Google Ads is fast to turn on and slow to get right

What does the first week actually buy you?

A new search campaign can serve its first impression within an hour of approval. That part of the pitch is true. What arrives with it is not qualified traffic; it is a list of the searches Google thinks your keywords are near, and most of that list is wrong at the start.

Treat the first three to four weeks as a diagnostic. You are buying a search terms report, not a return. Set the budget where a bad month is annoying rather than damaging, and plan to spend all of it learning what people in your area actually type.

Why the search terms report is the first month's real product

It sits under the keyword section of the account, and it shows what people searched rather than what you bid on. Open it twice a week for the first month.

What you will find, in roughly this order: people looking for a job at a company like yours, people looking up a salary figure, people trying to do it themselves, students, and competitors' business names. A new plumbing account with nobody watching it will typically pay for "plumber salary," "plumbing school," and three spellings of a competitor's name inside its first ten days.

Every one of those becomes a negative keyword. Build a single shared negative list at the account level so you maintain it once instead of per campaign, and add to it weekly for the first two months. This is unglamorous work and it is where nearly all of the early savings live.

Which settings decide where the money lands?

Location targeting has an option most accounts never touch. The default includes people who are merely interested in your area, which means a click can come from somebody two states away who was reading about the Jersey Shore. Switch it to the presence option so you pay for people who are physically here. On service accounts we take over, that one change routinely removes a tenth to a quarter of the waste.

The second is auto-apply recommendations, which is switched on for several recommendation types by default and will add keywords and broaden match types on your behalf. Turn off anything that changes keywords or targeting. Those suggestions optimize for the platform's definition of performance, which is not your cost per booked job.

While you are in there, confirm your search campaign is not quietly opted into display expansion, and make a deliberate decision about search partners instead of inheriting one.

Why does automated bidding stall on a small account?

Smart bidding learns from conversions. Below roughly thirty conversions in thirty days there is not enough signal, and the system swings between expensive experiments while it looks for one. A five-hundred-dollar-a-month account is not going to reach that in its first month, and no amount of waiting changes the arithmetic.

Start on manual bidding or maximize clicks with a short list of tightly matched phrase keywords. Move to a conversion-based strategy once the volume supports it. Doing it in the other order is how small accounts end up spending forty dollars on a click that was never going to book anything.

Make sure a conversion is a real one while you are at it. Count phone calls that last longer than sixty seconds rather than thirty, and do not count a thank-you page visit as a second conversion on top of the form submission. Double-counted conversions teach the system to buy the wrong traffic very efficiently.

Where do the leads actually get lost?

Between the click and the call. The ad promises emergency drain service and the click lands on a homepage listing eleven services. The page takes six seconds on cell service. The form asks for a street address before it asks for a phone number. Nobody answers after five, and the ads run until nine.

Match the landing page to the ad group, one service per page, with a tappable phone number at the top. If you cannot answer the phone during certain hours, use ad scheduling and stop bidding during them. Paying for a click that reaches voicemail is the most expensive thing that happens in a small account, and it happens every single night in accounts nobody has scheduled.

What should I expect by month three?

By the end of month one you should have a clean search terms report and a negative list that is doing work. By month two, a cost per call you can say out loud. By month three, a cost per booked job, which is the only number that decides whether the channel stays.

Digital Branding Blocks reports both numbers on one page each month, and if they are not moving by the third report we would rather change the plan than ask for more patience. Three months is long enough to know, and short enough that being wrong is survivable.

When is paid search the wrong answer?

Ads capture demand. They do not create it. If almost nobody searches for the thing you sell, no budget or optimization manufactures those searches, and the honest answer is a different channel entirely.

It is also poor value when the average job is small and the close rate is low, because there is no room between the click price and the margin. It works best where a job is worth several hundred dollars or more and the need is immediate, which is why New Jersey home service trades tend to do well with it and a shop selling twenty-dollar items does not. When the math already works and you need somebody to keep it working, that is the point at which paid search management pays for itself.