When the valuable work arrives

The busiest month isn’t the most valuable one, and the gap is wide enough to plan around.

Lead volume climbs through spring and peaks in May. Value runs on a different clock. In the first half of 2026 the average quote written against a February inquiry was $22,838, with 28% of February quotes at $20,000 or more. April, the month with more than three times February’s inquiry volume, averaged $12,323 with 15% above $20,000. The year before ran the same way. February averaged the highest of any month and 30% of its quotes cleared $20,000, against 16% in June.

So the early months bring fewer inquiries and better ones. February sat at 35.5% of the year’s peak month by volume in 2026, and 28.0% in 2025.

The intuition this cuts against is a common one. Budget tends to follow the phone, which means it lands in April and May when inquiry counts are highest. The figures suggest larger design and construction work is decided earlier, by people planning a season rather than reacting to one. A quiet January doesn’t mean a slow market. It means fewer inquiries that are more considered.

What each channel brings

Organic search is the largest single source of landscape inquiries in this index at 28.2% of leads, ahead of direct at 16.3%, Performance Max at 15.1% and paid search at 12.8%.

On quality the order changes. Paid search returns a quotable lead 34.0% of the time and Local Service Ads 34.3%, against 28.5% for Performance Max, which is the larger of the two paid channels by volume. That gap was wider a year earlier, when paid search ran at 33.2% against Performance Max’s 21.0%, so the automated channel has closed most of the distance without closing all of it. If your own mix has moved toward Performance Max, check whether your qualification rate moved with it.

Google Business Profile is the quiet performer. It’s the smallest meaningful source at 4.8% of leads, and its quotable rate of 26.4% is unremarkable, but the jobs behind it are big. It carries the highest average quote of any channel in the index, in both years. Read that dollar figure alongside the caveats on hand-entered quote values further down, but the direction is consistent, and it’s a reasonable argument for giving the profile more attention than its lead count alone would justify.

Local Service Ads deserve a caveat of their own. They are US-only, so that row describes a subset of the index rather than all of it.

How the lead arrives changes what it’s worth

How someone chooses to get in touch predicts what the inquiry is worth, and the effect is large.

A web form reaches quotable status 45.6% of the time. A phone call reaches it 19.4% of the time. The same ordering held a year earlier, at 42.9% against 18.6%. So a form is about two and a half times likelier than a call to turn into something worth quoting, and that ratio has been steady across both halves measured.

This isn’t an argument that calls are bad. Phone is the larger channel by volume, at 55.8% of all inquiries, and a call is how an urgent, smaller job tends to arrive. The point is that the two are different kinds of inquiry and averaging them together hides it. If your intake treats every lead as equivalent, your mix of forms to calls is quietly moving your numbers.

There’s also a less comfortable explanation for part of that gap, and it’s the subject of the next section.

The calls that go unanswered

Call outcome isn’t carried in the lead export, so this section draws on a separate call-tracking source covering April to June 2026, a shorter window than the rest of this report.

Roughly one call in three is never answered.

Across 34,390 tracked calls, 10,609 went unanswered or to voicemail. That’s 30.8%.

Set against the section above, it changes what that comparison means. Phone inquiries reach quotable status 19.4% of the time against 45.6% for web forms, and the easy reading is that people who call are simply less serious. If a third of those calls never reached a person, part of the gap isn’t about the caller at all. It’s about what happened after they dialed.

How much of it? Take the phone quotable rate and divide it by the share of calls that were answered, and it rises to 28.5%. That’s an illustration rather than an estimate, and it’s deliberately the most generous version: it assumes a missed call never qualifies, when in practice some callers try again or leave a message that gets returned. Read 28.5% as the ceiling on what better answering could recover, not as the answer.

Even at the ceiling, forms still qualify better than calls. The gap doesn’t close, but it narrows considerably, and the part that narrows is the part a business can actually do something about without buying a single extra lead.

That’s the reason this sits in a benchmark at all. Every other number in this report is about what the market costs and what it returns. This one is about a leak in the bucket, on the larger half of all inquiries, and it’s cheaper to fix than anything else on this page.

What the work is worth

Figures below are for 1 Jan – 30 Jun 2025, the most recent period whose sale outcomes have settled. The current half is still filling in, so publishing it here would understate every figure. Channels marked below that had too little history to appear on that basis are shown from the current half instead, with the number of quotes behind them.

Beyond what a lead costs, channels differ in the size of the job behind it, and the gap is wide enough to change where a business spends.

Local Service Ads produce the smallest work by a distance, at a $5,249 average quote against $17,475 from paid search and $22,507 from Performance Max, which puts LSA jobs at between a quarter and a third the size.

Google Business Profile sits at the other end. It’s the smallest meaningful source in the index by volume, but the work behind it is the largest of the settled channels at a $24,742 average, well above paid search. Organic search, the largest source by volume, sits high on value too at $21,191. The channels people don’t buy directly carry some of the best work in the index.

The highest average in the table belongs to none of them. Inquiries referred by an AI assistant averaged $45,071, roughly double Google Business Profile and about two and a half times paid search. That row is on a different footing from the others and the table says so: it comes from the current half rather than the settled one, so it could move a long way in either direction.

Treated with that caution, it’s consistent with everything else this report says about the channel. AI-referred inquiries also carry the highest quotable rate in the index at 48.3%. Few people arrive this way, but the ones who do seem to arrive further along, having already described the job to something that answered them, and they appear to be asking about bigger work.

This doesn’t rank the channels so much as sort them by the job each one is good for. If the problem is keeping crews busy through the season, the LSA economics work hard. If the problem is winning the larger build, a cheap lead that quotes small is the wrong thing to optimize for, and cost per lead alone will point you the wrong way.

A last caveat on timing. These figures are drawn from the settled half rather than the most recent one, because quote values keep arriving for months after a period closes. A current-period version of this table would understate every channel, and the ones with longer sales cycles most of all.

A signal worth watching

This is a small thing to watch rather than act on.

A year ago the referrals came from a single AI assistant. This year they come from four. AI-referred inquiries carried the highest quotable rate of any source in the index at 48.3%, and as the job-size table shows, they carry the highest average quote of any channel measured, at $45,071. They bring the fewest leads but the best qualified and largest jobs, on the thinnest evidence in the report.

The more important point is that this number is a floor, and probably a long way below the truth. Referral attribution only catches a lead whose browser passed the assistant along as a referrer. Someone who asks an assistant which landscaper to use, then searches the business by name, arrives as organic search. Someone who types the address arrives as direct, and direct grew from 13.3% to 16.3% of inquiries over the same period. None of that is proof. It does mean the measurable figure is the smallest possible version of this trend rather than the likeliest one.

First half 2026 compared with first half 2025, across the residential landscape campaigns in the index.

What the numbers say

Headline findings

Paid search got more expensive to enter and better to be in.

A click cost $2.91 in the first half of 2026, against $2.36 a year earlier. That’s a rise of 23.6%. Fewer of the people who saw an ad clicked it, too, with click-through falling from 3.17% to 2.62%. On the face of it, that looks like a worse market.

The rest of the funnel says otherwise. The people who did click converted more often, with conversion rate up from 1.87% to 2.12%, and they were far likelier to be worth quoting. Just under half of all leads reached quotable status in 2026, at 49.2% against 30.2% the year before. Nothing else in the dataset moved that far.

What follows from all that is the number to take away. A quotable lead cost $279.61 in the first half of 2026, down from $414.70. That’s 32.6% cheaper, in a market where the traffic itself got a quarter more expensive. In other words, qualification improved faster than media costs rose.

The falling click-through rate is the figure people ask about, so here is what this index can and can’t say about it. Two things would produce a fall like that, and only one of them is measurable here.

The measurable one is the buying mix. Performance Max took 40.3% of design-build spend in 2026 against 27.6% a year earlier, and it places ads across inventory where low click-through is normal. A blended figure covering both buying modes drops on that shift by itself, with nothing changing in how people behave on a results page. On its own it accounts for a move of this size.

The other is the explanation everyone reaches for first: AI answers now sitting above the results on a growing share of searches, taking the informational clicks and leaving the commercial ones. The data is consistent with that. Fewer clicks, converting more often and qualifying far more often, is what it would look like. It’s also what the mix shift looks like, and this index can’t separate them. Treat it as a hypothesis rather than a finding. Settling it would take click-through split by buying mode and by query type, which this data doesn’t carry.

What the fall doesn’t appear to mean is a worse market. Whichever mechanism sits behind it, the clicks that remained were worth more than the ones that went.

Cost and qualification

Cost per unique lead rose 9.8% to $137.60. That’s the one front-line number that moved against the buyer this year, and read on its own it looks like diminishing returns.

Next to the qualification figures it looks more like a deliberate trade: paying moderately more per lead to get a much better one. A lead cost 9.8% more, while a quotable lead cost 32.6% less. Whether that trade is worth making depends on what a quote is worth in your business, and your own numbers answer that better than an index can.

The typical advertiser in this index put 39.2% more into design-build search than a year earlier and saw leads per advertiser rise 27.3%. Leads grew more slowly than budget, which is the cost per lead telling the same story from the other direction.

Where the money goes

Where the money went moved sharply. Performance Max took 40.3% of design-build spend in 2026, up from 27.6%, while Search fell from 71.6% to 59.7%. Search’s share of leads went the other way, down from 67.4% to 61.8%. A larger share of the budget returning a smaller share of the leads.

That’s the clearest explanation of the rising cost per lead, and it needs separating from the market. A Performance Max lead cost $89.77 in 2026 against $51.42 a year earlier. Search moved in the opposite direction, $215.10 against $277.22. So Search remains much the more expensive of the two per lead, and it got cheaper. Performance Max remains much the cheaper, and it got more expensive. The index-wide figure rose partly because budget shifted toward the mode that was losing ground fastest.

None of that makes Performance Max the wrong choice. It still costs well under half as much per lead as Search. It does mean a shift in that direction bought less in 2026 than the same shift would have bought in 2025, and that a blended cost-per-lead figure hides which of the two moved. If your own mix shifted this way, the per-mode numbers are the ones to check.

Cost is only half the comparison, and the quality half runs the other way. In the first half of 2026, 53.9% of design-build leads from Search reached quotable status, against 46.3% from Performance Max. The value behind them separates further. Search returned $2,626 of quoted work per lead against $1,701 from Performance Max, and $4,870 against $3,674 once you narrow to the leads that were actually quoted. Search leads are likelier to be worth quoting, and the quotes are larger when they happen.

Set that against the cost and the gap closes a long way without closing entirely. Search costs about two and a half times as much per lead, $215.10 against $89.77, and returns about one and a half times the quoted value per lead. Per dollar spent, Performance Max still comes out ahead on quoted value, by roughly half again. But anyone treating cost per lead as the scoreboard is overstating that advantage by a wide margin, and a mix that has moved this much budget toward the cheaper mode has moved lead quality with it.

All comparable metrics

First half 2026 compared with first half 2025, across the residential landscape campaigns in the index.

What the numbers say

Headline findings

Maintenance search got cheaper to buy and better at qualifying at the same time.

A click cost $2.86 in the first half of 2026 against $3.09 a year earlier, down 7.5%. A lead cost $117.08 against $141.93, down 17.5%. The sharper move came further down the funnel: a quotable lead cost $211.12 against $363.16 the year before, a fall of 41.9%.

That came from qualification rather than volume. Just over half of maintenance leads reached quotable status in 2026, at 55.5% against 39.1% a year earlier. More than half of what now comes in is worth a quote, where before it was well under half. If your own lead-to-quotable ratio sits nearer 39% than 55%, that gap is the most actionable number in this report.

One measure moved the other way. Click-through fell from 3.98% to 2.69%, so fewer of the people who saw an ad clicked one. But more of those who did went on to become leads, with conversion rate rising from 2.17% to 2.44%, and more of those leads were quotable. The reach was narrower, but the traffic was better matched.

That fall is steeper here than in design-build, and the likeliest reason is visible in the same dataset. Performance Max took 36.3% of maintenance spend in 2026 against 20.3% a year earlier, a swing of 16 points toward a buying mode that places ads across inventory where low click-through is normal. A blended click-through covering both modes falls on that shift by itself, without anything changing in how people use a results page.

The AI-answers explanation from the design-build section applies here too, and this index can’t separate it from the mix shift for the same reason.

Cost and qualification

The cost picture is consistent across the funnel, which is what makes it worth trusting. Cheaper clicks, cheaper leads and much cheaper quotable leads all moved together, and the monthly view shows the qualification gain building through the half rather than arriving in a single month.

The typical advertiser in this index spent 13.3% more on maintenance search than a year earlier and saw leads per advertiser rise 38.1%. Spending moderately more and getting substantially more back is the pattern underneath every cost figure above.

A note on reading all of this: the cost and ratio figures (cost per lead, cost per quotable lead, lead-to-quotable, click-through, conversion rate) are computed within each period and are directly comparable year over year, so they’re the numbers to measure yourself against.

Where the money goes

Where the money went changed more than how much of it there was. Performance Max took 36.3% of maintenance spend in 2026, up from 20.3%, while Search fell from 79.7% to 63.7%. On leads the gap is wider still: Performance Max produced 60.1% of leads from 36.3% of spend.

That changes how the falling cost per lead should be read, because two separate things happened at once. Performance Max leads cost $70.73 against Search’s $186.91, so moving budget toward Performance Max pulls the blended figure down by itself, without anything actually improving. But Search also got cheaper on its own terms, $186.91 against $253.75 a year earlier, so the improvement is real as well as reallocated.

The counterweight is that Performance Max got more expensive per lead, $70.73 against $52.41. It remains much the cheaper of the two, but the gap narrowed even as its share grew. If you are weighing a shift in the same direction, that’s the trend to track, and to set against what Performance Max gives up, which is the keyword-level control Search keeps.

On lead quality, maintenance splits differently from design-build. Here the two buying modes qualify at almost the same rate: 56.4% of Performance Max leads reached quotable status in the first half of 2026, against 54.0% from Search. If anything that’s a slight edge to the automated mode, and it’s close enough that it should not be read as a gap at all.

The value behind those leads is where they part company. A Search lead carried $2,015 of quoted work against $1,319 from Performance Max, and $3,732 against $2,338 among the leads that were actually quoted. So Performance Max qualifies just as readily and quotes about two thirds as large, which points to the same kind of job at a smaller size. That fits what maintenance work is, a recurring and fairly standard service where volume matters more than ticket size, and it’s the opposite of the design-build picture, where Search both qualifies better and quotes larger.

Per dollar spent, that leaves Performance Max well ahead on maintenance. It costs well under half as much per lead, qualifies marginally better, and quotes about two thirds as large, so the cheaper lead more than covers the smaller quote. If maintenance is the part of the business you are trying to fill, the numbers back the shift that already happened.

All comparable metrics

What Local Service Ads cost per lead against Google Ads, for the same advertisers over the same period. First half 2026 compared with first half 2025.

What a lead costs

Local Service Ads cost less per lead than Google Ads, and the gap is remarkably stable.

In the first half of 2026 an LSA lead cost $37.02 against $60.61 from Google Ads, putting LSA at 61% of the Google Ads price. A year earlier it was $41.65 against $65.32, or 64%. Across two periods and a changing set of advertisers, the relationship barely moved.

That stability is what makes it worth acting on. Both figures come from the same advertisers in the same period, so nothing here depends on one group of businesses being compared against another. It’s the one comparison in this index that needs no caveat about who is in it.

The picture changes once qualification is taken into account. Google Ads turns a higher share of its leads into quotable ones, 50.7% against 44.4% in 2026, and 46.2% against 40.1% the year before. That’s a gap of roughly six points in Google Ads’ favor, in both years.

So the cost advantage narrows as you move down the funnel. Per lead, LSA runs at 61% of the Google Ads cost. Per quotable lead, it runs at 70%. LSA is still meaningfully cheaper, but about a third of the headline advantage is given back at the qualification step. Anyone judging the two channels on cost per lead alone is overstating the difference.

Through the season

There’s no clean seasonal pattern in what an LSA lead costs.

Across the first half of 2026 the monthly cost per lead moved between $30.32 and $43.11 without settling into a direction. The cheapest month was March, the dearest May. The year before ran differently again, starting at $57.28 in January and falling through the spring to $35.37 in May. If there’s a seasonal rhythm here, two half-years aren’t enough to show it.

Qualification did change. In 2026 the share of LSA leads reaching quotable status peaked in March at 55.3% and fell away through the summer to 38.3% by June. That’s the same shape the wider index shows, where earlier inquiries carry more intent. One year is a hint rather than a finding, but it lines up with what the lead data says elsewhere.

What the leads are worth

The cost advantage comes with a trade: Local Service Ads bring smaller jobs.

A note on sourcing first. The figures above come from the LSA dashboard. This section draws on the index’s lead-tracking data instead, which covers every channel on the same basis and carries the quote written against each inquiry. The two sets of percentages describe overlapping but not identical populations, so they should not be read against each other.

On that basis the average quote behind an LSA inquiry was $5,249, against $17,475 from paid search and $22,507 from Performance Max. That puts LSA jobs at somewhere between a quarter and a third the size. The ordering is the same in both years measured, which is what makes it something to act on rather than note.

An LSA lead costs 61% of a Google Ads lead and 70% of a Google Ads quotable lead, qualifies slightly less often, and carries a quote roughly a third the size. That’s a channel serving a different kind of demand: a job someone needs doing now rather than a project being planned. Neither is better in the abstract. Which is better for a given business depends on whether its capacity problem is filling the schedule or landing the larger build.

In practice, if you are trying to keep crews busy, the LSA cost per lead is doing real work for you. If you are trying to win design and construction projects, the cheaper lead is cheaper partly because it’s a smaller opportunity, and cost per lead alone will flatter it by roughly the factor the quote values differ by.

Two caveats on the dollar figures. Quote values are entered by hand, so these figures are computed only over the businesses that record them, and the ratio between channels is more reliable than the dollar level. They’re also drawn from the settled half rather than the most recent one, because quotes keep arriving for months after a period closes.