If you left Thumbtack a few years ago and are considering going back, the platform you remember is not the one you would be joining.
The lead prices are higher, which everybody knows. The more consequential change is quieter: what triggers a charge, and who controls it.
The billing trigger moved
Under the old model you paid per quote sent. You saw a job, decided it was worth pursuing, and spent money on that decision. Your spend was a function of your choices.
The current model, which Thumbtack calls Leads pricing, charges you when you respond to a lead or when a lead responds to your quote.
That second half is the part worth sitting with. A homeowner replying to something you sent can generate a charge, which means a portion of your monthly spend is now determined by other people's behaviour rather than your own.
Layer surge pricing on top, where costs rise with demand in a market, and monthly spend becomes genuinely difficult to forecast. Reported lead prices now run roughly $8 to $200 or more depending on trade and market, with figures that used to sit at $15 to $30 commonly reaching $25 to $100 in competitive metros.
The share button
This one is underreported and it is worth knowing before you sign up.
Homeowners have an option to share their project with additional pros. When that gets clicked, the contractors it reaches are charged, reportedly $30 to $50 each.
So a single click by somebody who has not yet decided anything can bill multiple businesses simultaneously for the same job. Reports describe up to fifteen pros charged against one request.
Nobody in that chain did anything wrong. The homeowner was told sharing gets more quotes. But the mechanism means your costs can be driven by a customer exploring options.
The numbers that decide it
Leads go to somewhere between four and ten contractors. Roughly 78% of customers book whoever responds first, which turns the whole thing into a speed contest that favours whoever is not currently up a ladder.
Then the arithmetic. At $35 to $100 per lead and a 15% close rate, cost per acquired customer lands somewhere between $233 and $667. At a $75 lead and a 10% close rate, it is $750.
Compare that with Google Local Services Ads, benchmarked around $53 per lead at a 43.9% book rate, giving roughly $233 per paying customer on exclusive leads. Same neighbourhood on cost per lead, very different on cost per customer, because one is shared with nine other people and the other is not.
For high-ticket work at $5,000 and above, a $400 acquisition cost can still work comfortably. For repair work at a $200 to $400 ticket, it rarely does.
The complaint pattern
Public sentiment is poor and unusually consistent. Trustpilot sits around 2.5 out of 5 and ConsumerAffairs around 1.5. The BBB has logged more than a thousand complaints through January 2026.
The recurring theme is being charged for leads that were never real: contacts who confirmed they had not submitted a request, customers who vanished after first contact, and refund requests declined. One frequently cited figure puts the ghost rate at around 75%, meaning three quarters of paid contacts go silent.
A contractor quoted in one review describes paying $60 for a lead whose message said, in effect, not interested and not in our budget.
Worth the standing caveat: most published analysis of Thumbtack comes from companies selling alternatives, and one comparison I read was a competitor's own marketing page with itself in the winning column. The complaint volumes and the structural mechanics are verifiable. The framing around them is not neutral anywhere.
The promoted results problem
There is a second spend layer. Contractors can pay to appear higher in results, which creates an arms race where visibility requires outspending competitors on top of the base lead cost.
That is a familiar shape. Any marketplace where position is purchasable eventually prices position at whatever the most desperate participant will pay.
When it still works
Narrowly, and with conditions.
You answer immediately, because 78% going to the first responder is the entire game. If notifications arrive while you are under a sink, you are funding somebody else's conversion.
Your ticket is high enough to absorb a $250 to $650 acquisition cost. That means remodels, replacements, and large installs rather than service calls.
You cap it. Contractors who let a shared-lead platform become 80% of their spend get hurt whenever pricing, the algorithm, or their visibility changes, and all three change without notice.
And you treat it as supply rather than as a pipeline. Nothing you build there transfers. The profile, the reviews, the ranking, the history all stay when you leave.
The honest read
Thumbtack in 2026 is a more expensive and less predictable version of what it was, with the billing trigger moved somewhere you do not control.
If you are considering it, run one calculation before anything else: your own close rate on shared leads, not your close rate on referrals, applied to the current lead price in your trade and market. That gives cost per booked job, and it is the only number that matters.
Then set a hard monthly cap, because the surge pricing and the share button between them mean the spend does not naturally limit itself.
And if you are on it now and it is working, keep going while building something you own alongside it. The contractors who get hurt by platforms like this are never the ones who used them. They are the ones who used only them.