Google Local Services Ads sit at the very top of the results page, above the map pack and above the text ads, and they charge per lead rather than per click. For the trades and professional categories Google has opened them to, that is the most valuable position on the page, and it is sold on a completely different mechanic from anything else in the account. That last part is what most advertisers miss. An LSA listing is not a Search campaign with a different layout; it has no keywords, no ad copy and no landing page, and almost everything that decides how it performs happens off the screen, in how the business answers its phone and collects its reviews. What follows is the operating manual, with the parts that change often flagged as things to read in Google's own documentation rather than trust from an article.
Who can actually run it
Eligibility is decided by category and by country, and Google adjusts both. The shape has been stable for years: home services trades — plumbing, heating and cooling, electrical, roofing, garage doors, locksmiths, pest control, cleaning and their neighbours — carry the Google Guaranteed badge, while professional service categories such as lawyers, financial planners and real estate agents carry Google Screened. Categories have been added and occasionally withdrawn, and availability differs between countries and sometimes between cities. So the first ten minutes of any LSA plan should be spent in Google's current Local Services Ads category list for your own market, not in a blog post. If your category is not there, nothing else in this article applies to you yet, and knowing that early is worth a quarter.
The second eligibility question is licensing, and it is stricter than most advertisers expect. Verification typically requires proof of a current trade or professional licence, evidence of general liability insurance at the level Google specifies for the category, a background check on the business, and — in many trades — background checks on individual technicians or practitioners. There is no way to accelerate this by spending more. It is paperwork, it takes as long as it takes, and nothing runs until it clears. The single most common self-inflicted delay is starting the applications after the budget has been approved rather than before.
The badge is the product
It is worth being precise about what the two badges promise, because they are not the same promise. Google Screened tells a searcher that the business, and often the named practitioner, has passed licence and background screening. Google Guaranteed adds a customer-facing guarantee: a customer dissatisfied with work booked through the ad may be reimbursed by Google, subject to a lifetime limit per business that Google publishes and that varies by country. Read the current limit in Google's documentation for your market rather than accepting a figure from anywhere else, including from a competitor's sales deck.
The advertiser-side consequence of the guarantee is operational rather than financial. Because Google is standing behind the work, complaint history feeds into whether you keep the badge, and losing the badge does not reduce your performance so much as end it — the whole reason the placement converts is the badge. That reframes customer service as a media buying activity, which sounds like a stretch until the first unresolved dispute. The businesses that do well here are the ones that treat a complaint as an account emergency rather than an admin task.
Why the ranking works nothing like a Search auction
In a Search campaign, placement comes out of a keyword auction in which your bid and your ad quality are the main determinants, and the things you tune are keywords, copy, extensions and landing pages. In LSA none of those exist. Google's published Local Services Ads documentation describes ranking as being influenced by proximity to the searcher, review score and the number of reviews, responsiveness to calls and messages, whether you are open at the moment of the search, and complaint history.
Read that list again and notice what is not on it: cleverness. There is no copy to write, no landing page to optimise, no keyword research that will move you up. The levers are all operational, which is genuinely good news for a well-run business and genuinely bad news for one that buys its way out of problems. Three consequences follow directly. Review recency is a lever you can pull every week, which is why a company adding fresh reviews steadily tends to do better here than one sitting on a large but ageing pile — the mechanics of that are the same ones behind review generation generally. Hours must be honest, because being listed as open and not answering is worse than being listed as closed. And a missed call is charged twice: once as the job you did not win, and once as a responsiveness signal you cannot get back.
LSA is not an auction you win with money. It is an auction you win with answered phones and fresh reviews.
Setting the budget and the lead price
Budget in LSA controls exposure rather than position: it decides how many leads you can absorb, not how trustworthy you appear. Google has offered both an automated bidding mode that aims to maximise lead volume and a manual mode where you set a maximum you are willing to pay per lead, and the interface for both has changed more than once, so confirm the current options in the account rather than in an article. The strategic point survives every interface change: the automated mode is a reasonable way to calibrate a first month, and operators who know their numbers almost always move to a manual price afterwards.
Set that price from your own arithmetic rather than from a suggestion. Take the gross margin on a typical job, multiply by the proportion of valid leads you actually close, and you have the point at which a lead stops paying for itself; your target sits below that line, and how far below is a decision about how quickly you want the money back rather than a number anyone else can give you. The reason this matters more here than in most channels is that a single price sits across job types of wildly different value. A lead price that is trivial against a full system replacement is ruinous against a small repair. Where your categories differ that much, separate what the platform lets you separate, watch the mix rather than the average, and revisit the price seasonally — demand in most trades is not flat across a year and neither is competition.
The dispute rhythm
You pay per lead, and not every lead is a lead. Wrong service, wrong geography, solicitors, job applicants, wrong numbers and duplicates all arrive through the same phone line and all get charged the same way. Google lets you dispute charges that were not genuine opportunities and issues a credit where the dispute is accepted. There is a window for raising one, stated in Google's documentation, and it is short enough that a monthly review will miss some of them.
So make it weekly and make it somebody's named job. The rhythm that works is a standing slot — one person, one sitting — in which the week's recorded calls and messages are reviewed, each lead is tagged valid or invalid with a reason, and the disputes are filed in one pass. Two things make it worth the hour. The credits themselves are real money that nobody else is going to claim on your behalf. And the tagging produces the only honest view of your true cost per valid lead, which is the number your bid should be set from and which no dashboard will calculate for you. An account that never disputes is not just donating the invalid leads; it is bidding on a cost per lead it has never actually measured.
Where LSA fits against Search and the map pack
Three products, one results page, three different mechanics. The confusion between them is the source of most bad local strategy, so it is worth laying them out side by side.
| Surface | What you buy | What decides position | What you control |
|---|---|---|---|
| Local Services Ads | A lead, charged per contact | Proximity, reviews, responsiveness, hours, complaint history | Job types and service area only. No keywords, no landing page |
| Search ads | A click, charged per click | Bid and ad quality in a keyword auction | Keywords, copy, extensions, landing page, audiences, schedule |
| The map pack | Nothing. It is earned, not bought | Relevance, distance and prominence | Your profile, categories, reviews, citations and site authority |
Ranking inputs as described in Google's published documentation, read September 2026. Not a CurrentAds finding.
LSA wins on urgent and commodity intent, where the caller wants a badge, a rating and a phone number and nothing else: the burst pipe, the dead air conditioning, the lock at midnight. Search wins wherever the decision is considered, because that is where you need a page that argues, a specific message, retargeting and control over which product the click lands on — none of which LSA offers at all. The map pack is the third thing and it is not bought: it is earned through relevance, distance and prominence, and it is the surface that keeps producing when the ad budget pauses. How that one is actually decided is set out in how the map pack actually ranks.
There is also a ceiling worth planning around. LSA volume is bounded by your review profile, your service area and the number of people searching in it, so past a certain point more budget stops buying more leads. Search volume is bounded mostly by budget and the size of the keyword market. That difference alone is why the honest configuration for most home service businesses is both: LSA as the always-on floor, Search layered on for high-ticket work and for the volume LSA cannot supply, with the map pack and the profile underneath holding up all three. Anyone selling you one of them as a religion is selling you their staffing model.
What LSA cannot do
An honest playbook needs this section. You cannot say anything specific in an LSA placement: no offer, no differentiator, no explanation of why you are the right firm for an unusual job. You cannot send the click to a page you designed. You cannot retarget the people who saw you and chose someone else. You get little in the way of keyword-level insight into what people were actually searching for. And you are dependent on a badge that a policy change or an unresolved complaint can remove. None of that is an argument against running it. It is an argument against running it alone, and against building a business whose entire lead flow sits inside one product owned by one company that can change its rules without asking you.
The weekly operating checklist
- 01 Review every charged lead, tag it valid or invalid with a reason, and file the disputes in one sitting.
- 02 Check missed and unanswered calls. Each one is a lost job and a responsiveness signal.
- 03 Count new reviews added this week. If the answer is zero, the close-out process is not asking.
- 04 Recalculate cost per valid lead from your own tags, not from the platform average.
- 05 Confirm the job types still match the work you want, and that nothing has been switched back on.
- 06 Confirm hours, licence documents and insurance certificates are current. Expiry pauses everything.
None of that is difficult and none of it is glamorous, which is exactly why the channel rewards it. Run LSA like an operator and it becomes the most predictable lead source in a local account. Run it like a directory listing and it becomes a quiet tax with a badge on it. It is one of the ten disciplines we run inside local and maps, and the price for that scope is published on the pricing page like everything else.
One warning about the weekly tagging, because it is where most LSA accounts quietly lose the plot. Tagging leads valid or invalid inside the LSA interface tells you what the channel cost. It does not tell you what happened to those leads afterwards, and a lead that was genuine, answered, and then arrived in the CRM with an empty source field is invisible to every calculation you are about to make. The nine places that source gets dropped, and the checks that find them, are in why leads arrive with no source.