Is Thumbtack Worth It for Contractors? 2026 Math

Thumbtack is worth it for contractors whose average ticket clears roughly ten times their cost per booked job, which in practice means about $2,200 and up. Below that number the platform reliably loses money, and no amount of profile optimization fixes it. The variable that decides this is not your trade, your reviews, or your response time. It is arithmetic, and four rule changes that landed in 2026 moved the arithmetic against most pros.

That last part is why the guides ranking for this question are close to useless right now. Nearly all of them describe a platform that no longer exists: they explain surge pricing and the share-with-more-pros button, then recommend building your own marketing. Meanwhile Thumbtack rolled out proxy phone numbers, rewrote the non-response refund rule, and kept lead prices climbing. If you are evaluating Thumbtack on a 2024 description of Thumbtack, you are pricing the wrong product.

The four 2026 changes that actually decide the answer

Thumbtack still charges you for contact rather than outcome. That has not changed and it is not the interesting part. What changed is the machinery around that charge, in four specific ways that each move money.

1. Proxy phone numbers put a stranger's caller ID on your call

Thumbtack now routes calls and texts through its own numbers instead of handing you the customer's line. When you call the lead you just paid for, the homeowner's phone displays Thumbtack, not your business.

Pros report the obvious consequence. A pro of three years wrote that the caller ID means "the client doesn't know if it's someone they already spoke to or a new pro reaching out, which makes them not want to pick up." Another with 15 years in home services described tracking "a significant drop in conversion since the mandatory shift to blind, proxy-based calling," and framed the mechanism precisely: a homeowner expecting a local business sees an unfamiliar platform name and reads it as a robocall.

There is a second, more expensive consequence buried in it. The same pro noted that Thumbtack records calls and messages on those numbers, and that "if you try to speak to the customer through your direct phone number you are not eligible for a refund." So the workaround that restores your answer rate is the one that forfeits your money back. A 12-year veteran on the platform put the tradeoff plainly: "I prefer to do everything via text message that way I've got something to refer back to. Now any and everything you say is on Thumbtack for them to cruise through. It's confusing as hell costing me work."

2. The non-response refund fires far less often than it reads

Thumbtack added a refund path for customers who do not respond. Read the condition carefully: the refund triggers when the customer fails to respond to anyone within the window, not when they fail to respond to you.

That distinction is the entire game, because a lead is typically sold to several pros at once. A photographer described the exact failure mode after limiting himself to three leads: "One hadn't responded to any pro for the first 36h or so, which made me hopeful I'd at least get that money back. I didn't, because she ended up responding to another." Two of his three leads left him on read, he was charged for all three, and he estimated the round at $200 for nothing.

Another pro found the flag itself unreliable, reporting a customer who "has never responded to me yet I'm marked as has responded, and I'm the only one showing that." The refund reads like protection. Structurally it is a coupon that only pays out when a lead was worthless to everybody, which is a much rarer event than a lead being worthless to you.

3. Approved refunds are credit unless you chase the cash

When a refund is approved, the money lands as credit on your Thumbtack balance by default. Credit can only be spent on more leads, which means an approved refund recycles your money back into the thing that lost it. To get it onto the card you paid with, you have to request that within 10 business days of the approval.

You have 45 days from the charge to file in the first place, but treat that as a backstop rather than a plan. The evidence that wins these requests is perishable: call logs get overwritten, disconnected numbers get reconnected, and message threads scroll away. File in the same week.

One pro's account shows how the credit default plays out in practice. Charged $129.84 plus tax on a job with a stated budget under $500, they went back and forth by email for almost a week, refused to pay, and Thumbtack "instead credited my account to refund the cost." They got the number reversed and still never got the cash.

4. The homeowner selects the job size that sets your price

Lead price tracks estimated job value. The estimate comes from what the homeowner clicks, not from what the job turns out to be. A pro on the platform 13 years described being "charged for 1,000 square feet of flooring when the actual amount is 450 ft," which doubled the lead fee on a job half the stated size.

This runs in both directions, and homeowners are being told about it. A post aimed at customers on the platform's own subreddit spelled out the arbitrage: "The bigger the project that you show, the more the lead price is to even talk to you. Fees can range from $9 up to $129+ for EACH lead. It helps out to just choose the smallest project size." Some homeowners are now deliberately understating jobs to keep pro costs down, and others are overstating them by accident. Either way, the number that sets your cost is a field on a form filled in by someone with no stake in its accuracy.

What a lead really costs once you count the misses

The listed price is not the price. The useful number is cost per booked job, and pros converge on a rough multiplier for getting there.

One pro stated the working assumption directly: "When setting up my lead pricing, I now have to assume the actual cost is 4x higher than listed. I'm paying for 4 garbage, dead-end leads just to land one decent job." That is a 25% close rate, which is at the optimistic end of what shared-lead marketplaces produce. Run it at three rates against a mid-band $55 lead:

The right-hand column is the answer to the question in the title. If you hold marketing to 10% of job value, a $55 lead at a realistic close rate demands an average ticket in the low thousands. That is why the same platform produces wildly opposite verdicts from pros in the same city: it is not a disagreement about Thumbtack, it is a difference in average ticket.

That simultaneity is the hidden cost. A licensed California electrical contractor who has spent just under $4,300 on the platform over two years described being charged $55 for what looked like a simple switch replacement, with "almost no scope of work, and at least three contractors received the same lead." Thumbtack collected around $165 on one switch. Another pro was charged $91 on a lead where at least two other pros were also charged, for a project the customer said was not starting for another two months, putting the platform's take near $273 on a job nobody could book yet. Our breakdown of exclusive versus shared leads covers why that structure caps your odds before response time matters.

The refund policy names six reasons, and none of them is a bad lead

Most refund requests fail because pros file against unfairness rather than against a category. The policy refunds leads that did not match your profile settings, not leads that did not turn into work. Six named reasons exist: Location, Timing, Job type, Charged twice, Phone disconnected or out of service, and Reason not listed.

Four common situations sit outside all of them:

1. The customer stopped responding. Contact happened, so delivery happened.
2. The customer hired someone else. That is the model working as designed.
3. The customer's budget was nowhere near your minimum. Frustrating, not a category.
4. The customer changed their mind, postponed, or did it themselves. One pro had a large bid accepted, then the client "got impatient and decided to do the work themselves over the weekend." Thumbtack said it did not meet refund requirements.

Work the named five in order and stop at the first one that genuinely fits. They are checkable facts: an address is inside your radius or it is not, a number rings or it does not. The catch-all asks a human reviewer for a judgment call instead, which is a harder ask, and pros reach for it constantly when Job type or Location would have described the same problem in language the policy already recognizes.

One more trap worth knowing if you are on the AI Smart Leads program. A pro reported that Thumbtack support confirmed in writing that "the automatic refund is only triggered if you select Not what I do and complete this flow before replying to the customer." Reply first, and the automatic refund never fires. Ordering decides the money.

Who Thumbtack still works for

The honest case is stronger than the anger on the platform's forums suggests, and it fits the math above rather than contradicting it.

The California electrician who spent $4,300 over two years also landed, from one of those leads, "the biggest project my company has done. It kept my company busy for almost a year." One high-ticket win pays for two years of misses. That is the whole argument for the platform, and it only exists at large ticket sizes.

A piano-related service business reported spending nearly $10,000 in a strong stretch against roughly $120,000 in revenue, a 12:1 return, and judged it clearly worth it at the time. A photographer said Thumbtack "reliably gotten me a few jobs a month" and that several of those clients rehired repeatedly, so she has "made more than I've spent by a lot," while remaining angry about price increases. Repeat business is the real lever there: it converts a one-time acquisition cost into a customer, which is the only way a low-ticket business survives a per-lead toll.

One pro two weeks in made the strongest structural argument for staying. He noted that a $100 per week lead budget runs about $5,000 a year, against the $10,000 or more a contractor would spend building a website, running ads, and doing competitive SEO, and that Thumbtack does that marketing work for you. It is a fair point about year one. It is a bad point about year five, because the spend never converts into an asset you own.

So the profile that works is narrow but real: a high average ticket, or strong repeat and referral rates, plus the discipline to move the relationship off the platform after the first job. You paid for the introduction. Do not pay for the second job too.

Who it does not work for

Low-ticket, one-and-done work. If your average job is a few hundred dollars, the table above is unsurvivable at any close rate, and the platform is structurally a price-comparison venue. The complaint recurs almost verbatim: "I have my hourly rate clearly stated on my profile. Customers can see it before they even select me. Yet, when I reply with that exact standard price, they hit me with that's too expensive."

Premium operators have a related problem. Being one of three quotes on a scope-free request is a race you win on price or not at all. If your positioning depends on being explained rather than compared, the format works against you. We wrote separately about how to stop competing on price, and the short version is that you cannot do it inside a comparison engine.

Anyone whose business depends on the account is exposed in a way worth naming. Multiple pros describe sudden deactivation with no warning, one mid-conversation with a customer, and the proxy number system means deactivation cuts you off from leads you already paid for. Chargebacks accelerate this. One pro who had "paid Thumbtack thousands" and disputed about $250 won 7 out of 7 with their card issuer and was removed from the platform. The disputes work. The account does not survive them, and neither does the review history you spent years accumulating.

A 30-day test that produces a real number

Do not decide this from forum sentiment in either direction. Thirty days of your own data settles it.

1. Set a fixed budget you can afford to lose. One pro capped himself at three leads to bound the experiment. Do the same in dollars.
2. Turn off automatic leads and browse manually. Review each opportunity and its price before committing, rather than letting the algorithm spend for you.
3. Log every charge in a sheet: date, listed price, job type, stated job size, number of other pros if visible, whether you reached a human.
4. Record response outcome within 48 hours, because that is the window where your evidence for a refund is still capturable.
5. File every legitimate refund the same week, under a named category, with a screenshot attached. Then set a calendar reminder for 10 business days after any approval to convert the credit to cash.
6. At day 30, divide total spend by jobs booked. That is your real cost per booked job. Divide your average ticket by it. If the result is under 10, the channel is not working for your business.

The honest summary

Thumbtack is not a scam and it is not a pipeline. It is a spot market for introductions, priced by a number the homeowner types, sold to several pros at once, with a refund policy that covers delivery errors rather than bad outcomes. In 2026 it added a proxy layer between you and the customer that costs you answer rate or refund eligibility, whichever you choose to give up.

At a high average ticket, with repeat business and the discipline to take relationships off-platform, it can pay for itself many times over on a single job. At a low ticket, on one-off work, it cannot, and the fix is not a better profile photo. It is a channel where the enquiry arrives in your own pipeline and stays there, so you pay for a customer once instead of renting access to them one job at a time. That is what lead generation you own is actually worth: not a lower cost per lead, but the end of paying a toll on every introduction for the life of the business.