Buying leads or owning the pipeline: an honest look at the US lead marketplaces
If you run a service business in the US, you have almost certainly paid for a lead that three competitors also bought, or paid for one that turned out to be a wrong number. That experience produces two equally wrong conclusions: that the platforms are a scam, or that a website will replace them. Neither holds up. The marketplaces do something real, they charge for it, and the trade they ask you to make is specific enough to evaluate. Here is how the models differ, what number you should actually be tracking, and what has to be true on your own site before reducing that spend is a plan rather than a hope.
Three models that put the cost in three different places
Shared lead marketplaces are the model most contractors know. A homeowner describes a job, the platform matches it to several pros, and each of you pays for the same request. You are buying a chance to be first and cheapest, and the customer is being handed a comparison shop by design. Responsiveness feeds the matching, which is why these platforms train you to answer within minutes and why they can quietly set your calendar.
Google's Local Services Ads sit above the regular search results and also charge per lead, but the mechanics differ. Eligibility involves license and insurance verification, and the Google Guaranteed badge that follows is a genuine trust signal in trades where trust is the barrier. Leads can be disputed when they are clearly invalid, and ranking is influenced by proximity, responsiveness and review profile rather than by a bid alone. Then there is pay per click, including Google Ads and Yelp Ads, where you pay for the visit and carry all the conversion risk yourself. That last model punishes a weak website hardest, and rewards a strong one most.
- Shared leads: lowest commitment, highest price competition, you pay to be one of several
- Local Services Ads: verification required, badge included, per-lead billing with a dispute path
- Pay per click: you buy attention, not intent, and your own page decides whether it converts
- All three reward speed of response; none of them accumulate anything you keep
Cost per lead is the wrong number
The number the platforms report, and the number most owners quote, is cost per lead. It is close to meaningless on its own. A cheap shared lead you close one time in ten is worse than a costly verified one you close one time in three, and no dashboard will tell you which is which because the dashboard does not know what you booked.
Track cost per booked job by source, and then gross profit per booked job by source. That means tagging where every job came from at intake and keeping it attached through to the invoice, which is a CRM discipline more than a marketing one. Once you have a season of that data, the decision stops being philosophical. Some sources will be carrying the business, some will be funding your competitors' calendars, and a couple will be producing jobs at a margin you would decline if the customer asked for it directly.
- Tag the source at intake, on every call and form, without exception
- Carry the tag through to the closed job and the invoice amount
- Compute cost per booked job and gross profit per booked job, by source, by season
- Watch the close rate separately: a falling close rate on shared leads is the model working as designed
- Reconcile against actual platform invoices rather than dashboard estimates
What the platform keeps that you never see
The recurring cost is visible. The structural cost is not. The reviews customers leave live on the platform's profile and build the platform's inventory, not yours. The customer relationship is theirs, which is why the follow-up prompts they receive next season are not about you. The ranking rules can change without notice, the price per lead can rise, and a profile can be suspended for a policy dispute you did not see coming. You have no equity in any of it.
That is a real cost, but it is not automatically the wrong trade. A new business with no reviews, no site and no reputation has nothing to rank and nothing to convert with. Buying its way to a full calendar while it builds a track record is a completely defensible use of money, and it beats waiting six months for organic anything. The problem is not starting there. The problem is being there five years later with the same exposure and no asset to show for the spend.
What has to exist on your own site before you can spend less
Reducing lead spend safely means your own channel has to be capable of catching demand you already generate. Some of that demand already exists: people who see your van, get a referral, or find you on a platform and then search your company name before calling. Today, many of them land on a site that answers none of their questions and offers no easy way to reach you.
The build that changes this is unremarkable and specific. A real page for each service you want to sell, written to answer what someone is deciding rather than to list what you do. Honest handling of the pricing question, whether that is ranges, a worked example, or a clear explanation of what changes the number. Proof that is actually yours: your work, your license and insurance status, your warranty terms in your own words. Then the conversion mechanics: click-to-call in the header on a phone, a form that asks only what you need to price the job, and an intake that reaches whoever is answering within minutes rather than sitting in an inbox until morning.
- A substantial page per service, matched to how customers describe the job
- Pricing addressed openly, even if the honest answer is a range and the reasons it moves
- Your own project photos and proof, not stock imagery of somebody else's crew
- Click-to-call above the fold on mobile, and a form short enough to finish at a stoplight
- Submissions routed straight to the person who responds, with an automatic acknowledgement
- Clear service area, so you stop paying attention to jobs you would not drive to
The honest verdict, including when to keep paying
Keep buying leads when you are new to a market and have no reputation there, when you have crews idle and need volume this month, when seasonality leaves a gap you cannot fill any other way, or when nobody on your team has the bandwidth to run anything else. In all four cases the spend is buying time, and time is the thing you are short of.
The math flips when you are consistently paying more per booked job than the job's margin comfortably supports, when your close rate on shared leads is falling because you are always the fourth quote, or when the platform is your only channel and that dependency has started to make decisions for you. Even then, a website will not replace that spend in the first quarter. What it does is compound: the same page keeps working next year, the reviews you collect on your own profile keep counting, and the customers you capture stay yours. Anyone promising to replace your lead spend by spring is selling. Run both, measure both by booked jobs, and let the numbers move the budget.
Frequently asked questions
Are the leads on shared marketplaces really sold to several contractors?
On the shared model, yes, and that is the product rather than a flaw. The homeowner submits once and gets several responses. It means you are competing on speed and price from the first second, and it explains why close rates on those leads are structurally lower than on someone who called your number directly.
Are Local Services Ads better than Angi or Thumbtack?
Different, not universally better. Verification and the Google Guaranteed badge carry weight in trades where homeowners are worried about who they let in the door, and the placement above the search results is prime. It still charges per lead and still stops the day you stop paying. Test it against your own numbers instead of against opinions.
How long before a website reduces our lead spend?
Longer than anyone selling one will tell you, and the honest answer depends on your market's competitiveness and how much demand already searches for your name. The first effect is usually not new traffic at all: it is converting more of the people who were already checking you out before calling. That shows up quickly, and it is worth measuring on its own.
Should we stop buying leads once the site is live?
No. Cut based on measured cost per booked job, one source at a time, in a period you can compare against last year. Turning off the whole spend the week a site launches is how businesses end up with a quiet spring and a wrong conclusion about what the website was worth.
What is the single highest-return fix on our current site?
Usually response speed and the phone. Make the number tappable at the top on mobile, cut the form to the fields you actually need to quote, and route submissions to a human within minutes with an automatic acknowledgement to the customer. Most service businesses lose more jobs to a slow reply than to anything a designer could change.
If you are spending on leads and want to know whether it still makes sense, send us your sources and what a booked job is worth to you. We will tell you which spend we would leave alone, what your own site would have to do to earn a share of it, and quote only that. If the answer is that your marketing is fine and your response time is the problem, we will say so.