---
title: "AI Setter vs SDR: Cost and Performance Compared"
slug: ai-setter-vs-sdr
author: "Leonardo Maldonado"
category: "AI Setter Operations"
articleType: comparison
tags: ["SDR cost per year","fully loaded cost of an SDR","AI appointment setter cost"]
publishedAt: 2026-10-17T16:00:00.000Z
updatedAt: 2026-10-17T16:00:00.000Z
canonical: https://setluca.com/blog/ai-setter-vs-sdr
---# AI Setter vs SDR: Cost and Performance Compared

> AI setter vs SDR is a comparison most people run on the wrong number. A human sales development rep costs roughly twice their base salary once benefits, payroll tax, tooling and re-hiring are counted. About half of what they do is work software cannot touch. An AI setter costs a published monthly fee, plus your own minutes if you choose to review its replies. Below about 60 inbound conversations a month, neither one is worth buying.

## Key takeaways

- Benefits and legally required costs run 28.5% of what an employer pays a sales or office worker (US Bureau of Labor Statistics, Employer Costs for Employee Compensation, June 2026 reference period). A $50,000 offer letter is therefore closer to a $70,000 employment cost before you have bought a single tool.
- No government occupation code exists for an SDR. The two codes that bracket the role pay medians of $35,450 and $69,990 a year. That is why one job title can mean two very different budgets.
- Turnover is a line item, not an accident. Median tenure for workers aged 25 to 34 is 2.7 years (BLS Employee Tenure, January 2024, read through the St. Louis Fed's FRASER archive), and the ramp cost recurs every time the seat empties.
- An AI setter and an SDR are not doing the same job. Reply-only software answers conversations you already have. An SDR creates conversations that did not exist.
- Luca's published pricing is $99, $249 and $499 a month for 1,250, 4,000 and 8,500 replies, with a 7-day free trial (card required). Auto-send is on by default on every channel, and you can switch any channel to review.

---

## What is the difference between an appointment setter and an SDR?

A salaried SDR costs roughly twice their base pay once benefits, payroll tax, tooling and re-hiring are counted. An AI setter costs a published monthly fee, but it only answers conversations you already have, while an SDR creates new ones. An appointment setter answers and books. An SDR builds the list, opens cold conversations, qualifies against a profile, books the meeting, and hands you a briefed prospect. The setter role in coaching is usually part-time and commission-heavy. The SDR role is salaried, carries a quota, and has a manager.

The setter-versus-SDR distinction changes every number below. Commission-only setters carry almost no fixed cost, a different comparison we ran in our [AI setter vs human setter cost breakdown](/blog/ai-setter-vs-human-setter). If the person you are picturing only gets paid when a call books, read that one instead.

Pay diverges too. Setters take a percentage of closed revenue or a flat fee per show. SDRs take a base plus variable pay tied to meetings booked. We broke both models apart in [how appointment setters get paid](/blog/appointment-setter-commission-rate)

## AI setter vs SDR at a glance

An AI setter and a human SDR overlap on one task, booking a call, and separate on almost everything else.

|  | AI setter | Human SDR |
| --- | --- | --- |
| What it does | Answers inbound, qualifies, books, follows up | Builds lists, opens cold conversations, qualifies, books, briefs you |
| Cash cost | $99 to $499 a month, published | Base plus variable, plus benefits and payroll tax |
| Your time cost | Reviewing replies (if you keep review on) | Hiring, training, one-to-ones, call reviews |
| Hours covered | Every hour of the month | About 160 worked hours a month |
| Outbound prospecting | No | Yes |
| Messy, emotional objection | Handled poorly | Handled well, if they are good |
| Best for | High inbound volume, slow reply times | Creating demand that does not exist yet |

## What does a human SDR actually cost per year?

A human SDR costs more than the salary line, by a predictable multiple. Start with base pay, then add the variable component at the attainment you actually expect rather than at plan. After that come the costs that never reach an offer letter: benefits, payroll tax, tooling, and amortised re-hiring.

Government statistics have no SDR code, which is the first honest thing to say here. The two occupations that bracket the role are Telemarketers and Sales Representatives of Services. O*NET's 2025 wage data puts their medians at $35,450 and $69,990 a year. A junior inbound SDR sits near the bottom of that band, a senior outbound SDR near the top. Anyone quoting one national SDR salary is picking a point on that range without saying which.

The model below uses a $50,000 base and a $20,000 variable target, so $70,000 on target. Redo it with your own inputs.

## Why is the loaded cost so much higher than the salary?

The loaded cost runs higher because wages are only about seven-tenths of what an employer pays. For private-industry workers in sales and office occupations, benefits and legally required costs make up 28.5% of total compensation. Wages and salaries cover the other 71.5% (Bureau of Labor Statistics, Employer Costs for Employee Compensation, June 2026). Turn that around and benefits add roughly 40 cents on top of every wage dollar.

Grouped bar chart of employer cost per hour worked, split into wages and benefits, for all private industry workers and for sales and office workers

One trap sits inside that 40%. The benefits figure already contains the employer's share of Social Security and Medicare, which the IRS puts at 6.2% plus 1.45% of wages in 2026. Plenty of hiring calculators apply a benefits multiplier and then add 7.65% again on top, charging themselves twice for one tax.

### Where these compensation figures come from

Three different routes, and the difference matters if you want to check them yourself. The wage medians come from O*NET OnLine, which republishes the BLS wage series, because no SDR occupation code exists to look up directly. The benefits share comes from the BLS Employer Costs for Employee Compensation release for the June 2026 reference period. We could not open bls.gov from our network. FRASER, the St. Louis Fed's archive of that release series, had only reached the March 2026 reference period when we checked. So we name the release and the figure, and claim no retrieval date for it. If your own copy of those tables shows a different share, use yours.

## What does turnover add to the bill?

Turnover adds a recurring charge you spread across the months the seat is filled. Every departure costs recruiting hours, a ramp period below target, and the weeks the seat sits empty while inbound piles up.

Tenure tells you how often to expect it. Across all US wage and salary workers, the median time with one employer is 3.9 years. For workers aged 25 to 34, the band most SDR hiring draws from, it is 2.7 years (Bureau of Labor Statistics, Employee Tenure Summary). Plan for a sales development seat to empty well inside that 2.7 years; the model below assumes 18 months, as a planning input rather than a statistic.

Modelling it is simple. Take the ramp cost, divide by the months you expect the seat to stay filled, and carry it as a monthly line. A ramp wasting $14,000 across an 18-month tenure is $780 a month, for as long as you run the seat.

### How solid are the turnover numbers?

The turnover numbers are solid enough to budget against, with two caveats we would rather state than hide. Employee Tenure is a biennial release, and the newest issue on file still carries January 2024 as its reference date. We read it through FRASER, the St. Louis Fed's document archive, rather than from bls.gov, which would not open from our network. So 3.9 and 2.7 years are the latest published figures, not fresh ones.

The second caveat is a number we are deliberately not giving you: cost per hire. The industry benchmark page we went to check returned a 404. We will not quote a recruiting figure we cannot open. Recruiting spend is therefore an input you fill in from your own last hire, not a statistic you borrow from this page.

## What does the AI side actually cost?

The AI side costs a published subscription plus your own minutes. Luca is $99, $249 or $499 a month for 1,250, 4,000 or 8,500 replies (setluca.com/pricing). Every Luca plan comes with a 7-day free trial (card required). Other tools price by contacts, conversation or seat, and the models sit side by side in [how much an AI appointment setter costs](/blog/how-much-does-an-ai-appointment-setter-cost)

Your own review time is the line vendors leave off, so we will put it on. Auto-send is on by default on every channel, and you can switch any channel to review. That default matters because the arithmetic here only holds if the AI never costs you a relationship. The cheapest guarantee is making a human the last step. So your time is a real input, and fifteen minutes a day, six days a week, is 78 hours a year.

Setup is a third cost, paid in effort. Connecting channels, writing the qualification logic and tuning the writing style takes a working afternoon; the sequence is in [how to set up an AI setter](/blog/how-to-set-up-an-ai-setter)

## What do the AI setter options actually charge?

AI setter tools charge between $25 and $3,497 a month, and the spread says more about what each vendor meters than about what the tools do. Five meters are in use: AI messages, replies, active contacts, credits, and sub-accounts. Two tools publish a free tier, and every one of them publishes at least one price under $500 a month. Every price below was read off the vendor's own pricing page on the date in the last column, and any of them can change without notice.

| Tool | Entry price | What the price meters | Top published tier | Trial | Checked |
| --- | --- | --- | --- | --- | --- |
| SetSmart | $25/mo (Light) | AI messages, 250 a month | $797/mo (15,000 messages) | 7 days on Pro and Scale | Oct 16, 2026 |
| ManyChat | $39/mo (Pro) | active contacts, 2,500 included; ManyChat AI included | $199/mo (25,000 contacts) | free plan, 25 contacts | Oct 15, 2026 |
| CloseBot | $0 (Free) | 100 messages, one seat, unlimited agents | $997/mo (Core, 100K messages) | 7 days | Oct 15, 2026 |
| Luca | $99/mo (Starter) | replies, 1,250 included | $499/mo (Scale, 8,500 replies) | 7 days (card required) | Oct 16, 2026 |
| Appointwise | $97/mo (Starter) | sub-accounts, 10 included on Growth | $297/mo (Growth, 10 sub-accounts) | 14 days | Sep 25, 2026 |
| Setter AI | $497/mo (500 credits) | credits | $3,497/mo (5,000 credits) | none published | Oct 15, 2026 |

### How to read one of these quotes

Convert every quote into one unit before you compare anything: dollars per inbound conversation at your own volume. A lead is not a conversation. Setter AI meters credits, starting at 500 a month for $497. Until you know how many credits one conversation actually uses, that number is either cheap or tight. An active contact is not a conversation either, and the vendor writes that definition, not you.

Three more questions decide whether the headline number survives contact with a real month. Ask what happens when you cross the included quota, because SetSmart tops up automatically at $40 to $100 per thousand extra messages. Ask whether the discounted price needs an annual commitment. Ask whether the AI sits on the tier you priced or the one above it. ManyChat, for one, includes its AI from Pro up, not on Essential.

## Worked example: Priya's 400 conversations a month

Priya is an illustrative online business coach. She sells a $4,000 twelve-week programme, closes on calls herself, and gets about 400 inbound DM conversations a month across Instagram and Telegram. She keeps review on and approves every reply before it sends, and she is choosing between one salaried SDR and a subscription. Every figure below is illustrative.

### Building the SDR column

Start from the offer letter and keep adding until nothing is missing. Base $50,000, variable target $20,000, budgeted at 85% attainment rather than plan. Variable pay lands at $17,000 and cash compensation at $67,000. The 40% benefits load adds $26,700, already covering the employer's payroll tax. A CRM seat and a scheduler run $1,200 a year.

Ramp comes next. She assumes the first 90 days produce about 40% of steady-state output, so 60% of one quarter's employment cost buys nothing. A quarter of $94,900 is $23,725; 60% of that is $14,235, or $9,500 a year across an 18-month tenure.

### Priya's SDR cost, line by line

| SDR cost line | Annual | Monthly |
| --- | --- | --- |
| Base salary | $50,000 | $4,167 |
| Variable at 85% attainment | $17,000 | $1,417 |
| Benefits and legally required costs (~40% of wages) | $26,700 | $2,225 |
| Tooling | $1,200 | $100 |
| Ramp, amortised over 18 months | $9,500 | $792 |
| **Cash and near-cash total** | **$104,400** | **$8,700** |
| Priya's management time, 3 hrs/week at $100 | $15,600 | $1,300 |
| **All-in** | **$120,000** | **$10,000** |

Her $50,000 offer letter costs $104,400 in cash: 2.1 times base, 1.5 times on-target earnings. That ratio is the one worth memorising.

### Building the subscription column

The other side of the ledger has two lines instead of five. Luca's $249 plan is $2,988 a year. Reviewing every reply, 15 minutes a day, six days a week, is 78 hours, or $7,800 at the same $100 an hour. All-in, about $10,800 a year, or $900 a month.

Per conversation, the SDR costs $21.75 in cash and the subscription costs 62 cents. That gap is real and also misleading, because only one of the two can go and find her another 400 conversations.

## How many extra booked calls does each option have to produce?

Each option has to produce enough booked calls to cover its own monthly cost, and the two answers sit an order of magnitude apart. Priya's $4,000 programme closing on a quarter of booked calls is worth $1,000 per call, so the break-even is monthly cost divided by $1,000. Illustrative assumptions, not benchmarks.

The SDR has to produce nine net-new booked calls a month in cash terms, or ten once Priya's own hours count. Net-new is the load-bearing word: calls that would have booked anyway do not count toward that nine. The subscription has to produce a quarter of one extra booked call, or nine-tenths with review time included.

A quarter of a call against nine sounds like the software winning by 30 lengths. It is not. The SDR's nine calls can come from people who have never heard of Priya. The AI's fraction of a call can only come from her existing inbox. We priced those conversations in [the ROI of an AI appointment setter](/blog/ai-appointment-setter-roi)

## How do the numbers change at 150 and 800 conversations a month?

The subscription side moves with volume and the salary side does not, which is most of this comparison in one sentence. Priya's SDR costs $8,700 a month in cash whether her inbox holds 150 threads or 800. The software costs a plan fee plus her review minutes, and both track the number of conversations. Running her model at three volumes shows the shape. Same illustrative assumptions as above. One booked call is worth $1,000, her own hours cost $100, and review time scales from the fifteen minutes a day she spends at 400.

| Monthly conversations | Luca plan | Her review time | Software all-in | Software break-even | SDR all-in | SDR break-even |
| --- | --- | --- | --- | --- | --- | --- |
| 150 | $99 Starter | 2.4 hrs, $240 | $339/mo | 0.3 net-new calls | $10,000/mo | 10 net-new calls |
| 400 | $249 Pro | 6.5 hrs, $650 | $899/mo | 0.9 net-new calls | $10,000/mo | 10 net-new calls |
| 800 | $499 Scale | 13 hrs, $1,300 | $1,799/mo | 1.8 net-new calls | $10,000/mo | 10 net-new calls |

### Where the break-even actually moves

The break-even moves, but not where most spreadsheets put it. The software's break-even climbs with volume, from a third of a booked call at 150 conversations to nearly two at 800. It never leaves that band. The SDR's break-even does not move at all. Ten net-new booked calls a month at every volume, because a salary has no idea how busy your inbox is.

What changes is what those ten calls would mean. Hire the person to work the inbox at 150 conversations, and roughly one thread in fifteen has to become a call that was never going to book. At 800 conversations it is one in eighty. Point that same person at outbound instead and the ten calls can come from strangers, which is the honest argument for making the hire.

### What each conversation costs on each side

A conversation costs about $2.25 on the software side at every volume we modelled, and anywhere from $58 down to $11 on the human side. The software number stays flat because the review minutes, not the plan fee, are the real meter. The SDR number falls fast because a fixed salary spreads across more threads. Those two lines only cross somewhere north of 3,800 conversations a month, and one person cannot answer that many inside 160 working hours. So the human never actually wins on cost per conversation in a coaching business. Hire one for the conversations that do not exist yet, and price that decision on new revenue instead.

## At what volume does each option make sense?

Volume decides the software question, and the job to be done decides the human one.

### Buy the software or hire the person?

- **Buy the AI setter** above about 150 inbound conversations a month. At that volume reply times slip past a few hours, and leads go quiet before you answer. That is a coverage problem, and coverage is what software is good at.
- **Hire the SDR** if the constraint is that not enough conversations exist. If inbound is thin and your growth plan is outbound lists, partnerships or event follow-up, no reply-only tool fixes that at any price.

### When should you do both, and when neither?

- **Do both** above roughly 400 conversations a month with an outbound motion. Reply shift on software, prospecting on the human, and stop paying a salary to answer "how much is it?" at midnight.
- **Do neither** below about 60 conversations a month. You can answer 60 people yourself in under an hour a day. Spend the $99 on [coaching lead generation](/blog/coaching-lead-generation) instead.

If you are unsure whether the gap is setting or closing, [should you hire a setter or use AI](/blog/hire-a-setter-vs-ai) sorts it faster than a spreadsheet will.

## Where a human SDR still beats an AI setter

A human SDR still beats an AI setter in four situations we can name. We sell the software, so read these as the places we would tell you not to buy it.

**Outbound.** Luca is reply-only by design. It answers people who message you, at human pace, because that is what keeps accounts safe. It will not build a prospect list or send a cold first message. Any comparison treating it as an SDR replacement is comparing two different jobs.

**Complex qualification.** Some offers change shape per client, involve a second decision-maker, or need someone to understand a business before judging fit. A good SDR handles all three in one call. No thread-reading model does it reliably.

### Where judgment and deal size still favour a person

**Emotional reads.** A strong SDR notices that a lead answered three questions in ten seconds then took nine hours on the budget question, and slows down. That is judgment, and the honest limit on every tool in this category including ours. It is why [human vs AI closer](/blog/human-vs-ai-closer) gets its own page.

**Low volume, high deal value.** At 40 conversations a month on a $25,000 offer, one extra close pays for a quarter of a person. Coverage is not the problem there.

## What changes about performance, not just cost?

Coverage changes completely and judgment barely moves. A 30-day month holds 720 hours, and a full-time employee works about 160 of them after holidays and leave. One SDR covers a fifth of the month's clock, and it is the fifth when your leads are at work.

Round-the-clock coverage matters more on messaging channels than it did on the phone, because the platforms put clocks on your replies. Meta's Messenger Platform policy gives a business 24 hours to respond to a user with promotional content. A human agent tag extends manual replies to seven days. Miss the window and your follow-up narrows to templates. Software does not miss it at 2am; a sleeping person does.

Quality on the hard messages does not change. Speed gets you more conversations to be good at without making you good at them, the argument we made in [do AI setters actually work](/blog/do-ai-setters-work)

## What about the awkward cases?

Four awkward cases come up often enough to plan for.

**The commission-only setter.** No benefits load, no payroll tax, no ramp amortisation, because the cost only appears when revenue does. Compare that against a subscription on cost per booked call, not per month.

**The overseas contractor.** The 40% benefits load usually does not apply, which is most of why the rate looks better. Worker classification rules are jurisdiction-specific, so price that arrangement with an accountant rather than a blog post.

**The seasonal spike.** A January launch is a six-week surge, and an annual seat to cover it means paying eleven months for six weeks of need. Subscriptions cancel; employment contracts do not. The year-end sequencing in [how to get coaching clients before the end of the year](/blog/how-to-get-coaching-clients-before-the-end-of-the-year) assumes no fixed cost locked against a temporary problem.

**The referral-only practice.** If most clients arrive through people who already trust you, both columns are the wrong purchase. Your constraint is reach, and neither a salary nor a subscription buys it.

## Troubleshooting: when your numbers stop matching reality

| Symptom | Likely cause | Fix |
| --- | --- | --- |
| Cost per booked call climbs after the hire | You budgeted at 100% attainment and pay fixed costs against below-plan output | Rebuild at actual attainment; if base alone beats your target cost per call, the seat is wrong, not the rep |
| The subscription is cheap, nothing improved | You bought coverage for a lead-supply problem | Count inbound for 30 days; under 60, spend on generating leads |
| Booked calls rose, revenue did not | Qualification is too loose, so the calendar fills with people who were never buying | Require a budget and timeline answer before the booking link appears |
| You hired an SDR and still answer DMs at 11pm | One person covers outbound and inbound, and inbound wins the day | Split the jobs: reply shift on software, prospecting on the human |
| A saving your bank account cannot find | Opportunity-cost hours were booked as cash | Keep cash and time in separate columns; only cash changes your balance |
| The model ignores the seat emptying | No ramp amortisation | Divide ramp cost by expected tenure in months and carry it monthly |

## Common mistakes when costing this comparison

Six mistakes turn up in almost every buy-versus-hire spreadsheet we get sent. Three of them distort the pay math, and three leave a real cost out of the model altogether.

### Mistakes in the pay math

**The Offer-Letter Number.** Quoting base salary as the cost of the hire. It understates the real figure by roughly half, which is why most buy-versus-hire posts reach the wrong answer.

**The OTE Illusion.** Budgeting variable pay at 100% attainment. Plan is a target, not a forecast, so model the attainment you would bet money on.

**The Double-Counted Payroll Tax.** Applying a benefits multiplier and then adding 7.65% on top. The multiplier already contains the employer's Social Security and Medicare share.

### Mistakes in what the model leaves out

**The Free Manager.** Pricing your own supervision at zero. One-to-ones, call reviews and script feedback are hours you are not coaching, and they belong in the model even though no money moves.

**The Permanent Seat.** Modelling a hire as if nobody leaves. Total separations ran 3.2% of US employment in the single month of July 2026 (BLS series JTSTSR, read on FRED, the St. Louis Fed's data service). Compounded across twelve months, that is about a third of all seats turning over in a year. Yours is not exempt.

**The Same-Job Comparison.** Putting an outbound SDR and a reply-only AI in one column and declaring a price winner. They overlap on booking and diverge everywhere else.

## How to run the AI setter vs SDR numbers for your own business

Five steps and one afternoon gives you a defensible answer instead of a feeling.

1. **Count your inbound.** Thirty days of unique conversations, not followers or impressions. This number decides the software question on its own.
2. **Write down the real hire.** Base, variable target, and the attainment you honestly expect. Multiply wages by 1.4 for benefits and payroll tax, then add tooling.
3. **Amortise the ramp.** Estimate the output shortfall over the first 90 days. Take that share of one quarter's cost, then divide by the months you expect the seat to last.
4. **Price your own hours twice.** Once for managing a human, once for reviewing replies if you keep review switched on. Keep both in a column separate from cash.
5. **Convert both to booked calls.** Divide each monthly cost by the expected revenue of one booked call. Then ask which option produces that many calls you would not have had anyway.

### What to do with the answer

If the AI setter vs SDR arithmetic lands on software, start on the plan matching your message volume. Auto-send is on by default on every channel, and you can switch any channel to review. Telegram, Instagram and Messenger are live today. If it lands on a human, hire for outbound and stop asking that person to babysit your inbox. The wider case sits in [the complete guide to AI setters for coaches](/blog/what-is-an-ai-dm-setter). Whether the role survives the decade is a question we took seriously in [will AI replace appointment setters](/blog/will-ai-replace-appointment-setters)

Price the software side properly before you write anyone an offer letter. [Start a 7-day free trial of Luca](/pricing) (card required) and run one real month of conversations through it first.


## FAQ

### How much does an SDR really cost compared with an AI setter?

A $50,000 base becomes roughly $104,000 a year in cash once benefits, payroll tax, tooling and amortised re-hiring are added. That is about 2.1 times the salary line. A published AI setter subscription runs $99 to $499 a month, plus your own time if you choose to review its replies.

### Is an AI setter cheaper than hiring an SDR?

On cash cost, yes, by roughly an order of magnitude. On capability, they are not equivalent. Reply-only software answers conversations you already have. An SDR creates new ones through outbound work no reply-only tool performs, so cheaper does not automatically mean better.

### What is the difference between an SDR and an appointment setter?

An SDR builds prospect lists, opens cold conversations, qualifies and books, usually on a salary plus variable pay. An appointment setter typically answers inbound and books, often part-time on commission. The SDR carries fixed cost and a quota; the setter mostly carries neither.

### How many leads do I need before an AI setter is worth it?

In our model, about 150 inbound conversations a month is where coverage starts to pay for itself, and above 400 the argument gets hard to refuse. Below roughly 60 a month, answer them yourself and spend the subscription money on generating more leads.

### Can an AI setter do outbound prospecting?

Luca does not. It replies to people who message you, at human pace, because unsolicited first messages at volume are what gets accounts restricted. If outbound is your growth plan, that job belongs to a person, and no comparison on price changes it.

### What percentage should I add to a salary for benefits?

Roughly 40% of wages for a US sales or office employee. That comes from benefits making up 28.5% of total employer compensation cost. The share already contains the employer's Social Security and Medicare, so do not add payroll tax again on top.

### How long does a new SDR take to ramp?

Plan for a full quarter before output reaches steady state, and treat the shortfall as a real cost. Assume 40% of target output across the first 90 days. Roughly 60% of one quarter's employment cost then buys you nothing, every time you fill the seat.

### Should I hire an SDR or buy an AI setter first?

Buy the software first if you have inbound conversations going unanswered, because it is reversible in a month. Hire first if your problem is that too few conversations exist at all, since that is the one thing a reply-only tool cannot fix.


## Sources

1. [US Bureau of Labor Statistics. "Employer Costs for Employee Compensation, June 2026" (news release, private industry total compensation $46.89 per hour worked; wages 70.0%, benefits 30.0%), and Table 4, private industry workers by occupational and industry group (sales and office occupations, June 2026: total compensation $36.34 per hour, wages $26.00, benefits $10.34, benefits 28.5% of total). and https://www.bls.gov/news.release/ecec.t04.htm, NOT retrieved: bls.gov did not resolve from our network, so these figures are cited from the release rather than fetched from it.](https://www.bls.gov/news.release/ecec.nr0.htm)
2. [Federal Reserve Bank of St. Louis (FRASER). "Employer Costs for Employee Compensation" release archive (mirrors the BLS release series; newest issue on file at the time of writing is the March 2026 reference period, released June 12, 2026, which is why the June 2026 figures above carry no retrieval date).](https://fraser.stlouisfed.org/title/employer-costs-employee-compensation-9615)
3. [O*NET OnLine. "Summary Report for: 41-3091.00, Sales Representatives of Services" (2025 median annual wage $69,990; projected 2024-2034 outlook). Retrieved October 17, 2026.](https://www.onetonline.org/link/summary/41-3091.00)
4. [O*NET OnLine. "Summary Report for: 41-9041.00, Telemarketers" (2025 median annual wage $35,450; projected 2024-2034 outlook). Retrieved October 17, 2026.](https://www.onetonline.org/link/summary/41-9041.00)
5. [US Bureau of Labor Statistics. "Employee Tenure, 2024" (USDL-24-1971, released September 26, 2024; median tenure 3.9 years for all wage and salary workers and 2.7 years for workers aged 25 to 34, both as of January 2024), read through the FRASER archive of the Federal Reserve Bank of St. Louis because bls.gov did not resolve from our network. and https://fraser.stlouisfed.org/files/text/releases/tenure/tenure_20240926.txt, Retrieved September 25, 2026. Original release location: https://www.bls.gov/news.release/tenure.nr0.htm](https://fraser.stlouisfed.org/title/employee-tenure-9641/2024-707079)
6. [US Bureau of Labor Statistics, Job Openings and Labor Turnover Survey. "Total Separations: Total Nonfarm" (series JTSTSR; July 2026 rate 3.2%, seasonally adjusted, series updated September 1, 2026), read on FRED, the data service of the Federal Reserve Bank of St. Louis, because bls.gov did not resolve from our network., Retrieved September 25, 2026. Original release location: https://www.bls.gov/news.release/jolts.nr0.htm](https://fred.stlouisfed.org/series/JTSTSR)
7. [Internal Revenue Service. "Topic no. 751, Social Security and Medicare withholding rates" (employer share 6.2% Social Security to a $184,500 wage base plus 1.45% Medicare, 2026). Retrieved October 17, 2026.](https://www.irs.gov/taxtopics/tc751)
8. [Meta for Developers. "Messenger Platform Policy Overview" (24-hour standard messaging window; human agent tag extending manual replies to 7 days).](https://developers.facebook.com/docs/messenger-platform/policy/policy-overview)
9. [Luca. "Pricing" (Starter $99/mo for 1,250 replies, Pro $249/mo for 4,000, Scale $499/mo for 8,500; 7-day free trial, card required).](https://setluca.com/pricing)
10. [SetSmart. "Pricing" (Light $25/mo for 250 AI messages, Pro $97/mo for 1,000, Scale $297/mo for 4,000, Enterprise $797/mo for 15,000; automatic top-ups at $40 to $100 per 1,000 extra messages; 7-day trial on Pro and Scale). Retrieved October 17, 2026.](https://www.setsmart.io/pricing)
11. [ManyChat. "Pricing" (Free 25 active contacts, Essential $17/mo for 250 contacts, Pro $39/mo for 2,500 contacts with ManyChat AI included, Business $99/mo for 7,500, Advanced $199/mo for 25,000; checked October 15, 2026).](https://manychat.com/pricing)
12. [CloseBot. "Plans for Businesses and Agencies" (Free plan $0 with 100 monthly messages, one seat and unlimited AI agents; Business Core from $97/mo, scaling to $997/mo by volume, with a 7-day trial; Agency Core from $397/mo, quoted separately).](https://closebot.com/plans/)
13. [Appointwise. "Pricing" (Starter $97/mo, Growth $297/mo for 10 sub-accounts with 10 AI agents each, $29/mo per extra sub-account, Enterprise quoted on request; 14-day free trial). Retrieved October 17, 2026.](https://www.appointwise.io/)
14. [Setter AI (TrySetter). "Pricing" (credit plans: $497/mo for 500 credits, $2,297/mo for 2,500, $3,497/mo for 5,000; from $397/mo billed yearly; no free plan or free trial listed; checked October 15, 2026).](https://www.trysetter.com/pricing)

---

Published by SetLuca.