Asked to choose between Tesla and Apple, ChatGPT declined to name a winner. What it produced instead, according to the MoneyLion piece syndicated on Yahoo Finance, was a diagnostic: Apple for the investor who wants stable compounding and will accept lower headline gains, Tesla for the one with a long horizon, genuine conviction in autonomous driving and robotics, and the stomach to hold through a stock that "can drop 30% on a missed delivery number." The article judged that framework more useful than a pick, and on that point it was correct.
The same exchange occurs whenever a business owner asks an AI which marketing channel to buy. The honest answer is never a channel. It is a set of conditions under which each channel makes sense, because the variable that decides the question is not the channel's quality but the buyer's own tolerance, horizon, and willingness to supervise the position after purchase. Google Local Services Ads illustrate this with unusual clarity.
What Local Services Ads actually are, and what they cost
Local Services Ads are Google's pay-per-lead placement, which sits above the map pack and carries the Google Guaranteed badge. Unlike Google Ads, where the advertiser pays per click and controls keywords, LSA charges for the lead itself while surrendering nearly all targeting control: the advertiser selects service categories and a service radius, and Google determines the rest.
The economics are well documented. A February 2026 benchmark from SearchLight Digital, drawn from $6.72 million in observed spend across 888 contractors and 126,650 leads, put the blended cost per lead at $53, ranging from $39 in electrical to $59 in drain and sewer work. Data-Driven Trades reported a May 2026 blended figure of $63.29 per unique lead across 1,011 home services businesses. Most contractors land somewhere between $20 and $80 per lead, though water damage restoration in major metros can reach $150.
Set against traditional search advertising, the comparison favors LSA on the metric that matters. Cost per paying customer runs approximately $233 through LSA against $472 for blended Google Ads, with an average book rate of 43.9%. Whether a particular business realizes that advantage is a separate question.
The Tesla problem: high intent, low control
LSA behaves like the volatile position in the portfolio. Intent runs high, since a caller who taps a Google Guaranteed listing is generally ready to hire, but control runs low, because the advertiser cannot bid on specific terms, exclude search queries, or dictate which calls arrive. Lead flow can therefore swing in ways that a keyword-level campaign does not, and the business absorbs that variance directly.
The comparison to a high-volatility equity holds under scrutiny. Both instruments reward the buyer who can tolerate the swing and punish the one who cannot, and in neither case does the underlying asset's quality settle the question of whether it belongs in a given portfolio. A shop that can absorb an uneven week and work every call that arrives extracts real value from LSA. A shop that cannot answer the phone reliably converts the same spend into billed leads that were never worked.
Where the AI now grades its own homework
The parallel acquires practical consequences at this point. Google formerly permitted advertisers to dispute leads manually, a process through which contractors recovered credits on more than half of what they contested. That mechanism has been deprecated. Google now applies its own AI to identify unqualified leads and credit them automatically, and the results are considerably less generous: practitioners managing LSA accounts report feedback-to-credit conversion in the range of 15% to 25%, meaning roughly one credit for every four to six bad leads flagged.
An automated system that adjudicates the fees it also charges is precisely the arrangement that warrants human supervision. The AI is not malicious; it is simply optimizing something other than the advertiser's cost per booked job, and nobody outside the account has an interest in flagging what it misses. Someone has to review the leads, submit the feedback, and reconcile what was actually credited against what was actually junk, week over week. Automating the disputes did not retire that obligation. It made the obligation easier to neglect, which is a different thing entirely.
This is the argument developed at greater length in AI with a human in the loop, which examines the same principle in investigative work: automation is trustworthy in proportion to how clearly a person can intervene when it errs.
The criteria that actually decide it
LSA suits a business that answers its phone consistently, operates in a trade with sufficient lead volume to smooth the variance, carries margins that survive a $53 average lead cost, and has someone reviewing lead quality against credits every week. Where any of those conditions fails, the channel underperforms regardless of its benchmark averages, and the failure will be attributed to the channel rather than to the conditions.
Velora's Local Services Ads Management exists to hold the fourth condition, since it is the one businesses most reliably drop. The service covers profile setup and Google verification, budget and bid management, weekly lead-quality review and feedback, and a monthly cost-per-booked-lead report, at $500 to set up and $350 a month. The report is the point: cost per lead is the number Google shows you, and cost per booked job is the number that determines whether any of this was worth doing.
Frequently asked questions
Are Google Local Services Ads worth it in 2026?
For most home and local service businesses, yes, on cost grounds. Cost per paying customer averages roughly $233 through LSA against $472 for blended Google Ads, with an average book rate near 43.9%. The channel underperforms, however, where a business answers its phone inconsistently or leaves lead quality unreviewed.
How much do Local Services Ads cost per lead?
A February 2026 benchmark across 888 contractors and 126,650 leads put the blended cost per lead at $53, with electrical at $39, HVAC at $51, plumbing at $57, and drain and sewer at $59. Most contractors pay between $20 and $80, though restoration work in large metros can reach $150.
Can you still dispute bad leads on Local Services Ads?
Not in the original sense. Google deprecated manual lead disputes and now uses its own AI to identify and credit unqualified leads automatically. Advertisers submit feedback rather than formal disputes, and practitioners report credits on roughly 15% to 25% of what they flag, well below the recovery rate manual disputes once produced.
How do Local Services Ads differ from Google Ads?
LSA charges per lead and places the listing above the map pack with a Google Guaranteed badge, but permits almost no targeting control beyond service category and radius. Google Ads charges per click and allows keyword-level bidding and negative keywords. LSA generally delivers higher intent with less control.
Should you ask an AI which marketing channel to buy?
An AI is useful for producing the criteria that decide the question and largely useless for making the decision, much as ChatGPT declined to choose between Tesla and Apple and returned a framework keyed to the investor's own tolerance and horizon. The determining variables are internal to the business, and the AI has no access to them.
Related reading
For the same supervision principle applied outside advertising, see AI with a human in the loop. For where an owner's week actually goes, and why weekly review tasks are the first to lapse, see where a small business owner's day actually goes.
Where to look next
Local Services Ads Management runs $500 to set up and $350 a month, and it includes the weekly lead review and the monthly cost-per-booked-lead report. If you would rather establish first whether your business meets the conditions above, the discovery call is free and it begins with your current call handling and margins.