What an AI SDR is
An AI SDR is software that does the job of a sales development rep: it picks up a new lead, researches the person and company, replies, asks qualifying questions and books a meeting with an account executive. The good ones also log everything in the CRM and hand the seller a short brief before the call.
The name covers two different jobs, and most of the confusion around the term comes from mixing them up.
Inbound AI SDR vs outbound AI SDR
| Inbound AI SDR | Outbound AI SDR (AI BDR) | |
|---|---|---|
| Who it talks to | People who filled in a form, booked a demo, started a chat or downloaded something | Target accounts that have not contacted you |
| What triggers it | A form fill, chat or website visit | A list of target accounts |
| What matters most | Speed and accuracy of the first reply | Relevance of the first message, and protecting your sending domain |
| Main risk | A slow or wrong answer to someone ready to buy | Spam complaints and a damaged domain |
| Typical channels | Email reply, website chat, SMS, calendar link | Email sequences, LinkedIn, sometimes phone |
This page covers the inbound side, which matches how most sales teams use the SDR title: following up on demand that marketing already created. The outbound side has its own guide: AI BDR: what it does, what it can’t, and how to run one with a human in the loop. For the agents that take over after the first meeting, such as call briefs, recaps and CRM updates, see our guide to AI sales agents.
Why speed is the whole point of an inbound AI SDR
Inbound leads go cold fast, and human teams are slow. In research published in Harvard Business Review, the authors audited 2,241 US companies by sending each a web lead:
- 37% responded within an hour.
- 24% took more than 24 hours.
- 23% never responded at all.
- The average response time, among companies that answered within 30 days, was 42 hours.
In a separate study of 1.25 million leads in the same article, firms that tried to contact a lead within an hour were nearly seven times as likely to qualify it as firms that waited even an hour longer, and more than 60 times as likely as firms that waited 24 hours or more.
The study is from 2011, so treat the exact numbers as dated and the direction as the lesson. An AI SDR closes that gap: a demo request at 9 pm gets researched and answered before the seller wakes up.
What an inbound AI SDR does, step by step
- Captures the lead from the form, chat or calendar tool.
- Researches it: company size, industry, tools used, the person’s role, and whether the account is already a customer or an open deal.
- Scores it against your ideal customer profile, often with simple rules like BANT or your own fit criteria.
- Routes it to the right seller by territory, segment or round robin. See lead routing.
- Replies: sends a confirmation and calendar link, answers a simple question, or drafts a personal reply for the seller to approve.
- Follows up if there is no answer, and flags leads that sit untouched.
- Hands off with a short brief so the account executive does not start from zero.
What it can’t do well yet
- Know your exceptions. It will treat a churned customer, a competitor or a partner like any other lead unless your CRM fields say otherwise.
- Answer commercial questions safely. Pricing exceptions, contract terms and roadmap questions need a person.
- Run discovery. It can ask qualifying questions in writing. It cannot read a hesitant buyer on a call.
- Fix bad data. If your forms let people enter “test” in the company field, the research step has nothing to work with.
An illustrative inbound flow
Illustrative, not a benchmark. The company is fictional.
Larkspur HR sells onboarding software to companies with 100 to 1,000 employees. At 9:12 pm on a Tuesday, an HR director at a 450-person logistics company requests a demo.
| Time | What happens |
|---|---|
| 9:12 pm | Form submitted |
| 9:13 pm | The AI SDR finds the company size, that it is hiring 30 warehouse roles, and that it is not a customer or open deal |
| 9:14 pm | Scores it as a strong fit and routes it to the mid-market rep for the region |
| 9:15 pm | Sends a confirmation with the rep’s calendar link |
| 9:15 pm | Drafts a personal note for the rep, mentioning the 30 open roles and onboarding at that volume |
| 8:05 am | The rep reads the brief, edits one line of the draft and sends it |
| 8:40 am | The buyer books a call for Thursday |
The software did the first hour of work overnight. The person did the one step where a wrong word would have cost something.
How Growigami runs inbound follow-up
Our inbound lead follow-up service works like the flow above: each inbound lead is researched, scored and sent to the right seller with a reply drafted, at any hour. The seller checks it and sends it, and a lead that sits untouched goes back to the team before it goes cold.
Common questions about AI SDR
What is an AI SDR?
Software that does the work of a sales development rep. It picks up new leads, researches them, sends or drafts the first reply, asks qualifying questions and books meetings for account executives. Most tools also sort replies and update the CRM.
What is the difference between an AI SDR and an AI BDR?
Vendors use the names loosely. In human sales teams, SDRs usually follow up on inbound leads and BDRs prospect outbound. Applied to software, an inbound AI SDR responds to people who filled in a form or visited your site, and an AI BDR starts conversations with accounts that never raised their hand.
Will AI take over SDR jobs?
It is taking over the fastest, most repetitive part: the first reply, the research and the scheduling. Judgment calls stay with people: whether a lead is worth a seller's hour, how to handle a tricky question, and the discovery call itself. Expect fewer SDR hours spent on typing and more on conversations.
How fast should an AI SDR respond to an inbound lead?
Within minutes, and certainly within the hour. In a study published in Harvard Business Review, firms that tried to contact a lead within an hour were nearly seven times as likely to qualify it as firms that waited even an hour longer.
Should an AI SDR send replies without a human?
For scheduling, confirmations and simple product questions, it is reasonable to let it. For anything that commits the company, like pricing, discounts or custom terms, keep a person approving the draft. A wrong answer to a warm lead costs more than a reply that arrives ten minutes later.
