Networx vs Hearth vs Thumbtack for Roofing Leads: What Each One Actually Sells You

Three names come up over and over when a roofing owner starts pricing out lead sources: Networx, Hearth, and Thumbtack. They get lumped together in the same sentence, the same spreadsheet, the same "where should I spend my marketing money" debate. The problem is that they are not the same kind of thing at all. One sells you contact records. One sells your customer a loan. One rents you a spot in a marketplace where you bid against everyone else in your zip. Comparing them head to head without understanding that is how a contractor burns three grand in a month and concludes that "buying leads doesn't work" when the real issue was buying the wrong product for the wrong reason.

I've watched roofers run all three, sometimes at the same time, and I've pulled the numbers afterward to see what actually closed. What follows is the practitioner version: what each platform is, how the mechanics really work, what it costs once you account for the leads that go nowhere, where each one fits (some of them do have a fit), and the math on the alternative that most owners skip because it requires a little more setup. I'm not going to tell you all three are garbage, because that's lazy and untrue. I'm going to tell you exactly what you're buying so you can stop guessing.

The one-paragraph answer, then the long version

If you want the short version before the 7,000 words: Networx is a lead aggregator that sells you shared homeowner contact info, usually sold to three to four contractors at once, priced per lead. Hearth is not a lead source at all in the traditional sense; it's a homeowner financing and contractor sales toolset that helps you close jobs you already have by offering monthly payments, and it markets some lead-gen features on top. Thumbtack is a service marketplace where homeowners post a request, you pay to send a quote, and you compete on profile, reviews, and response speed. None of the three gives you an owned, repeatable acquisition channel. They give you transactions. The contractors who win long-term build a list of homes they target on purpose, document the roof, and own the relationship from first touch. That's the part the three platforms structurally cannot sell you, and it's where the back half of this piece goes.

Now the long version, because the details are where the money is.

What you're actually comparing: three different businesses

Before the feature-by-feature, internalize the category each one lives in. This single distinction explains 90% of the confusion.

When a contractor says "I tried Networx vs Hearth vs Thumbtack and none of them worked," what usually happened is they expected all three to deliver the same outcome: a homeowner who wants a roof, on the phone, ready to book. Only one of the three (Networx, and arguably Thumbtack) is even trying to do that. Hearth is solving a completely different problem. Judging Hearth on "did it send me leads" is like judging a forklift on gas mileage.

Networx: the shared-lead aggregator

How it works

Networx runs websites and ad campaigns that capture homeowners looking for home-improvement work, including roofing. When a homeowner fills out a form, that record gets sold to a handful of contractors who service the area. You set a budget, define your service zips and project types, and leads get delivered to your phone and inbox as they come in. You're typically billed per lead, and the standard model is shared — the same homeowner is sold to multiple contractors, commonly three to four.

What it really costs (the honest math)

The sticker price per roofing lead from aggregators in this category generally lands somewhere in the $25 to $100+ range depending on project type, zip competitiveness, and whether it's a repair or a full replacement inquiry. Storm-driven full-replacement leads in a hot market cost more. But the per-lead price is not your cost of acquisition. Here's the real chain, with conservative numbers you can swap for your own:

• You buy 40 shared roofing leads in a month at $45 each = $1,800.
• Shared means 3 other contractors got the same homeowner. Realistically you connect with (actually reach a human on) maybe 50% = 20 conversations.
• Of those, the homeowner is still in-market and hasn't already signed with the contractor who called first: maybe 40% = 8 real opportunities.
• Of those 8, you set an inspection on 50% = 4 appointments.
• You close roofing appointments at, say, 30% = roughly 1.2 jobs.

So $1,800 bought you a little over one job. Your cost per acquired customer is around $1,500 on that cohort. Whether that's good or terrible depends entirely on your average job value and margin. On a $14,000 replacement at 35% gross margin, $1,500 CAC is workable. On a $600 repair, you just lost money to win the customer. This is the single most important thing about buying shared leads: the unit economics only survive on high-ticket replacement work, and only if your speed-to-lead and close rate are genuinely good.

Where Networx-style aggregators actually fit

• You have a dialer-ready sales process and someone who calls a new lead within five minutes, every time. Speed is everything on shared leads because you're racing three other companies.
• You're chasing replacement volume, not repairs.
• You can stomach variable lead quality and have a disciplined process for disputing junk leads (wrong number, out of area, renter, not actually roofing). Every aggregator has a credit/return policy — learn it and use it religiously, because un-disputed bad leads quietly wreck your cost-per-job.
• You want to fill gaps in a slow week rather than build your whole pipeline on it.

Where it goes wrong

The failure pattern is predictable: an owner turns on a lead budget, doesn't have a sub-five-minute call process, lets leads sit for an hour, talks to the homeowner who's already had two other roofers out, blames the source, and quits. The lead was fine. The response time wasn't. The second failure pattern is running aggregator leads for repair-sized tickets where the math can't work. The third is never disputing bad leads, so the effective cost per usable lead is 30-40% higher than the invoice suggests.

The dispute discipline most roofers skip

Every aggregator gives back credits for leads that meet their bad-lead criteria, but the window is short and the burden is on you. Treat it like a real process. Build a one-line log — date received, time of first call attempt, outcome, reason for dispute — and have whoever works the leads flag bad ones the same day. The legitimate reasons usually accepted are wrong or disconnected number, geography outside your service area, a renter who can't authorize work, a duplicate of a lead you already bought, and a project type you don't do. Reasons that almost never get credited: "they didn't answer," "they went with someone else," "they weren't ready to buy." Those are sales problems, not lead-quality problems, and trying to dispute them just burns goodwill with the rep and gets you nowhere. A shop that disputes diligently often recovers 15-25% of spend in credits over a season; a shop that never disputes pays the full sticker on leads that were never workable. That gap alone can flip an aggregator from unprofitable to profitable.

What a workable aggregator process looks like, step by step

1. Lead lands and triggers an instant push notification to a phone, rather than an email that gets read at lunch.
2. The assigned closer calls within five minutes. If no answer, leave a specific voicemail ("This is Dana with [company], you asked about a roof estimate — I can come by tomorrow morning or afternoon, which is better?") and immediately send a text. Voicemail alone gets ignored; the text is what gets the callback.
3. Second attempt within the hour, third attempt same day, then a tapering cadence over the next week. Most contractors quit after one try; the homeowner who looked dead on Monday often answers Thursday.
4. Log every touch against the lead with its source tag intact.
5. Same-day dispute of any lead that meets the credit criteria.
6. Anything that reaches a human and stays in-market moves into your real pipeline and gets the standard appointment-setting workflow.

That process is the difference between aggregator leads being a slow bleed and being a controllable, marginally profitable supplement. The leads are rarely the problem. The operating discipline around them is.

Hearth: not a lead source, a closing tool

How it works

Hearth is fundamentally a homeowner financing platform plus contractor sales tools. The core value is letting you present a roofing project as a monthly payment instead of a $14,000 lump sum, by pre-qualifying the homeowner through Hearth's network of lending partners without affecting their credit to check rates. Around that, Hearth has built sales features: digital quotes, e-signature, invoicing, payment collection, review requests, and some lead-management functionality. It's sold as a subscription.

Why people put it on this list at all

Because Hearth markets itself adjacent to lead generation and because financing genuinely increases close rate and average ticket on the leads you already have. That's a real, measurable effect. When a homeowner is on the fence about a full replacement, "$210 a month" lands very differently than "$14,000 today." Financing also lets a salesperson present good-better-best options without the top tier scaring everyone off. So in the funnel, Hearth doesn't add leads at the top — it widens the conversion in the middle and bottom.

The honest read for a roofer

Don't buy Hearth expecting your phone to ring. Buy it (or any financing/sales tool like it) if:

• A meaningful share of your lost deals die on price/affordability, not on trust or scope.
• You sell replacement and want to lift average ticket by presenting upgrades as a payment delta.
• You want to collect payment and signatures digitally instead of chasing paper.

And understand the tradeoffs. Financing has dealer fees — the lender's cost of offering the loan is generally borne by the contractor as a percentage of the financed amount, which comes straight out of your margin. You have to price that in. A 6-7% dealer fee on a $14,000 job is roughly $900-$1,000 you don't see. If financing converts a deal you'd otherwise have lost, that fee is the cheapest customer you'll ever buy. If you're offering financing to people who'd have paid cash anyway, you're donating margin. Discipline matters.

So: Hearth and Networx aren't competitors. They're complements. One could fill the top of the funnel; the other tightens the bottom. Comparing them as if you must choose is the wrong frame.

A worked example of what financing actually does to a deal

Walk through a real close. A homeowner has a 19-year-old three-tab roof, two leaks, and a quote at $13,500 for a full replacement in architectural shingles. Cash-only, the conversation stalls at "let me think about it," and "think about it" closes at maybe 25-30%. Now reframe it as financing. Pre-qualify them with a soft pull, and the same $13,500 becomes roughly $190-$230 a month depending on term and rate. The objection shifts from "I don't have $13,500" to "is $210 a month worth not worrying about the roof," which is a far easier yes — close rates on financed presentations commonly run meaningfully higher than cash-only.

Financing also lets you sell up, beyond simply closing the base job. Present three tiers: the base replacement, a mid tier with an upgraded shingle and better underlayment, and a top tier with a full ventilation correction and a longer workmanship warranty. As a lump sum the top tier is scary. As a payment delta — "the better package is about $35 more a month" — a real share of homeowners take the upgrade, and your average ticket and margin dollars per job rise. The dealer fee is real and comes out of margin, but on incremental deals you'd have lost and on tickets you'd never have lifted, it pays for itself. The discipline, again: don't finance the buyer who was going to write a check anyway. Read the room.

What financing does not do

It does not fill your calendar. It does not bring you a single homeowner who wasn't already in front of you. If your problem is "not enough at-bats," Hearth is the wrong purchase — you need a top-of-funnel source, owned or rented. If your problem is "good at-bats that die on price," financing is one of the highest-ROI tools in roofing. Diagnose which problem you actually have before you spend.

Thumbtack: the marketplace auction

How it works

Thumbtack is a marketplace where homeowners describe a job — "replace asphalt shingle roof, single-story, ~1,800 sq ft" — and the platform surfaces local pros. You set your services and preferences, and when a matching request comes in, you can send a quote. You pay per lead, and the price is dynamic: it's set by an auction-like system that factors demand, competition in your category and area, and how many pros are competing for that job. Higher-value jobs and more competitive markets cost more per lead. Reviews, profile completeness, and response speed drive how often you get surfaced and chosen.

What it really costs

Thumbtack lead prices for roofing swing widely — a small repair inquiry might be a few dollars, while a full-replacement lead in a competitive metro can run $30 to $90+ per lead, and you can be charged when the customer responds or when multiple pros are competing. Like Networx, a single homeowner is usually talking to several pros, so it carries the same shared-lead dynamics: speed and reviews decide who wins. The platform has changed its pricing model more than once, which is its own risk — your cost structure can shift out from under you.

Where Thumbtack fits

• Repairs and smaller jobs, especially for newer or smaller roofing operations building a review base. Thumbtack's review engine is genuinely useful early on; a wall of strong reviews compounds.
• Markets and seasons where you want incremental volume and can babysit response time.
• Operators who will respond within minutes, because the marketplace rewards it explicitly.

Where it goes wrong

Same core issue as Networx — shared demand, so speed-to-lead is everything — plus marketplace-specific risks: pricing model changes, charges for leads that never respond, and a fundamental ceiling: the homeowner is Thumbtack's customer, not yours. You don't get a durable channel. You get transactions, and the platform keeps the relationship and the data. The moment you stop paying, the faucet stops. There's no compounding asset.

How to run Thumbtack without bleeding

If you do use it, run it tight. Lock your service preferences down hard — exact job types, exact travel radius, project sizes you actually want — so you stop paying for leads you'd never take. A complete profile with real project photos, a clear license/insurance display, and a steady stream of recent reviews changes both how often you're surfaced and how often the homeowner picks you over the three other quotes. Reply within minutes, with a real message, not a templated blast; the marketplace and the homeowner both reward it. And set a weekly spend cap so a sudden flood of low-quality requests can't drain a month's budget in three days. Watched closely, Thumbtack can be a fine early-stage repair-and-reviews engine. Left on autopilot, it quietly charges you for the privilege of competing.

The reviews asset is the one thing here that compounds

Worth calling out: of everything Thumbtack gives you, the review base is the only piece that keeps paying after you stop. Those reviews build your reputation, and a strong review profile lifts conversion on every other channel you run — your website, your Google profile, your mail. So if you use Thumbtack early, treat review generation as the actual goal, not the jobs themselves. Ask every satisfied customer, make it one tap, and migrate the momentum to a review profile you control (your Google Business listing) so the asset isn't trapped inside one marketplace.

Side by side, with the parts the sales pages skip

The table makes the real conclusion obvious: these three are not a menu where you pick one. A coherent operation might run Thumbtack early for reviews and repairs, layer in an aggregator like Networx when it has the speed discipline to win shared leads, and offer financing (Hearth or a competitor) at the point of sale to lift close rate. They sit at different points in the funnel.

But all three share one structural flaw, and it's the most important point: none of them is an owned channel. Every one of them is rented. You're paying per transaction or per month, and the day you stop, you go back to zero. There's a fourth way to get roofing work, and it's the one that compounds.

Shared vs exclusive: the comparison that actually decides profitability

The shared-versus-exclusive question matters more than which brand name is on the invoice, and most owners underweight it. A shared lead is sold to multiple contractors — three to four is typical for Networx and effectively true on Thumbtack. An exclusive lead is sold to you alone. Exclusive costs more per lead, often two to four times more, and the instinct is to balk at the price. That instinct is usually wrong, and here's the math that shows why.

Take the shared cohort from earlier: 40 leads at $45, ~50% reach rate, ~40% still in-market because three competitors are racing you, working out to roughly 8 real opportunities and about 1.2 jobs for $1,800 — a cost per won job near $1,500. Now take exclusive leads at $150 each. You buy 12 for the same $1,800. Because no one else is calling, your reach rate climbs (say 70%) and your in-market rate climbs hard (say 75%) because the homeowner hasn't already signed with the fast contractor. That's roughly 6 opportunities from 12 leads, 4-5 appointments, and at a 30-35% close, about 1.5 jobs — for the same money. Same budget, fewer leads, more jobs, and far less wasted phone time chasing people who already hired someone.

The numbers are illustrative — plug in your own reach and close rates — but the shape holds across markets: exclusive usually wins on cost per won job and almost always wins on wasted labor, even though it loses on cost per lead. This is exactly why comparing Networx, Hearth, and Thumbtack on sticker price is the wrong analysis, and why an owned outreach channel — where every lead is exclusive to you by construction — tends to come out ahead once you measure jobs instead of leads.

The thing none of these three sell: an owned, ranked acquisition channel

Here's the mindset shift. Networx, Hearth, and Thumbtack all start from inbound — a homeowner who already decided to look for a roofer raised their hand, and you're paying to be in the room. That's fine, but it's the most expensive and most competitive moment to show up, because by definition everyone else is in that room too.

The contractors who get off the lead-buying treadmill flip it to outbound on purpose: instead of paying for whoever happened to fill out a form, they decide which homes to go after, based on the two things that actually predict a roofing job — how old the roof is and whether it's been hit by storms — and they reach those homeowners before the homeowner has started shopping. No three-way share. No auction. No platform that owns the relationship. You own the list, the contact, and the data.

That's the part worth building, and it's specifically where RoofPredict fits.

Targeting: rank the homes instead of buying the leftovers

RoofPredict scores every home in a service area by roof-age band — recent, mid-life, due, overdue — combined with per-roof storm exposure and an overall opportunity score, and produces a ranked target audience: a house-by-house list of which roofs are most likely due for replacement, each with a "why this home" evidence chain so your team isn't knocking blind. You can import addresses by CSV, draw a territory on a hex map, and filter by storm hits so you're working the streets that actually took weather.

Be clear-eyed about what that scoring is and isn't, because honesty is the whole point. Roof age comes back as a range, not an exact install date — it's a band, not a birth certificate. Storm exposure is a probability of impact, odds, not proof that a given roof is damaged. This is heuristic targeting — roof-age and storm-exposure signals stacked into a priority order — not a magic prediction that a specific roof will leak Tuesday. What it does is let you spend your knocking, mailing, and calling energy on the overdue, storm-exposed homes first instead of spraying a whole zip. That's the difference between a 2% response and a 6% one.

From ranked list to actual contact: mail, microsites, QR, and field routes

A ranked list only matters if you do something with it. RoofPredict turns the due-roof list into work:

• Tracked direct mail. It turns the list into a mail campaign with personalized mail proofs (brand, copy, and address checks before anything prints), vendor release, per-piece delivery and return tracking, and a cost quote up front so you know your spend before you commit. You're mailing the overdue, storm-hit homes — not the whole carrier route.
• Personalized microsites + reports + QR. Every targeted home gets its own personalized report (roof profile, storm history, risk, and a cost-of-waiting framing) as a PDF and a public microsite with a lead-capture form. Each home and mail piece gets its own QR code, so a homeowner who scans from a postcard or a door hanger lands on their roof's page, not a generic site. That's a far warmer first touch than "licensed local roofer, call now."
• Canvassing and field routes. Build door-knock routes from the ranked list, assign canvassers, and run a mobile field app with next-stop navigation, outcome forms, voice notes, and leave-behind QR codes, with live route progress so a sales manager sees the territory get covered in real time.

Compare the unit economics honestly against the shared-lead model. With Networx you paid ~$1,500 to net one job out of 40 shared leads you didn't choose. With a ranked mail-plus-canvass push, you pick the 800 most-overdue, storm-exposed homes in a territory, mail them at a known cost, track every piece, and the leads that come back are exclusively yours and self-selected by scanning their own roof report. The response rate is lower per-contact than an inbound form — these people weren't shopping yet — but the leads are exclusive, the cost is fixed and visible, and the channel compounds: the list, the data, and the brand presence in that neighborhood are yours next season too.

A worked example of an owned-channel campaign

Make it concrete with numbers you can sanity-check against your own market. You draw a territory of newer-to-mid neighborhoods, filter to the overdue and due roof bands that also took a recent storm, and land on 800 homes. You mail a personalized postcard to each — the homeowner's own roof age band, the storm date their street saw, a QR code to their roof's microsite, and a cost-of-waiting line. At a fully-loaded mail cost in the neighborhood of $0.75-$1.10 per piece (design, print, postage), call it $800. Direct mail to a cold list typically responds in the low single digits; to a targeted, personalized, storm-relevant list with a roof-specific microsite, you should beat generic mail. Say a 3% response on 800 = 24 inbound contacts that are exclusive to you, plus QR scans you can see in the funnel even from people who don't fill the form yet (those are retargeting and door-knock fuel). Of 24 exclusive, warm-ish contacts, you set maybe 12 inspections and close 30-40%, so 4-5 jobs from $800 of mail. That's a cost per won job in the low hundreds — and you still own the list to knock, re-mail, and work next season.

The honest caveat: this channel is slower to spin up than flipping on a lead budget, and the first mail drop teaches you more than it sells. You refine the territory, the offer, and the creative over two or three drops before it hums. But once it hums, it's yours, it's exclusive, and it doesn't reset to zero when you stop paying a marketplace.

Don't lose the leads you paid for: the pipeline and CRM problem

Whichever sources you run, the second-biggest money leak after slow callbacks is leads dying in the gap between the source and the CRM. A Networx lead hits an inbox, a Thumbtack message sits in an app, a door-knock outcome lives on a canvasser's phone, and three weeks later nobody remembers to follow up on the homeowner who said "call me after the holidays."

This is the other place RoofPredict earns its keep. It runs a real lead pipeline — new, contacting, appointment, inspected, won/lost — with an immutable first-touch source stamped on every lead, so when you finally do the math on Networx vs Hearth vs Thumbtack vs your own mail, you can actually see which source produced which won jobs instead of guessing. And it does two-way sync with 13 CRMs, including HubSpot, ServiceTitan, JobNimbus, AccuLynx, Jobber, Housecall Pro, Salesforce, Pipedrive, Leap, Roofr, SalesRabbit, and CompanyCam (plus Zapier and CSV), so the system you already run stays the source of truth and nothing falls through the gap between a marketplace and your back office.

That immutable first-touch source is what makes the next section possible, because you cannot compare lead sources you don't measure.

Actually measuring your sources: cost per lead is the wrong number

Most owners compare Networx, Hearth, and Thumbtack on cost per lead, which is close to useless because a $25 shared lead and a $25 exclusive lead are not the same product. The number that matters is cost per won job by source, and almost nobody tracks it cleanly because the data is scattered across three platforms and a CRM.

RoofPredict's results funnel is built for exactly this: delivered → views → form → calls → leads → wins, with cost-per-lead and cost-per-win computed per source, plus actual vs estimate vs industry benchmark so you can see whether a channel is beating or trailing what's normal, and A/B campaign variants so you can test two mail pieces or two offers against each other honestly. Run that for a quarter and the Networx-vs-Thumbtack debate stops being a vibe and becomes a number.

Here's a simple worked framework you can build in a spreadsheet today even before any tooling, so you're at least measuring the right thing:

1. Tag every lead with its true source at first touch and never overwrite it.
2. Track each lead to a terminal status: won, lost, or dead.
3. For each source, compute total spend ÷ jobs won = cost per won job. (Include the platform fee AND the labor cost of working those leads — a shared lead that takes five callbacks costs more than the invoice.)
4. Compute average gross margin per won job by source — replacement-heavy sources earn their higher CAC; repair-heavy ones may not.
5. Kill or shrink any source where cost per won job approaches gross profit per job. Grow the ones with the widest gap.

When you do this honestly, the usual finding is that the cheap-per-lead sources are expensive per job, and the channel everyone ignored because it "takes setup" — owned, targeted outreach — wins on cost per won job even though it looks slower at the top.

What experienced roofers get wrong about all three

A few patterns show up again and again, even among shops that have been buying leads for years:

• Confusing activity with progress. A full inbox of shared leads feels productive. If your reach rate is 40% and you're not disputing junk, that inbox is mostly cost. Measure jobs, not lead volume.
• No source attribution, so every comparison is a guess. If you can't say which source produced which won job, you're allocating budget on feel. Stamp first-touch source on every lead and never overwrite it.
• Running every source at full tilt at once. You can't read a test where three variables move together. Cap and stagger.
• Blaming the source for a sales-process problem. Slow callbacks, weak voicemails, no follow-up cadence — those kill leads from every source equally. Fix the process before you change the source.
• Ignoring the channel that compounds because it's slower to start. Owned outreach takes two or three drops to tune, so impatient owners abandon it after one and stay on the rented treadmill forever.
• Treating financing as a lead source. It lifts conversion; it does not fill the funnel. Buy it for the right reason.

A 30-day plan to figure out your own answer

Stop reading reviews and run a controlled test. Here's a month that gives you real data instead of opinions.

Week 1 — Instrument first. Set up source tagging in whatever CRM you run (or a clean spreadsheet). Write down your current numbers: average job value, gross margin, close rate, speed-to-lead. You can't improve what you won't measure.

Week 2 — Fix speed-to-lead before you spend a dollar on shared leads. Whoever buys Networx or Thumbtack leads without a sub-five-minute callback process is lighting money on fire. Build the process: instant notification, a dialer, a script, and a person whose job is to call now. Test it on your existing inbound.

Week 3 — Run small, parallel tests. Put a capped budget on one shared source (Networx or Thumbtack, not both at full tilt). Separately, stand up one owned test: pick a territory, pull the most overdue and storm-exposed homes, and mail or knock 200-400 of them with a personalized roof report and a QR code to their own microsite. If you sell replacement and lose deals on price, turn on financing as the closing tool. Keep every source tagged.

Week 4 — Read the cost-per-won-job, not the cost-per-lead. Compute it for each source per the framework above. Be ruthless. The winner is whatever produces won jobs at the lowest fully-loaded cost relative to the margin on those jobs — and remember the owned channel keeps paying off after the test while the rented ones reset to zero the day you stop.

After 30 days you'll have your own answer to "Networx vs Hearth vs Thumbtack" that's worth more than anyone's blog ranking, including this one — because it's your market, your team's speed, your job mix, and your margins.

A note on storm and "claims" language, because it'll come up

The second you start targeting storm-exposed roofs, the temptation is to lean into insurance-claim messaging in your mail and door pitch — and there's a hard line you do not cross. As a roofing contractor you may inspect a roof, document damage thoroughly, and prepare an accurate repair estimate for your own scope of work, and you may state facts about your scope to the carrier. You may not, for a fee, negotiate or "handle" the claim, interpret the homeowner's policy or coverage, promise a specific payout or approval, promise that a deductible is waived or absorbed, or advertise a "free roof." Doing any of that is functioning as an unlicensed public adjuster, and it's the fastest way to put your license at risk and your mail in front of a state regulator.

The compliant frame is simple: you document, you write an accurate, Xactimate-aligned estimate for the work, and you hand it to the homeowner. The homeowner files the claim and the insurer decides coverage. Your storm targeting tells you which roofs likely qualify by age and storm exposure and drives a thorough photo-and-scope documentation workflow — it never represents the homeowner against their insurer. Keep the do-not-say list above on the wall by the sales team's desk. It protects the business that all this lead spend is meant to grow.

So which one should a roofer pick?

If you force me to a one-line recommendation per situation:

• Newer/smaller shop building reputation and doing repairs: start with Thumbtack to stack reviews and catch small jobs, with a brutal response-time discipline. Don't over-spend.
• Established shop with a real sales team chasing replacement volume: test a capped Networx (or comparable aggregator) budget only after speed-to-lead is fixed, and dispute bad leads relentlessly.
• Any shop losing replacement deals on affordability: add Hearth (or another financing tool) as a closing layer — it's not a lead source, it's a conversion lift, and it pairs with either of the above.
• Any shop that wants to stop renting its pipeline: build the owned, ranked outreach channel underneath all of it — target by roof age and storm exposure, mail and knock the overdue homes with personalized reports and QR microsites, capture every lead with an immutable source in a real pipeline synced to your CRM, and measure cost-per-won-job in a results funnel. That's the channel that compounds and the one the marketplaces structurally can't sell you.

The marketplaces give you transactions. An owned channel gives you a customer-acquisition machine you control. RoofPredict is built to be that machine — it ranks the homes that are actually due, turns that list into tracked mail, microsites, QR codes, and field routes, captures every lead with a permanent first-touch source synced two-way to the CRM you already use, and shows you cost-per-win by source so the next time someone asks "Networx vs Hearth vs Thumbtack," you answer with your own numbers. Start by mapping one territory's overdue, storm-exposed roofs and mailing the top of that list — then compare what comes back to whatever the marketplaces sent you. The math tends to settle the argument.

FAQ

Are Networx, Hearth, and Thumbtack the same kind of lead source?

No, and treating them the same is the most common mistake. Networx is a shared-lead aggregator that sells you homeowner contact records, usually to several contractors at once. Thumbtack is a service marketplace where you pay per quote to compete on a homeowner's posted job. Hearth is not a lead source at all in the usual sense; it's a homeowner-financing and contractor-sales toolset that helps you close jobs you already have. They sit at different points in the funnel.

What do roofing leads actually cost on these platforms?

Per-lead prices for roofing on aggregators like Networx and on Thumbtack generally run roughly $25 to $100+, with full-replacement and storm-driven inquiries in competitive metros at the high end and small repairs cheaper. But the per-lead price isn't your real cost. After accounting for shared leads you never reach, homeowners who already signed with a faster contractor, and bad leads, your cost per actually-won job is often well over $1,000 — which only works on high-ticket replacement jobs with a strong close rate.

Are these leads shared or exclusive?

Networx leads are typically shared, sold to three to four contractors at once. Thumbtack is also effectively shared — multiple pros quote the same homeowner. That's why speed-to-lead is everything on both: the first contractor to reach a human usually wins. If you can't call a new lead within about five minutes every single time, shared leads will underperform for you regardless of source.

Is Thumbtack worth it for roofing?

It can be, in a narrow lane. Thumbtack is most useful for newer or smaller roofing operations building a review base and catching repairs and smaller jobs, because its review engine compounds and the marketplace rewards fast responders. The downsides are real: shared demand, charges for leads that may not respond, periodic pricing-model changes, and the fact that the homeowner is Thumbtack's customer, not yours, so you never build a durable channel.

Hearth doesn't seem to send leads — what is it for?

Correct. Hearth's core job is offering homeowners monthly-payment financing for their roof and giving you digital quoting, e-signature, invoicing, and payment tools. It lifts your close rate and average ticket on leads you already have rather than adding new ones at the top of the funnel. Buy it if you lose replacement deals on affordability — just price in the dealer fee the lender charges, which comes out of your margin, and don't finance customers who'd have paid cash anyway.

Why is cost per lead a misleading way to compare sources?

Because a $25 shared lead and a $25 exclusive lead are completely different products. The number that matters is fully-loaded cost per won job by source, including the platform fee and the labor of working those leads. Tag every lead with its true first-touch source, track it to won/lost/dead, and divide total spend by jobs won. Cheap-per-lead sources frequently turn out to be expensive per won job, and that's the figure that decides where to spend.

What's the alternative to buying leads from these platforms?

Build an owned, targeted outreach channel instead of paying for inbound leftovers. Rank the homes in your service area by roof-age band and storm exposure, then proactively reach the overdue, storm-hit homeowners with tracked direct mail, personalized roof-report microsites, QR codes, and door-knock routes before they start shopping. The leads come back exclusive to you, the cost is fixed and visible, and the list and brand presence compound season over season — which rented marketplaces can't offer.

How does RoofPredict compare to these three platforms?

It's a different category. Networx, Hearth, and Thumbtack are point tools — a lead marketplace, a financing tool, a marketplace. RoofPredict is the targeting-plus-outreach-plus-CRM-plus-measurement platform underneath them: it ranks which roofs are due by age and storm exposure, turns that list into tracked mail, microsites, QR codes, and field routes, captures every lead with an immutable first-touch source synced two-way to 13 CRMs, and reports cost-per-won-job by source. You own the channel instead of renting it.

Is roof-age and storm targeting a guarantee a roof needs replacing?

No, and any honest vendor will tell you that. Roof age comes back as a range, not an exact install date, and storm exposure is a probability of impact — odds, not proof that a specific roof is damaged. It's heuristic targeting that puts the most likely-due, storm-exposed homes at the top of your list so you spend your mailing and knocking energy where it pays off, rather than a prediction that any single roof will fail.

Can I tell homeowners I'll get their roof covered or waive their deductible?

No. As a contractor you may inspect, document damage, and prepare an accurate repair estimate for your own scope, and state facts about that scope to the carrier. You may not, for a fee, negotiate or handle the claim, interpret the homeowner's policy, promise a payout or approval, promise the deductible is waived or absorbed, or advertise a 'free roof' — that's unlicensed public adjusting. Document thoroughly, write an Xactimate-aligned estimate, hand it to the homeowner; the homeowner files and the insurer decides coverage.