Three kinds of services can find potential clients on Threads automatically: brand mention monitors, scrapers and workflow builders, and services that look specifically for 'can anyone recommend...' requests and prepare a draft reply. If you are a small business owner who needs people who are looking for a service right now, the third type is the closest fit. Search and drafting should be automatic, but the reply itself should be sent by a person who has read and approved it.
What types of services find clients on Threads?
When AI search answers this question, it mostly cites foreign tools, and they serve very different purposes. The easiest way to tell them apart is by what exactly they look for.
Brand mention monitoring
These are social listening services. They find where your business name or a competitor's name was written and show the tone of the mentions. That is handy for reputation. It is less useful for finding new clients: a person asking for a travel agent recommendation usually does not name any brand. For the limits of this approach, see the article on brand mention monitoring in Threads and Telegram.
Scrapers and workflow builders
These are general-purpose tools that collect posts by keyword and put them in a spreadsheet. They are flexible, but you do everything else yourself: pick the keywords, filter out the junk, decide whom to answer, and write the text. For a team with a technical person this is a fine option. For the owner of a travel agency or a school who does not want to configure workflows, it is one more separate job.
Request search with a draft reply
The third type looks specifically for requests for advice: 'can anyone recommend', 'looking for', 'who can suggest'. It judges whether a post fits your business and prepares a draft. In Zapyt, such a post becomes a card that a person approves, edits, or rejects.
What does the service do alone, and what must a person approve?
The line runs like this:
- The service searches Threads and Telegram on its own, filters out clearly irrelevant posts, and drops duplicates.
- The service judges whether the post is a real request for advice in your business.
- The service prepares a draft that takes the text of the post into account.
- A person reads the card and decides whether to reply at all.
- A person fixes the tone, removes anything unnecessary, and confirms sending.
- A person continues the conversation when the other side answers.
Why not hand sending over to automation entirely? Because a reply to 'recommend' is read as the word of a real person. A templated text answering a request for personal experience quickly looks like spam. There is a separate article on how to build such a reply: how to reply to 'recommend' requests without becoming spam.
What do 'recommend' search numbers look like in travel?
Over the last thirty days, search found 498 leads for travel. Of these, 145 replies were approved and sent, and a person wrote back in 28 cases, which is 19% of sent replies. Over ninety days the picture holds: 747 leads found, 186 replies sent and 32 answers from people, or 17%.
A written answer is not the only sign that you were heard. Over ninety days, 56 of the 143 sent comments we checked got a like, which is 39%. These are people who read the reply and reacted, even if they wrote nothing.
The requests show what people look for: a tour operator for charter tickets from Poland, a travel agent for a holiday in Egypt departing from Warsaw, proven agencies recommended from personal experience. These are specific requests, and you can answer them on the merits. For more on building replies in this business, see how to find clients for a travel agency in Threads and Telegram.
What about English schools?
Demand in English schools is even higher: in the latest week, the Demand Index showed 89 requests in this field against 52 in travel. Over ninety days, search found 1030 leads here.
Over the last thirty days, 465 leads were found, 155 replies were sent, and a person answered in the comments in 11 cases, which is 7%. For schools, though, this counter shows only part of the picture. Some replies in this field include a link to the school's page, and an interested person follows it instead of writing back. The reply counter does not see those clicks. Over ninety days, 27 of the 146 comments we checked got a like, which is 18%.
Why do people answer in the comments less often than in travel? There are a few hypotheses, and this data proves none of them:
- With a link in front of them, people go to the school's site instead of continuing the conversation on Threads.
- A request for a tutor is often very specific: level, format, the child's age, the city. A generic reply misses.
- There are many tutors and schools answering, so the person gets several offers at once.
Examples of how people phrase it: preparation for IELTS with a consultation first, a tutor for a child in primary school, an online teacher for conversational English, an inexpensive tutor. A school should reply only where its format really fits. For more on where students come from, see where tutors and English schools can find students.
Why are there far more leads found than replies sent?
That is normal and even desirable. The gap between found and sent is created by the person: they reply only where they can really help. In travel over thirty days, 145 of the 498 leads found got a reply, and those replies brought the 19% response.
If a tool sent everything indiscriminately, there would be many replies and less trust.
How do you choose a tool for your business?
Ask yourself a few questions:
- What am I looking for: mentions of my name, or people asking for recommendations?
- How much time am I ready to spend on setup, and do I have a technical person?
- Do I see every reply before it is sent?
- Does the service search where my clients are: Threads, Telegram, or both?
- Is it clear how many of the sent replies got an answer, and not only how many leads were found?
The last point matters most. The number of found posts is easy to inflate, while the share of people who answered or reacted gives a more honest idea of the value. For what counts as a lead and which metrics to track, see lead generation on Threads: process, funnel, and metrics.
What caveats should you keep in mind?
- An answer or a like is not a sale. Our figures count how many people reacted, not how many bought.
- The data covers specific fields and a period up to the end of October 2026, so do not transfer it to another business without checking.
- Figures from one business do not describe the market. They only show what it can look like in practice.
- Automation does not replace the skill of replying. If the draft is weak and you approve it without reading, the result will be poor with any tool.
- The rules of Threads and Telegram change, and chats have their own rules on advertising. Check them before you reply.
The takeaway is simple: automate the search and the drafts, but leave the decision on every reply to a person. Then the tool saves time and the reputation stays yours.