Why the Same Home-Repair Lead Gets Sold Five Times: The Economics of Angi, Thumbtack, and HomeAdvisor

What a shared lead actually is

A shared lead is the core product of a home-services lead marketplace. When a homeowner submits a job on Angi, Thumbtack, or HomeAdvisor, that contact is not handed to one contractor. It is distributed to several who match the trade and service area, each of whom is charged for the introduction. The homeowner has effectively become inventory sold multiple times over.

This is a deliberate design, not a defect. The marketplace maximizes revenue per inquiry by charging every contractor who receives it, and it maximizes the homeowner's odds of a fast response by putting several businesses in a race. The consequence for the contractor is that the price paid for a lead buys a chance at the job, not the job. If five businesses receive the same contact, the arithmetic is unforgiving: on average each wins it one time in five, but pays every time.

The title of this piece says five for a reason of illustration, not precision. The number of contractors a given lead reaches varies by trade, density, and the marketplace's own routing logic, and the platforms do not publish it. The point that survives the uncertainty is structural: on a shared-lead model, the cost you see per lead is not the cost per customer, because the lead was never exclusively yours.

The reported cost of a booked job

The number contractors care about is not the price of a lead but the price of a job that closes, the effective cost per booked job. Because a shared lead converts only a fraction of the time, the true acquisition cost is the lead price divided by the win rate, plus the cost of every lead that went nowhere.

Why the marketplace model produces these economics

The shared-lead marketplace sits between the homeowner and the contractor as an intermediary that charges a toll on demand it did not create the underlying need for. Its incentives point in a consistent direction. Selling each lead to more contractors raises revenue per inquiry while lowering the value of that lead to any one buyer, because the same contact now has more competitors attached to it.

This is a familiar pattern in platform economics. Wherever an aggregator inserts itself between local supply and local demand, from ride-hailing to food delivery to home-services leads, the intermediary captures a fee that the businesses on the platform experience as a tax on every transaction. The contractor rents access to demand that the marketplace controls, and the rent is charged again on the next lead and the next.

The countervailing force is disintermediation. When a business can be found directly by the buyer, through its own profile, its own reviews, and its own presence in the answer a search engine returns, the toll can be routed around. That is not a moral claim about marketplaces. It is an observation about where the cost sits and who controls the channel.

The complaint pattern: duplicate leads, fake leads, opaque pricing

Contractor dissatisfaction with the shared-lead model is documented in the aggregate. The Better Business Bureau profile for Angi Leads carries an average customer rating of about 1.96 out of 5, and the recurring themes in contractor complaints are consistent: leads that turn out to be duplicated or fake, contacts that never wanted the service, and pricing that is difficult to predict or dispute.

These complaints are self-selected and should be read as such, since satisfied contractors are less likely to file a review. But the pattern matters because it maps directly onto the economics above. A shared-lead system that charges per lead has a structural tolerance for low lead quality, because the revenue is booked at the point of distribution, not at the point where a job actually closes. The contractor absorbs the quality risk that the pricing model does not.

The owned-surface alternative: the local pack and AI answers

The alternative to renting demand is owning the surfaces where local buyers now decide. Two surfaces dominate: the local map pack that appears above the classic results, and the synthesized answer that engines increasingly return in place of a list.

The map pack captures a disproportionate share of local-intent attention. Aggregated Google local-search behavior studies report that searchers click the local three-pack roughly 44 percent of the time, against about 29 percent for organic links and 19 percent for paid, and that the top map-pack position takes around 17.8 percent of clicks. Businesses that appear in the pack have been reported to receive materially more traffic and more user actions, calls, direction requests, and site visits, than comparable businesses outside it. The direction of this finding is well established; the precise magnitudes are secondary-sourced and should be treated as indicative rather than exact.

The AI-answer surface is newer and behaves differently. A business that ranks in the classic local pack has less than even odds of also being named in an engine's AI local recommendation, and getting cited in a tool like ChatGPT has been reported as far harder to obtain than ranking in the map pack. Consumer use of these tools for local discovery is climbing quickly, with one annual survey reporting that 45 percent of consumers had used an AI tool to find a local business in the trailing year, up from 6 percent a year earlier. Both figures are single-vendor and single-survey, and are labeled emerging on that basis, but they point at a surface a contractor cannot afford to be invisible on.

Rented demand versus owned demand

The cleanest way to read the whole comparison is as two different kinds of spending. A marketplace lead is an operating expense that recurs. You pay for each contact, you pay again for the next, and when you stop paying the flow stops that day. Nothing you bought accumulates.

An owned local-pack and AI-answer presence behaves more like a capital asset. The work of making a business read as one consistent, well-reviewed, credential-accurate entity across its profile, its site, and the third-party sources engines trust is done once and compounds. It keeps returning buyers after the initial build, and it is not re-billed per lead. The relevant question for a contractor is therefore which channel is cheaper this month, and, just as important, which one they own at the end of the year.

This does not make marketplaces worthless. For a new business with no reputation and no presence, a shared-lead marketplace can supply demand immediately while the owned surfaces are still being built. The mistake is treating rented demand as a permanent strategy rather than a bridge, and never measuring what the booked-job cost actually is against the cost of owning the surface instead.

How to read these numbers honestly

The field is thick with confident figures that do not survive a check of their sources.

The shared-lead routing mechanism itself is well documented and not in dispute. The specific dollar figures, the roughly $1,400 cost per booked job and the four-to-five-times premium, rest on contractor anecdote and marketing-vendor comparisons rather than an audited dataset, and are labeled emerging throughout for that reason. The local-pack click and traffic magnitudes are established in direction but secondary-sourced in their exact numbers. The AI-answer visibility gap and the jump in consumer AI use come from a single vendor and a single survey. None of this makes the picture wrong. It makes the picture directional, and it means the only figure a contractor should fully trust is their own: the real cost of a booked job in their own market, measured against the cost of owning the surfaces where their buyers now search.