The best time to sell a small business is while it is still boring and growing — which is precisely when it feels least urgent. Most people sell after the decline has started and the numbers have recorded it.
Signal 1: you've stopped shipping
Look at your commit history over twelve months. If meaningful work stopped six months ago and revenue is holding, you have a window. Buyers pay for trailing performance; a business coasting on past work looks identical to a maintained one — for about two more quarters.
Signal 2: growth has flattened but not turned
Flat is a fine time to sell. Declining is not. The difference in multiple between "flat for twelve months" and "down 15% over twelve months" is roughly a full turn, and it can't be argued away — the chart is the chart.
Signal 3: the next step needs a skill you don't have
Plenty of small products hit a ceiling that isn't about the product. It needs sales, or SEO, or paid acquisition, or enterprise support — and you don't want to become that person. That's a genuinely good reason to sell, and buyers respect it, because it's the version of the business they're buying.
Signal 4: it's become a chore
Not romantic, but it's the most reliable predictor of what happens next. Resented businesses get neglected, and neglect shows up in the numbers within two quarters. Selling at the point of boredom captures value that selling at the point of resentment does not.
The bad reasons to wait
"Let me hit $10k MRR first." If growth is 2% a month, that's eighteen months of a business you've stopped enjoying, to gain maybe half a turn.
"Let me finish the rewrite." Buyers pay for revenue. A rewrite is cost you incur and they benefit from.
"The market might improve." Multiples move slowly; your business's trajectory moves faster. Bet on the thing you control.
The one good reason to wait
Preparation. If your numbers can't be verified at source, your costs are mingled with personal spending, and nothing is documented, then six months of preparation is worth more than six months of growth. That is the only delay that reliably pays for itself.