A sales development representative finds people who might buy, contacts them cold, works out whether they are worth a salesperson's time, and books a meeting. They do not close deals. That is the entire job, and the boundary is the point.
Every page you will find explaining this role was written by a company selling sales software, and they all describe the same list of responsibilities. What none of them explains is why the job exists at all, because it is not obvious. For most of commercial history the person who found the customer was the person who sold to them.
The arithmetic that created the role
Prospecting and closing are different skills, and more importantly they have different hourly values. A closer running a demo, handling objections and negotiating terms is doing work that only they can do. A closer building a list of 200 companies and sending first-touch emails is doing work almost anyone can do, at the same salary.
Split the two and you get specialisation on both sides. The closer spends their whole week in conversations that can produce revenue. The prospector gets good at the narrow, repetitive, genuinely difficult skill of starting conversations with people who did not ask to hear from you.
That is why you do not see SDR teams selling $200-a-month software to small businesses. The economics do not survive it. Below roughly a five-thousand-dollar deal, outbound prospecting as a separate salaried function loses money, and the companies doing it profitably are either running much higher contract values or have found a cheaper way to do the prospecting.
What the day actually looks like
Published benchmarks put the median SDR at around 46 dials a day with a 6.1% connect rate. Work that through: roughly three live conversations a day with someone who did not expect the call, from about four hours on the phone. The rest of the day is list building, research, writing sequences and updating the CRM.
| Metric | Median | What it means |
|---|---|---|
| Dials per day | 46 | About four hours of actual calling |
| Connect rate | 6.1% | ~3 live conversations daily |
| Meetings booked per month | 14.6 | Roughly one per working day |
| No-show rate | 20% | ~12 meetings actually held |
| Ramp to 80% quota | 3.9 months | Before which output is partial |
| Median tenure | 17.6 months | Then they leave or get promoted |
Put the last two rows next to each other and you get the number that explains most of what is wrong with the function. A rep takes 3.9 months to reach 80% of quota and stays 17.6 months. That is 22% of their entire tenure spent below full productivity, and then you hire again and pay it a second time.
Nobody selling SDR software puts those two numbers in the same table. Separately they are benchmarks. Together they are the argument for changing how the function works.
SDR, BDR, and the titles that mean nothing
You will see SDR, BDR, MDR and ADR used as if they are distinct roles. In practice the only distinction that consistently holds is direction of traffic: some teams call the inbound-facing rep an SDR and the outbound-facing one a BDR, and roughly as many teams do it the other way round.
Treat the title as noise and ask two questions instead. Does this person contact people who have not raised a hand? And do they hand off rather than close? Those two answers tell you everything the acronym was supposed to.
Why this matters before you look at any software
Every AI SDR product is priced against the number at the top of this page. The pitch is always some version of half the cost of a human, and that comparison only means something if you know what the human costs, what they produce, and which parts of their job the software actually replaces.
The next chapter itemises the cost properly, including the three line items that are almost always left out of the comparison.