The telemarketing industry generates about $16 billion in annual revenue in the United States. It employs, depending on the estimate, somewhere between 400,000 and a million people. More calls are made to American consumers and businesses today than at any point in history — by a large margin. In 2023, an estimated 50 billion robocalls were placed in the US alone. That number has held roughly steady since the TRACED Act was supposed to fix it.
So when people ask whether telemarketing is dead, the first honest answer is: it clearly isn't.
But the second honest answer is more complicated. Because what is dying — what is genuinely, structurally, irreversibly dying — is the version of telemarketing that built entire industries on the premise that if you call enough people, enough of them will say yes. That model is running out of runway in ways that have nothing to do with culture and everything to do with physics.
The Machine That Wouldn't Stop
To understand where telemarketing is going, it helps to understand why it never went away in the first place.
The basic economics are brutal and they work. If you can buy a list of 100,000 business phone numbers for a few thousand dollars, auto-dial them at a rate of 200 calls per hour, connect with 2% of them, and close 5% of those connections, you have 100 customers. If each customer is worth $500 in commission, you've made $50,000 from a $3,000 list. The math pencils even if you throw away 99,800 of the numbers.
This is why every law aimed at telemarketing has produced a variant of the same outcome: the model adapts, costs rise slightly, the worst actors absorb the fines as a cost of doing business, and calling continues. The Do Not Call Registry launched in 2003 with real teeth — over 200 million registered numbers — and the industry worked around it through consent-farming, business-to-business exemptions, and the patient tolerance of regulatory agencies with limited enforcement budgets.
The TCPA, which allows for $500–$1,500 per call in statutory damages, produces enormous class action settlements and genuine deterrence for companies with something to lose. It produces almost nothing against the fly-by-night operations that generate most of the call volume. You can't extract damages from a shell company in a strip mall that's already moved on by the time the lawsuit is filed.
Every declared death of telemarketing has run into the same problem: the model is simple enough that it regenerates. Anywhere the economics work and enforcement is uneven, the calls come back.
What Is Actually Changing Now
That said, something is different this time. Not one thing — several things converging simultaneously, and some of them are not working around the way previous obstacles did.
The Carrier Layer Is Moving Against Callers
For most of telemarketing's history, the phone carrier was a neutral pipe. It rang numbers without judgment. The only friction was legal, and legal friction was manageable.
That changed with STIR/SHAKEN.
STIR/SHAKEN — the authentication framework that the FCC mandated in 2021 — requires carriers to cryptographically verify the origin of calls and assign an attestation level: A (fully verified), B (partially verified), or C (unverified). Calls with C-level attestation get flagged. Unverified caller ID gets labeled. And increasingly, carriers aren't just flagging these calls — they're blocking them outright before the phone rings.
This is categorically different from any previous regulatory constraint. TCPA threatens you with money after the fact; STIR/SHAKEN stops the call from going through in the first place. You can't sue your way past it. You can't argue consent in a post-call hearing. The call doesn't complete.
The consequence: any operation that spoofs caller ID, uses rotating numbers, or lacks a clean origination path is increasingly reaching voicemail — or nothing. The carrier-level infrastructure is now an enforcement mechanism that the legal system never was.
The Smartphone Killed B2C Cold Calling
This one is cultural but it's irreversible.
A cold call to a landline in 2001 reached someone who had the phone at their desk, expected calls, and had no caller ID that told them to be suspicious. A cold call to a mobile phone in 2026 reaches someone who:
- Has immediate visual indication that the number is unknown
- Has been conditioned by years of spam calls to reject unknown numbers by default
- Has a carrier that may have already labeled the call "Spam Likely" before they see it
- Has a voicemail system they check less frequently than email
- Has a phone that can send all unknown numbers to voicemail automatically with a single toggle
The connect rate on a cold call to a random mobile consumer number is now, by most estimates, somewhere between 1% and 3%. That number has been declining steadily for a decade. It's not coming back because the conditioning that drives it — the association between "unknown number" and "someone trying to sell me something" — is now deep and well-earned.
This is why B2C cold calling in financial services is essentially over as a scalable model. Mortgage lenders learned this. Auto finance learned this. Consumer banking learned this. The funnel math doesn't work when fewer than one in fifty calls connects with anyone willing to listen.
AI Calling Is Poisoning the Well for Everyone
The most recent and most significant development is the explosion of AI-generated voice calls.
AI voice agents can now make calls that are, in brief interactions, indistinguishable from a human caller. They can handle objections, answer basic questions, and transfer qualified leads to human agents. They can run at scale — thousands of simultaneous calls — at a cost per call measured in fractions of a cent. The economic model that made human telemarketing attractive looks expensive next to this.
The problem is that the majority of AI calling operations are not running compliant campaigns. They are flooding the lines with calls that are illegal under the FCC's February 2024 ruling that AI-generated voices require prior express consent, that are not authenticated under STIR/SHAKEN, and that are burning through phone numbers faster than carriers can block them.
The effect on the broader calling ecosystem is what engineers would call an externality: AI calling is making everyone's calls worse. The carrier blocking systems respond to call behavior patterns, and patterns driven by AI spam volumes are causing legitimate business calls to be swept up in the same blocks. A commercial finance broker making compliant, human-agent calls from a verified number is losing connection rate not because of anything they did, but because the infrastructure is responding to a flood of garbage they didn't cause.
This is not a solvable problem for individual operators. It's systemic. And it's accelerating.
The B2B Exception — And Why It's Shrinking
Here's where commercial finance sits in an awkward position.
Business-to-business calls have always occupied a different legal space than consumer calls. TCPA's cell phone restrictions apply most forcefully to consumer personal numbers. A call to a business's main line, to a business owner's business-registered number — that's a different analysis. The commercial finance industry built its outbound model partly on this distinction.
The problem is that the distinction has eroded.
Small business owners in 2026 don't have a desk phone at a business number that's separate from their personal cell. They have one phone. That phone is their business phone, their personal phone, and the phone that gets called by anyone who bought a list with their contact information. When a broker dials the number on a "small business owner" lead list, there's a meaningful chance that number is a personal cell phone, owned by a natural person, with full TCPA consumer protections.
The FCC has not provided clear guidance on how to classify these calls. Courts have gone different ways. The practical result is that a broker who relies on the "B2B exemption" to justify outbound calling to small business owners is carrying legal risk that may not be manageable if the owner has ever used that number for personal purposes — which is nearly always.
This is why TCPA class action filings in the commercial finance space have been rising. The legal theory is available, the per-call damage is significant, and plaintiffs' attorneys have discovered that MCA brokers and commercial lenders make good targets.
What the Numbers Actually Show
The volume of outbound commercial finance calls has not decreased — if anything, it has increased as more participants enter the market. But several indicators suggest the model is under serious strain:
Connect rates are falling. Industry insiders report that outbound connect rates on purchased commercial lead lists have dropped from the 5–8% range common in 2018–2020 to 2–4% in recent years, with some operators reporting below 2% on heavily dialed categories like MCA and working capital. The same number of calls produces fewer conversations.
Cost per funded deal is rising. When you need more calls to produce the same number of conversations, and more conversations to close the same number of deals, the math on purchased list dialing starts to break. Brokers who built their model on a certain cost-per-close are finding that the model requires more volume, more compliance infrastructure, or both.
Attrition in call-dependent operations is high. The culture of volume-based outbound sales — dial through the objections, hit the numbers, find the ones who say yes — produces sales teams with high turnover, significant mental health overhead, and increasing resistance among the people being asked to make the calls. Experienced brokers are moving to inbound-dependent models not because they can't handle outbound, but because they've decided it's not what they want to build a career on.
Regulatory enforcement is becoming less predictable. When TCPA enforcement was concentrated in a few large class actions with predictable settlement ranges, you could model the risk. With state mini-TCPA laws multiplying, with courts interpreting TCPA independently post-Chevron, and with the FCC under sustained regulatory pressure to do more, the variance in outcomes has increased. Companies that thought they had their compliance framework dialed in are discovering their assumptions were wrong.
Where the Volume Goes
So if cold calling is breaking down as a scalable model, where does sales volume come from?
The answer, which has been developing for a decade and is now becoming the dominant model for lending originations at any scale, is inbound lead generation: the borrower contacts the lender, not the other way around.
This happens through search — a business owner Googles "working capital loans" or "MCA lender" and lands on a site that captures their information. It happens through content marketing. It happens through referrals from accountants, attorneys, and other professionals who trust specific lenders. It happens through verified marketplaces where borrowers submit funding requests and lenders compete for the relationship.
The structural advantage of inbound over outbound is not primarily cost — in a direct comparison, a well-run outbound operation can generate leads more cheaply than SEO at scale. The advantage is consent. A borrower who came to you has demonstrated intent. They want to talk about financing. The conversation starts from a completely different position than one that starts with "Hi, I'm calling from..."
For lenders and brokers, the consent question is also legal rather than just commercial. A borrower who submitted a funding request has documented their willingness to be contacted. The TCPA exposure is minimal. The state mini-TCPA exposure is minimal. The liability that has attached to outbound calling doesn't exist in the same way.
This is the shift. Not from calling to not calling — lenders who win inbound leads still call them, still have human conversations, still close deals on the phone. The difference is that the call happens with someone who expected it.
The Commercial Finance Industry's Structural Problem
Commercial finance — MCAs, working capital loans, equipment financing, factoring — built itself on outbound sales culture in a way that most other lending verticals didn't. The product moved through brokers who moved it through the phone. The broker channel remains important; the question is whether it can survive a continued erosion of the outbound model.
Some of it will. Brokers who have built referral networks — who get their leads from accountants, business attorneys, and existing clients — are largely insulated from the trend. Their leads are warm by definition. Their compliance exposure is minimal. Their close rates are high because the borrower has already been pre-qualified by the relationship that referred them.
Some of it won't. Brokers and ISOs who are entirely dependent on purchased list outbound — who have no inbound capacity, no referral network, and no compliance infrastructure — are operating a model whose margins are compressing and whose legal exposure is growing. The economics that made the model work at volume are degrading on both ends simultaneously: fewer calls connect, and each call carries more risk.
The ones caught in the middle are the most interesting case: experienced commercial finance professionals who know how to have a conversation with a business owner about financing, who can structure a deal correctly, who understand the products and the underwriting. Their skill is real. The channel they built it on is what's becoming unreliable.
What those people need is not a crash course in compliance — they need access to borrowers who have already raised their hand. The model where a lender or broker receives a qualified, consented funding request, reaches out from a verified position, and competes on terms and service rather than on who-called-first is not hypothetical. It's where originations are moving.
So Is It Dead?
No. And yes.
The call volume is not decreasing. The industry employing people to make calls is not shrinking. A business owner in a genuine capital crunch, reached by a broker who knows what they're doing and has a product that fits, will sometimes take that call.
What's dying is the premise. The premise that volume solves for everything — that if you make enough calls, something will work out. The premise that compliance is someone else's problem. The premise that a purchased list of cell phone numbers and a power dialer is a business model rather than a liability waiting to be filed.
The commercial finance industry grew on that premise. The lenders who are growing now are mostly growing on something different: inbound intent, documented consent, relationships that started because a borrower chose to start them. The brokers who are thriving are the ones who found a way to be the person the borrower calls rather than the person calling the borrower.
Telemarketing isn't dead. The specific model of telemarketing that treated consent as an obstacle rather than a requirement — that one is dying. And it's not being killed by any law or any regulator. It's being killed by the cumulative effect of millions of phone owners deciding, silently and individually, that the unknown number is not worth answering.
That decision, once made, doesn't reverse.
FundScout is a lead marketplace where borrowers submit funding requests and qualified lenders compete to serve them. The contact is always borrower-initiated, the consent is documented, and the lenders are vetted before they ever see a lead. If you're a commercial lender or broker looking for a channel that doesn't require a call list, learn more about how FundScout works.
Sources
- Telemarketing industry revenue and employment ($16 billion annually; 400,000–1,000,000 employees): IBISWorld, Telemarketing & Call Centers in the US Industry Report (2024)
- 50 billion robocalls placed in the US in 2023: YouMail Robocall Index — robocall.youmail.com
- TRACED Act — Pub. L. 116-105 (2019); directed FCC to require STIR/SHAKEN call authentication
- STIR/SHAKEN framework — FCC mandate effective June 2021; 47 CFR Part 64 Subpart AA; cryptographic call authentication and carrier attestation levels
- National Do Not Call Registry — FTC; 200+ million registered numbers; FTC DNC Data Book FY2023
- Connect rate decline (5–8% in 2018–2020 to 2–4% in 2024–2026): reported by industry practitioners; see RAIN Group, What It Takes to Win with Outbound Sales Calls (2020)
- TCPA class action surge — WebRecon LLC TCPA Litigation Tracker; webrecon.com
- Insurance Marketing Coalition v. FCC, No. 24-10277 (11th Cir. January 24, 2025) — vacated FCC one-to-one consent rule; courts now interpret "prior express consent" without Chevron deference
- Loper Bright Enterprises v. Raimondo, 603 U.S. 369 (2024) — eliminated agency deference under Chevron
- FCC FCC 24-17 (February 2024) — AI-generated voices classified as "artificial" calls under TCPA requiring prior express consent