If you’ve been running Local Services Ads and you’re now looking at Performance Max pay-per-lead, the first thing worth saying is that the switch is more of a change in reporting than a change in what your phone does. The leads still come from people searching for your trade in your area. What changes is where the lead is recorded, how the cost shows up, and how much of it you can see.
I’ve walked a few home service companies through this now, so here’s an honest week-by-week picture of the first 60 days, plus the handful of things I’d check at each stage. None of it is complicated. It’s just that nothing in the interface tells you when to look.
Before day one: get your history out
The single most important thing to do before the move is export your Local Services Ads lead history. Your lead log, your dispute history, your monthly charges and your booking outcomes if you’ve tracked them. Once a dashboard is retired, the export button goes with it, and that history is the only benchmark you’ll have for judging whether the new setup is doing better or worse.
At minimum, pull the last twelve months and write down three numbers: total spend, total leads, and your average cost per lead. Two more if you have them: how many of those leads booked, and what a booked job is worth to you on average. Those five numbers are your before picture.
Week one: learning, and it looks bad
The first week is the one that makes people nervous. Volume is uneven, cost per lead jumps around, and some days produce nothing. This is the learning period doing its job. The system is testing which searches, placements and audiences produce a lead for you, and early on it’s paying tuition.
What I’d do in week one: almost nothing to the campaign, and quite a lot to the tracking. Confirm that every way a lead can reach you is actually being recorded. Form submissions, phone calls from the ads, calls from your Google Business Profile, chat if you have it. This is the point where problems get caught cheaply. I’ve had a brand-new site go live with the tracking fields never filled in, and weeks of data went nowhere before anyone noticed, because nothing in the platform prompts you to check.
One number to write down at the end of week one: how many leads the platform says you got, versus how many your team says they handled. If those two disagree by more than a little, fix that before you judge anything else.
Weeks two and three: leave it alone, watch the mix
By week two the daily swings usually start to settle. Cost per lead is still not meaningful yet, but the mix is. Look at where the leads are coming from and what people are actually asking for. If you do foundation repair and half the calls are asking about fence posts, that’s a targeting signal, not a lead quality problem.
Resist the urge to change budgets or targets in these two weeks. Every meaningful change restarts some of the learning, so a week of edits can turn a 3-week ramp into a 6-week one. The most common way I see this go wrong isn’t a bad setup, it’s a nervous week three.
Week four: your first real read
At about 30 days you have enough to compare against your Local Services Ads history, with one caveat: compare cost per lead only if you’re counting leads the same way. Local Services Ads counted a charged lead. Performance Max counts a conversion, and conversions include things you may have set up yourself, like a form fill or a 30-second call. If your new number counts more things as a lead, of course your cost per lead looks better. That’s a definition change, not a win.
So make the comparison honest. Pick one definition, ideally a phone call over 60 seconds or a submitted form from a real person, count both periods that way, and then compare.
Here’s the worked math I’d use at day 30. Say you’re at $95 per lead and about a third of leads become jobs. That’s roughly $285 in ad cost per booked job. If your average ticket is $6,000, you’re spending under 5% of revenue to book the work, which is comfortable. If your close rate is closer to one in six, that same $95 lead costs you $570 per booked job, and now the conversation is about the phone, not the ads.
Weeks five to eight: the tuning window
This is when changes actually pay off. In order of what usually matters most:
- Location. Check performance by area and trim the places that produce calls you can’t profitably serve. Drive time is a cost, and the platform doesn’t know that.
- Search terms. Look at what people typed. Add negatives for the trades and services you don’t do. This is the cheapest fix in almost any account.
- Budget. If you’re regularly running out of budget before the end of the day, the number to look for is impression share lost to budget. That means your daily budget ran out mid-day, not that customers picked a competitor.
- Assets. Swap the weakest headlines and images. Add real photos of your crews and your work if you have them; stock images underperform in the trades pretty consistently.
- Target. Only now, once cost per lead is stable, is it worth tightening or loosening your target.
One at a time, roughly a week apart, so you can tell what did what.
What good looks like at 60 days
- Cost per lead within about 20% of your Local Services Ads number, measured the same way. Often a bit higher at first, then better as the targeting sharpens.
- Day-to-day lead volume steady enough that a slow Tuesday doesn’t feel like a crisis.
- Impression share lost to budget under about 20%, unless you’re deliberately capping spend.
- Every lead source tracked, and the platform’s lead count matching what your team handled.
- A cost per booked job you can say out loud, not just a cost per lead.
If you’re not there at 60 days, the honest next step is usually diagnosis rather than more budget. Two things worth checking first: whether tracking is double-counting, and whether the lead-handling side is losing bookings that the ads already paid for. I’ve seen a business with perfectly healthy lead volume where only about two thirds of leads ever reached a scheduled appointment. Closing that gap added around fifty appointments a month with no extra ad spend at all.
A short checklist you can run yourself
Click into your campaign, then work through this list. Insights and reports for the search terms and locations, campaign settings for the budget and target, and your conversion actions list for the tracking.
- Are all my lead types recorded as conversions, and only once each?
- Does the platform’s lead count match what my team handled last week?
- What percentage of impression share am I losing to budget?
- Which locations produced leads I couldn’t serve?
- What did people actually search for, and what should I add as a negative?
- What’s my cost per booked job, not just per lead?
If you’d like a second set of eyes on the first 60 days, that’s what a marketing audit is for. And if you want to get sharper on the numbers behind all of this, start with cost per lead vs cost per booked job, which is the comparison that makes every other decision easier.