it when you should be asking whether this one clears the bar.
The economist who studied grocery shoppers, not ad agencies.
In 1956, the economist Herbert Simon coined a word for this: satisficing, a blend of "satisfy" and "suffice." He was describing how real people make decisions when they don't have infinite time to compare every option, which is most people, most of the time. Instead of searching for the mathematically best choice, you set a threshold for what's acceptable, and you stop the moment something clears it.
Simon didn't frame this as the lazy cousin of optimizing. In his Nobel Prize speech, he said decision makers can satisfice "either by finding optimum solutions for a simplified world, or by finding satisfactory solutions for a more realistic world." Searching for the perfect ad means pretending you're in a simplified world where you can see every possible version and rank them. You're not. You're in the real one, where the cost of one more revision pass competes with everything else on your plate this week.
Here's how this plays out in real decisions outside academia. The clearest evidence comes from an unlikely place: used car lots. A study of 628 dealers found 97 percent followed a satisficing pattern almost exactly matching Simon's model. They'd price a car in the middle of the range for comparable vehicles, wait roughly 24 days, and if it hadn't sold, cut the price about 3 percent. They didn't hunt for the perfect price up front. They set a reasonable starting point, gave it a fixed window, and adjusted only if the car hadn't sold by the end of their target window.
The dealers weren't guessing worse than the perfectionists. They were just not pretending they had the time to guess perfectly.
A minority, about 19 percent and mostly smaller dealers, picked one price and never touched it. The dealers who adjusted did something an ad reviewer should notice: they shortened their waiting period as local competition increased. More competitors meant less patience for a slow sale. The threshold wasn't fixed forever. It moved based on real conditions, not on a feeling that the price could still be better. Ad owners could use the same kind of signal, a read that adjusts to how the ad is actually doing, and an outside read on your ad is what Knowing is built to give.
Why one more edit rarely finds the real problem.
Here's the part that should change how you spend that last hour before launch. If the ad is going to fail, it's probably not failing because of a word choice. It's failing because it makes people feel nothing.
System1 is a neuroscience and ad-testing firm that scores advertising on emotional response. Their research, summarized in a report on their Cost of Dull findings and built on work led by chief creative officer Orlando Wood, makes a case that should reorder your priorities. The research ties ad effectiveness directly to emotional response, not execution polish. System1 scores ads on discrete emotions, happiness, surprise, fear, disgust, anger, and produces two scores: one for long-term brand building, one for short-term sales impact. The surprising finding isn't that negative ads fail. It's that neutral, emotionally flat ads are the real failure mode, producing neither short-term nor long-term results.
The cost of that flatness is large and specific. System1 estimates that dull, emotionally flat advertising requires an extra 189 billion dollars in global media spend to match the impact of work that actually makes people feel something. Wood put it plainly: "Dull advertising leaves little trace." That's not a line about bad grammar or a clunky logo placement. It's a line about an ad that could be removed from someone's feed and leave no mark that it was ever there.
The research found certain categories chronically underperform on emotional engagement: business services, home and garden, car services. These are exactly the categories a lot of small business owners operate in. A plumber, an accountant, a landscaper are all fighting a category-wide habit of making flat, informational ads, which means the bar to stand out emotionally is lower than it looks.
What to actually do with your last hour.
If you're satisficing correctly, you set your bar before you start reviewing, not after. The bar isn't "does this feel finished to me tonight." It's a short, specific list: does the ad make someone feel something rather than nothing, is the offer clear, does it look like it belongs to your business and not a stock template, is there one idea instead of four competing for attention.
Once an ad clears that list, more editing time mostly produces motion, not improvement, the same way an extra week on a car's price tag mostly produces anxiety, not a better sale. The dealers in the study didn't keep second-guessing a price that was reasonable. They set it, watched, and moved only when there was real evidence to move on.
The research on dullness also tells you where to spend that last hour if you insist on spending it. Not on wording, which is what most owners default to, but on whether the ad tells a small story, uses drama, or uses music, the research found that storytelling, drama, and music outperform voiceover, on-screen text, or a straight testimonial. For shorter formats like a 15-second video or a single static image, that same research found a consistent, recognizable element, a face, a tagline, a jingle, outperforms trying to build emotional impact from scratch in a few seconds.
The real risk isn't launching too soon.
Owners who keep polishing usually believe they're protecting themselves from a bad ad. More often they're protecting themselves from the discomfort of publishing something imperfect. Simon's model predicts something almost funny about this: two nearly identical ads, reviewed by two different owners, can sit in draft for wildly different amounts of time, purely based on which owner set a looser threshold going in. The ad isn't better for having waited. It's just waited longer.
The actual risk sitting in most ad drafts isn't a typo or a slightly weak call to action. It's the flat, competent, forgettable version that clears every grammatical bar and fails the only one that counts: did anyone feel anything. That's worth one more look before you hit publish. The fifth look at your headline font probably isn't.
Frequently asked
Common questions.
How do I know when an ad is good enough to stop editing?
Set your bar before you start reviewing: clear emotional response, a clear offer, one main idea, and a look that matches your business. Once an ad meets all four, more editing tends to produce motion rather than real improvement. Treat the bar as fixed, not something you quietly raise each time you reread the ad.
Is it better to launch an imperfect ad or keep revising?
Research on decision-making, including Herbert Simon's work on satisficing, suggests setting a reasonable threshold and acting once an option meets it, rather than searching for a theoretically perfect version. For most small business ads, an acceptable ad that runs beats a flawless one that never ships.
What actually makes an ad fail, if not weak writing?
System1's research on the 'Cost of Dull' found that emotionally flat ads, not poorly written ones, are the real failure mode, producing neither short-term sales lift nor long-term brand impact. An ad can be grammatically fine and still leave no impression, which is the costlier problem.