Spam and Bot Leads on Thumbtack: How to Spot Them Before You Pay

Not every bad lead is a real person who changed their mind. A share of what lands in your Thumbtack inbox was never a customer at all — spam, bots, and outright scams engineered to look like a genuine request just long enough to get you charged. The frustrating part is that you often pay before you can tell the difference. The good news is that fake leads leave fingerprints, and once you know what to look for, most of them are obvious in the first ten seconds. Here are the red flags that give spam leads away, and how Thumbtack's own spam detection works alongside the refund dispute you file yourself.

Real customers describe real problems. They mention the room, the timeline, the specific thing that broke, the neighborhood. Spam and bot leads read like a template — a bare service category with no specifics, no urgency, and nothing that ties the request to an actual home or job. If a "lead" could have been auto-generated by picking your category from a dropdown and nothing else, treat it with suspicion.

The tell is that a genuine request almost always contains a detail only a real person would include: "the disposal started leaking under the sink last night," "we're closing on the house in three weeks." A fake one stays vague because there is no real project behind it to describe. Vagueness alone is not proof, but combined with the other flags below it is a strong signal.

Unreachable numbers are one of the most reliable indicators of a fake lead. The number is disconnected, rings endlessly, goes straight to a generic voicemail that never gets returned, or simply is not a working line. A real customer who took the time to request a pro answers or calls back; a bot does not have a phone to answer.

A related pattern is worth watching: multiple "different" customers with nearly identical phone numbers — the same area code and prefix with only the last digits shuffled. Pros have reported clusters of fake leads whose numbers differ by a single digit. That is not a coincidence of local demand; it is a sign the leads came from the same source, not from separate households.

When you call the number back or search the name, nothing lines up. The name does not match the number's registered owner, the address does not exist or sits far outside any plausible service area, or the person has no digital footprint of any kind. A single mismatch can be innocent, but a request where the name, number, and location all fail to correspond to any real record is almost never a real job.

This is why the fastest fake-lead filter is a phone call. Calling immediately — which you should be doing anyway to win the lead on speed — doubles as your verification step. If the number is dead on the first attempt and the details do not check out, you have caught the fake before you have invested anything but a minute.

The clearest sign of coordinated spam is repetition. If two or three leads arrive with word-for-word identical phrasing but different names, you are not looking at three customers — you are looking at one script sent three times. Real requests vary because real people write them differently. Copy-pasted language across supposedly unrelated leads is a signature of bulk-generated fakes.

Timing reinforces this. Fake leads tend to arrive in clusters — several within the same hour or the same day, often after a long quiet stretch — rather than trickling in the way organic demand does. A sudden burst of similar-looking, similarly-worded requests that all go unreachable is a pattern worth reporting, not just refunding one at a time.

Thumbtack runs two separate systems for bad leads, and it helps to know which is which. The first is the Trust & Safety team, which proactively detects spam and fraudulent leads across the platform. When their systems flag a lead as spam, Thumbtack can refund it automatically — the amount goes back to your balance without you filing anything. This is the platform catching fakes at scale, and it is why you will occasionally see a credit appear that you never requested.

The catch is that automated detection is not perfect and does not catch everything in real time. Plenty of spam slips through to your inbox and gets charged before any system flags it. When that happens, the Trust & Safety net has failed you, and recovering the money moves to the second system — the one you drive yourself.

The second system is the case-by-case refund dispute. When a spam or bot lead gets charged and Trust & Safety did not auto-refund it, you request a refund yourself within 45 days of the charge — the same window that governs every other bad lead. Spam fits squarely under the "fake or low-intent request" refund reason, but this reason demands specifics, not a hunch. Support will not credit "I think this was a bot"; they credit a documented case.

So capture the evidence as you spot the flags: screenshot the vague or copy-pasted request text, the call log showing the unreachable number, and any cluster of near-identical leads side by side. Pros can send spam disputes to refunddisputes@thumbtack.com, and a request that shows the pattern — dead number, template wording, no real project — is far more likely to be approved than a vague complaint. Report obvious spam so Trust & Safety can catch it upstream, and dispute the ones that got through so you get your money back.

Fake leads are not random bad luck — they are a recognizable category with recognizable tells: generic requests with no real detail, phone numbers that go nowhere, names and addresses that match no record, and identical wording arriving in clusters. Spot those before or right after you pay, verify with a quick call, and screenshot the evidence the moment it is fresh. Thumbtack's Trust & Safety team catches some of this automatically, but the spam it misses is on you to dispute inside the 45-day window.

RefundMyLead is built for exactly the ones that slip through. It matches a spam or bot lead to the correct approved refund reason and writes the dispute in the structure Thumbtack support actually approves — so the fakes that beat the platform's own detection do not get to keep your money.