Know what to expect before you spend a dollar.

When your campaign launches, Facebook enters what it calls the learning phase. Its system is testing your ads against different people in your area to figure out who actually fills out the form. During this period results swing hard day to day - a $90 lead on Tuesday, two $20 leads on Wednesday. That is the system working, not failing.

Every edit restarts the clock. Changing the budget, swapping an image, or pausing and resuming tells Facebook to start learning again from scratch. The single most expensive habit in contractor advertising is fiddling with a campaign during its first two weeks.

Our system is watching the whole time. We track every ad daily and have automatic rules for killing real losers and shifting budget to winners. The difference between us and you checking the dashboard hourly: we act on statistical evidence, not on one bad morning.

Early in a campaign, Facebook has not yet learned which of your neighbors are likely to become leads, so it shows your ads broadly and your cost per lead runs high. As the system collects results, it narrows in on lookalikes of the people who already converted, and cost per lead typically drifts down over the first several weeks.

This is why judging a campaign on week 1 numbers almost always leads to the wrong decision. The fair checkpoint is after 2-3 weeks of uninterrupted running.

It is also why we run multiple ad angles. Different messages attract different homeowners, and the early data tells us which angle deserves your budget. Killing the campaign early means paying for the lesson without collecting the reward.

A lot happens behind the dashboard. Every day we pull your real performance numbers from Facebook and score every ad. When an ad goes stale (3+ weeks old) or its click rate drops meaningfully, we rotate in fresh creative. When one angle clearly beats another, we shift budget toward the winner. When an ad would violate Facebook's housing-advertising rules, we catch it before it ships.

Every call to your tracking number is logged, recorded where legal, and turned into a lead in your inbox automatically. Miss a call and the caller gets a text within a minute so the lead does not go cold.

What we never do: change what you are spending without your say, or launch anything you have not approved. Your budget and your approval are the two controls that stay in your hands.

Industry studies are consistent on this: a lead contacted within 5 minutes is many times more likely to become a job than one contacted after an hour. Homeowners fill out three forms and hire whoever calls back first.

We close part of this gap for you - new form leads get an instant text from us (where they have agreed to texting), and missed calls trigger an automatic text back. But the call itself is yours to make. The contractors who win with ads treat a new lead like a fire bell, not an email.

A practical rule: if you cannot answer between jobs, return every lead call at lunch and at the end of the day, same day, every day.

Raising your budget Monday, cutting it Thursday, and raising it again next week feels like careful management. To Facebook's system it looks like instability, and it responds by re-entering learning mode - the expensive, erratic phase - over and over.

If you want to scale up, the safe move is one increase of up to about 20-30%, then let it settle for several days. If you need to cut back, one decisive cut beats three small ones. And if money is tight for a season, pausing cleanly and resuming later beats starving the campaign at $5/day.

When you change your budget in TopKnock, we apply it as a single clean change on the Facebook side for exactly this reason.