Most local service accounts swing between two bad habits. Either nobody logs in for six weeks, or the owner logs in every morning, sees a $94 cost per lead on a Tuesday, and starts pausing keywords.
The second habit is worse. Automated bidding needs a stable account to learn from. Google's own Smart Bidding guidance says to wait at least two weeks without changes during the initial learning period, and that campaigns with fewer than 30 conversions in the measurement window can take up to four weeks to settle. If you change the target CPA on Monday and the budget on Thursday, you are resetting the clock on a system that was never given a chance to work.
So the goal of a weekly routine is not to make a lot of changes. It is to catch problems early, remove waste, and then leave the account alone.
Here is the routine. It takes about 20 minutes. Do it the same day every week — Monday morning works well because you can see the full prior week — and set the date range to the last 7 days unless a step says otherwise.
Minutes 0–3: Did anything break?
This is the step that saves the most money and takes the least time. You are looking for things that are broken, not things that are suboptimal.
Open the account overview and check four things:
- Are campaigns actually spending? A campaign showing $0 for three straight days is either paused, out of budget, or has no eligible ads. That is a dead week of lead flow you will never get back.
- Any disapproved ads or assets? A disapproval on your only responsive search ad in an ad group means that ad group is not running at all.
- Billing status. Expired cards are a shockingly common cause of "my phone stopped ringing."
- Any notification banner about an account change. Google is actively migrating Search campaigns to AI Max. Campaigns that used automatically created assets or the campaign-level broad match setting were auto-upgraded starting September 2026, and Google's documentation says Dynamic Search Ads campaigns are scheduled to upgrade automatically starting February 2027. If a banner appears telling you a setting is changing, read it. That is not optional housekeeping.
If everything is green, move on. Do not start optimizing here.
Minutes 3–8: Search terms
This is the highest-value five minutes in the whole routine for a local service business, because this is where you find out what you are actually paying for.
Go to Campaigns → Insights and reports → Search terms. Set the range to the last 14 days so you have enough volume to judge.
Sort by cost, highest first. Read down the list and ask one question per term: would I want the phone to ring for this?
Three outcomes:
- Clearly wrong intent — jobs you don't do, DIY searches, "free," "how to," job seekers, wrong city, competitor brand names you don't want. Add as a negative. Campaign-level for anything broadly irrelevant, ad group level for anything that just belongs somewhere else.
- Right intent, converting, not in your keyword list — add it as an exact match keyword in the most relevant ad group.
- Right intent, no conversions yet, small spend — leave it. Five clicks and no call is not a signal.
One caveat worth knowing: the search terms report does not show everything. Google omits terms that don't have enough query activity to clear its privacy thresholds. So the report is a sample of your waste, not a complete inventory. That is one of the reasons a standing negative keyword list matters more than reactive cleanup — our negative keyword guide for local service businesses covers building that list up front.
Realistically, most weeks you will add between zero and ten negatives. If you are adding forty every week, the problem is upstream in your match types and structure, not in your weekly maintenance.
Minutes 8–11: Budget and where the ceiling is
Add two columns to your campaign view if they aren't there already: Search lost IS (budget) and Search lost IS (rank). These tell you why you aren't showing up more, and they point at opposite fixes.
Lost impression share to budget means you're winning auctions but running out of money before the day ends. If those campaigns are producing leads at a cost you're happy with, this is the cleanest growth lever you have — raise the budget.
Lost impression share to rank means you're in the auction and losing it. More budget does nothing. That is an Ad Rank problem: bid, ad relevance, landing page quality.
Two discipline rules here. First, don't raise a budget on a campaign that isn't profitable — you're just buying more of a bad outcome faster. Second, change budgets in steps, not leaps. Large sudden budget swings can disturb a bidding strategy that has already learned your account.
Minutes 11–14: Does Google's conversion count match your reality?
Open your conversions column, then open your actual lead source — your CRM, your call log, your inbox. Compare the week.
You are not auditing attribution to two decimal places. You are checking for a mismatch big enough to mean something is broken. Google reporting 22 conversions when you logged 6 real inquiries usually means you're counting the wrong thing — page views, form loads, short calls, or a tag firing twice. Google reporting 3 when you booked 11 jobs means calls aren't being tracked at all.
This matters more than it looks, because every automated bidding decision in the account is made from this data. Bad conversion data doesn't produce slightly worse results, it produces confidently wrong ones. If this check keeps failing, stop doing weekly optimization and go fix the tracking — our articles on conversion tracking setup and call tracking for phone-driven businesses are the place to start.
Minutes 14–17: Ads and assets
Open the Ads tab. You're doing a quick creative check, not a rewrite.
Look at Ad Strength per ad group. Google's guidance is to run at least two responsive search ads rated "Good" or "Excellent" per ad group, and it reports that advertisers who move an RSA from "Poor" to "Excellent" see 15% more conversions on average. Treat Ad Strength as a completeness checklist rather than a performance prediction — it measures asset variety, not whether your offer is compelling.
Then click View asset details on your main ad and look at the asset performance ratings. If a headline is consistently marked "Low," swap it for something new. One swap. Not six.
If you change creative this week, that is your change for the week. Don't also move the bid target.
Minutes 17–20: Write down what you did
Keep a running log — a note file, a spreadsheet, anything. Date, what you changed, why, and what you expect to happen.
This sounds like busywork until the week your cost per lead jumps 40% and you need to know whether it was something you did, a competitor entering the market, or seasonality. Without a log you will guess, and you will usually guess wrong and "fix" something that wasn't broken.
If you want this as a structured system — the account build, the tracking setup, and the management cadence as one thing rather than a checklist bolted onto whatever you already have — that's what the Launchpad program is.
What does not belong in the weekly routine
A good routine is defined as much by what you skip.
Don't change bid strategy targets weekly. If you adjust target CPA every seven days, the system is permanently in learning. Review targets monthly at most, and move them in increments of 10–20%, not in half.
Don't pause keywords on thin data. A keyword with 8 clicks and no conversions has told you nothing. For most local service businesses, you need a month or more of data on a single keyword before the number means anything.
Don't chase optimization score. The score measures how much of Google's advice you've adopted. Dismissing a recommendation raises it just as much as accepting one. Some recommendations are genuinely useful — conversion tracking upgrades, adding assets. Others, like broadening match types or raising budgets, exist to increase spend. Read them, apply the ones that fit your goals, dismiss the rest with a clear reason, and ignore the percentage.
Be careful with auto-apply. If Google is applying recommendations automatically, it can change targeting or budgets without you seeing it. At minimum, know which toggles are on. For most small local accounts, keeping auto-apply limited to low-risk hygiene items is the safer default.
The monthly version
Once a month, spend another 30 minutes on the things that need more data than a week provides: geographic performance by city or zip, device splits, day and hour patterns, bid strategy targets, landing page conversion rate, and auction insights.
That's the whole system. Twenty minutes weekly to catch breakage and remove waste. Thirty minutes monthly to make strategic calls. The rest of the time, let the account run.
The accounts that do worst are not the neglected ones. They're the ones that get touched every day by someone reacting to noise.