Yes, an AI appointment setter can handle the routine replies from cold outreach (answering common questions, offering times, sending reminders), as long as people still own the process around it. Replies are where many campaigns lose meetings. In our first two months for Digital Phosphenes, a US personal-branding agency, we booked 3 calls and 2 were no-shows or unqualified. By month 6, we booked 10 calls a month.
The fixes were templated replies, reminders before each call and mock sales calls. These were process changes made by our team, and none of them used AI. They fixed the stage where that campaign was losing meetings. Today our AI appointment setters do the reply part of this work, and our team still runs the call preparation.
Where do cold outreach campaigns lose meetings?
A campaign can get plenty of replies and still book few meetings if those replies are handled slowly. Replies go unread for days, and positive leads wait in a shared inbox until the prospect has moved on. We see poor management as the main reason outbound fails, and reply handling is usually where it shows first.
A new client campaign we started in October 2026 shows how much work arrives at once. In its first days of sending, it reached 1,191 leads and got 57 replies (4.79%). About 34 of those replies were positive, which is 60% of all replies. Within the first week, 16 calls were booked, or 1.3% of the leads contacted.
Each of those 34 positive replies needed an answer, a few proposed times and often a follow-up before a call went into the calendar. If one person handles that alongside their normal job, some of those conversations will slip. Many firms give up on outbound for this reason, and we covered it in more detail in why outbound overwhelms your team.
What changed for Digital Phosphenes between month 2 and month 6?
Digital Phosphenes is a US personal-branding agency. In the first two months of its campaign, we booked 3 calls, and 2 of them were no-shows or unqualified. At the start, the founder closed about 10% of the calls they took.
We made three changes. The SDR started answering replies with templates written for the common questions, so prospects got consistent answers without waiting for someone to write each one. Prospects with a booked call received reminders before it, and we ran mock sales calls to prepare for the conversations themselves.
The show rate rose above 75% by month 4. By month 6, the same system was booking 10 calls a month. Over the six months, the campaign produced 673 replies, 24 meetings (3.6% of replies), 17 opportunities and $105,000 in signed contracts. The full numbers are in the Digital Phosphenes case study.
The changes we made were all in how replies and booked calls were handled. Our AI appointment setters on email and LinkedIn now do the reply part of this work for our clients.
Which parts of reply handling can AI do well?
We built our own infrastructure for the work that happens after a reply, because this is where we saw campaigns fail most often. It has four parts, and each one removes a delay that used to depend on someone being free.
| Task | What the AI does | What a person still does |
|---|---|---|
| Answering replies on email and LinkedIn | Reads each reply, answers common questions and offers meeting times | Writes and approves the reply templates and takes over conversations that need judgement |
| Meeting briefs | Writes a brief automatically when a call is booked | Reads the brief before the call |
| CRM updates | Sends positive leads to the client’s CRM with enriched details | Works each lead from the CRM after the first call |
| Call reviews | Analyses each sales call and writes coaching notes | Decides what to change in the pitch |
A manufacturing client shows what this looks like at volume. In 30 days, our system put 70 leads into its CRM. Each lead was enriched automatically, and the account executives received a briefing before each meeting. Every call was then analysed, with coaching notes for that specific call.
The client pays a retainer of €8,000 a month, which works out at about €114 per opportunity. I do not think ads, SEO, content or LinkedIn posting could match that cost and volume for this client.
What should a person still own?
An AI appointment setter can only work with the replies a campaign produces. If the audience is wrong or the offer is weak, it will answer polite refusals very quickly. We use the NPOT Methodology™ to set the niche, the problem, the outreach offer and the target outcome before any reply handling starts.
Your reply numbers tell you which part to fix. A reply rate under 1% usually points to a deliverability problem, because copy cannot be tested until the emails land. A reply rate over 1% with fewer than 10% positive replies usually points to the message or the offer.
The October campaign above had a 4.79% reply rate and 60% positive replies. In NPOT terms, it already had the niche, problem and outreach offer right, so an AI appointment setter could focus on the target outcome: turning positive replies into booked calls.
A person should also own the decisions that need judgement. That includes which replies deserve a personal answer from someone senior and when to change the reply templates. The AI follows the rules it is given, so someone has to update those rules when prospects start asking new questions.
Are old replies worth following up?
Every campaign builds up a list of prospects who replied once, showed interest and then went quiet. When we re-engage those past conversations, we usually book 2 to 5 calls per 100 leads.
The message is short and gives a clear reason to talk again: “We spoke a while back about [outcome] for [company]. I wanted to reconnect because we [new change], which means [benefit]. Open to chatting later this week?” It works when the change is real, such as a new service or a new case study.
This is a good task for AI because it follows the same steps for every lead. The system can find every lead that went quiet, send the message and pass anyone who answers to the appointment setter. Few people have time to do this by hand for hundreds of old threads.
How can you check your own reply handling this week?
These steps assume you already run cold outreach and can see last month’s replies in your inbox or CRM.
- Count last month’s positive replies and the calls they turned into. A large gap between the two means meetings are being lost after the reply.
- Check how long your last 20 positive replies waited for an answer, using the timestamps in your inbox.
- Write reply templates for the questions prospects ask most often, such as price and examples of past work.
- Send a reminder before every booked call, and track your show rate each month.
- Pull every lead who replied with interest in the past year and send them the re-engagement message above.
- Once these steps work by hand, give the repeated parts to an AI appointment setter and keep the judgement calls with your team.
If you would like a second view on how your replies are handled, request a call and talk through your outbound with our team.
FAQ
Should an AI appointment setter book meetings without a person checking?
It can for routine replies, such as a prospect asking for a time or a short summary of the offer. We still recommend that a person reviews replies that raise objections, ask about price or come from a very senior buyer. Set clear rules for which replies the AI answers alone and which it passes to your team, and review those rules every month.
When does the AI write a meeting brief?
Our system writes a brief automatically when a call is booked. For one manufacturing client, the account executives received a briefing before each meeting, and every call was then analysed with coaching notes for that call. The person taking the call only needs to read the brief before the meeting starts.
Will an AI appointment setter fix a low reply rate?
No. It works on the replies a campaign already gets. If your reply rate is under 1%, check deliverability first, because copy cannot be tested until the emails land. If replies are above 1% but fewer than 10% are positive, the message or offer usually needs work. Fix those first, then improve how the replies are handled.
Can AI handle replies on LinkedIn as well as email?
Yes. We run AI appointment setters on both channels. LinkedIn suits audiences who are active there, such as tech founders who use LinkedIn messages to talk to partners and investors. Email suits larger audiences because it scales further. The same reply rules and templates can apply on both channels, so prospects get consistent answers wherever they reply.
Do you still need an SDR if AI handles replies?
You still need someone to own the process. That person updates the reply templates, handles conversations that need judgement and decides what to change after the call reviews. The AI takes over the repeated work, such as answering common questions, sending reminders and moving positive leads into the CRM, so your team spends its time on the conversations that need a person.
Want to Turn Outbound Into a Predictable Source of Pipeline?
Thanks for reading this article, I hope you found it useful. I'm Tudor Dumitrescu, founder of TANDA Digital (P.S. follow me on LinkedIn here).
If you're looking to generate more qualified pipeline through outbound, here are two ways we can help:
1. Get a Complimentary Outbound Campaign Drafted For You
We open every potential engagement with a complimentary outbound campaign that includes the messaging and 25 target accounts (+ decision-makers) created for you. Get Your Free Outbound Campaign
2. Book a Call
If you're not sure if outbound is the right move, you can book a call with our team. We'll assess your current situation and tell you if outbound could generate pipeline for your business.
