AI illustrationWealth and investment migration · LinkedIn
One in three wealth and migration professionals replied with a real answer
A referral network that connects wealth and migration advisors with firms that serve their clients ran Sellinger agents on 26 LinkedIn profiles of its team and member firms. They opened peer-to-peer conversations with private bankers, wealth managers and golden-visa advisors across the UK, Europe and the Gulf, and booked calls for the network’s team.
- Client
- Referral network for wealth professionals (anonymized)
- Audience
- Private bankers, wealth and golden-visa advisors
- Channel
- LinkedIn, 26 profiles
- Window
- Nov 2025 to Sep 2026
- Data as of
- 10 October 2026
reply rate: 4,392 of 9,808 professionals messaged, quick replies included
replied with more than a pleasantry: 3,119 of 9,264 the agent wrote to first
of 1,751 “thanks for connecting” replies the agent turned into real conversations
calls booked and corroborated, April to August 2026, once the calendar was connected
In short
- A referral network lives on trust, so its agents wrote as peers, not vendors: a compliment or one question about the prospect’s clients, no pitch and no link in the first message.
- LinkedIn quick replies inflate any reply rate. Of 1,751 “thanks for connecting” replies, the agent turned 1,035 into real conversations by asking how the prospect handles their own clients.
- Recruiting private bankers and golden-visa advisors worked best. Profiles run for individual member firms replied at 27.4% and booked under 1%.
9 min read
A referral network that runs on trust, built one peer conversation at a time
The client is a referral network with two sides. Advisors (private bankers, wealth managers, IFAs, migration lawyers, relocation specialists) join free and are paid when a referral closes. Firms that serve their clients (developers, investment-migration firms, wealth offices, legal and tax firms) get introductions.
Both sides had to be won on LinkedIn, by a peer, not a vendor. Sellinger’s agents wrote from 26 LinkedIn profiles of the network’s team and member firms, to prospects across the UK, Europe and the Gulf, and booked short calls for the network’s team.
AI illustrationWhat
Warm introductions to HNW clients
Advisors refer clients to vetted firms and are paid when a referral closes.
AI illustrationWho
Private bankers and golden-visa advisors
Wealth managers, IFAs, migration lawyers and relocation specialists in the UK, Europe and the Gulf.
AI illustrationHow
A peer-to-peer chat on LinkedIn
From the profiles of the network’s team and member firms, with a short call as the only ask.
professionals sent a connection request
of requests accepted since March 2026 (5,969 of 18,362)
messaged on LinkedIn after connecting
Not a warm list. 95% of the prospects the agent wrote to first came from cold connection requests, and they replied at 42.7%. First-degree connections replied at 47.0%, third-degree at 43.6%.
The playbook: a peer opener, a question about their clients, a 15-minute call
The median gap between the first message and the first follow-up was 5.0 days.
Day 0
Connection request
From one of 26 profiles; 47% of the prospects the agent wrote to first got a note.
On accept
A peer-to-peer opener
A compliment on their work, or one question about their clients. No pitch, no link.
+5 days
Follow-ups
About five days apart, up to four in the current setup, plus a revival nudge after about 90 days.
Any reply
A role-specific question
Even after “thanks for connecting”: how do your clients handle this today?
Interest
A 15 to 20 minute call
“Tuesday or Wednesday?” plus a booking link; often handed to a partner from the network’s team.
Who the agents wrote to
- Private bankers, wealth managers, family-office staff, IFAs, golden-visa and migration advisors, found by LinkedIn searches on role and country (“Golden Visa Advisor, UAE”, “Family Office, Switzerland”).
- Imported lists, and people engaging with a founder’s posts or with large brands’ pages.
- Firm-side profiles also wrote to individual bankers, so the two audiences overlapped.
Rules the agents work by
- A peer practitioner, never salesy. Outcomes first, then trust and discretion; how it works only if asked.
- First messages of three or four sentences. Stop when a prospect is irritated.
- English by default; Portuguese, Spanish or Greek when the prospect writes in it.
Thousands of replies, 3,119 real conversations, 95 booked calls
The reply rate counts every reply, including LinkedIn quick replies such as “thanks for connecting”. So every first reply was classified: 2,025 said something real, 1,751 were pleasantries. Calls are counted only from April 2026, when the calendar was connected, and attendance was not recorded, so these are calls booked, not held.
AI illustrationThe funnel
LinkedIn threads the agent opened, 14 Nov 2025 to 18 Sep 2026
- Messaged first by the agent9,264
- Replied4,012 (43.3%)
- Replied with substance3,119 (33.7%)
- Calls booked (Apr to Aug)75
Counting prospects who wrote first after accepting a connection request as well, 4,392 of 9,808 replied (44.8%). 75 of the 95 booked calls came from threads the agent opened; the other 20 from prospects who wrote first after a connection request.
What the first replies said
3,997 classified first replies (keyword-based)
- Something real51%
- Pleasantry or quick reply44%
- Negative3%
- Reaction3%
Reply rate by month of first message
share of prospects first messaged that month who replied (all channels; email is under 1%)
- Nov 202547%
- Dec 202558%
- Jan 202646%
- Feb 202643%
- Mar 202648%
- Apr 202655%
- May 202648%
- Jun 202635%
- Jul 202636%
- Aug 202633%
What the numbers mean
- A quick reply is a door, not a result. The agent answered pleasantries with a question about the prospect’s clients; 1,035 of 1,751 became real conversations.
- The call count is a floor. Before April 2026, calls were agreed in chat over pasted video links, outside the calendar, so none of them is counted here.
- Real conversations, not one-liners. 49.8% of replied threads ran to three or more prospect messages.
- Profiles run for individual member firms started in May 2026 and drew fewer replies. Outreach stopped on 18 September 2026.
Recruiting advisors worked best. Profiles run for single member firms did not.
The network ran three kinds of profile: advisor recruitment (free to join, paid on closed referrals), firm sales (signing up firms that serve those clients), and from May 2026 profiles for individual member firms, each pitching its own market to win advisor referrals.
By program
| Program | Messaged | Replied | Real replies | Booked |
|---|---|---|---|---|
| Advisor recruitment (free to join, paid on closed referrals) | 2,730 | 49.6% | 38.4% | 2.47% |
| Firm sales (firms that serve advisors’ clients) | 4,312 | 43.8% | 34.0% | 2.29% |
| Member-firm profiles (one firm each, May to Aug 2026) | 1,228 | 27.4% | 20.2% | 0.93% |
Agent-first threads, all time. Booked = calendar-flagged bookings among prospects first messaged 15 April to 31 August 2026, for comparison only. One more profile (935 messaged) switched from advisor recruitment to firm sales in February 2026.
Booking rate by the prospect’s profession
prospects first messaged 15 Apr to 31 Aug 2026 (4,467), from their LinkedIn headline; calendar-flagged bookings, for comparison
- Capital, investment2.81% (38.8% replied)
- Immigration, golden visa2.58% (47.4% replied)
- Private banking, wealth2.39% (43.5% replied)
- Real estate1.90%
- Legal, tax1.45%
Reply rate by program
same prospects, first messaged 15 Apr to 31 Aug 2026
- Advisor recruitment49.3%
- Firm sales39.6%
- Member-firm profiles28.3%
- A free offer opened doors. Advisors were asked to join free and be paid on closed referrals, and the cost question became the best-converting objection: 14 bookings from about 75 threads.
- Member-firm profiles often met the wrong advisors. Cyprus real estate produced no external meeting from 66 prospects, Greece real estate none, Greece hotel investment one (two counting a prospect who wrote first). One answer: “[...] I am a portfolio manager i only deal with financial assets”
Follow-ups added 49% more replies. From message four, they cost goodwill.
Counting only replies to the first message, the reply rate would have been 29.0%. With follow-ups about five days apart it reached 43.3%: 1,325 of 4,014 replies came after a follow-up. From the fourth message on, 16% of replies were negative.
Reply rate at each touch
of prospects who got that touch without having replied yet
- Intro29.0% (2,689/9,270)
- Follow-up 112.1% (722/5,960)
- Follow-up 26.1% (291/4,746)
- Follow-up 36.1% (216/3,522)
- Follow-up 45.0% (73/1,467)
Which touch earned the first reply
4,014 prospects who replied to outbound
- Intro67%
- Follow-up 118%
- Follow-up 27%
- Follow-up 3+8%
Did follow-ups annoy people?
| Touch | Recipients | Replies | Negative first replies |
|---|---|---|---|
| Intro | 9,270 | 2,689 | 10 (0.4%) |
| Follow-up 1 | 5,960 | 722 | 20 (2.8%) |
| Follow-up 2 | 4,746 | 291 | 28 (9.6%) |
| Follow-up 3+ | 3,522 | 312 | 50 (16.0%) |
Negative = the first reply was negative or asked to stop (keyword match).
The follow-up that mattered most came after the link. Many prospects agreed to a call, got the booking link, and went quiet. A short nudge (“Did you find a moment to check the calendar link, or would you prefer I suggest some new times?”) brought them back: 48 of 111 booking threads with a link booked only after at least one nudge, and 24 bookings came more than 14 days after the prospect’s first reply.
of 116 booking threads began with a reply to a follow-up, not the opener
median gap between the first and second follow-up
replied to the fifth and sixth follow-ups of earlier, longer sequences (11 of 373, 6 of 190)
Answers in minutes, while the prospect is still on LinkedIn
Prospects replied a median 7 hours after the first message. When they did, the agent answered in a median 2.5 minutes across 12,621 replies. Replies typed by the network’s team took a median 2.7 hours.
median agent reply over the whole window; 60% within 5 minutes
median since 21 May 2026 (two-thirds under 5 minutes)
median for replies typed by the team
of replied threads opened after 21 May 2026 had no message tagged as typed by a person (974 of 1,192); in 43% every message was tagged as the agent’s
From the prospect’s first reply to a booked call
116 external booking threads
- Within 1 hour25 of 116
- Within 24 hours51 of 116
- After 14+ days24 of 116
A migration agent in Australia went from “Thank you for connecting” to a booked call in 13 minutes. When a prospect wrote first, usually after reading the connection note, the agent answered 615 of 799 such threads with a median first response of 4.1 minutes.
Speed was not the weak point; operations were. In about 14 booked calls the prospect joined and found no host. The agent books the call; a person still has to host it.
Six patterns behind the booked calls
Counts are of the 116 external threads with a calendar booking, each read in full, except pattern 6, which compares all prospects first messaged from mid-April. A thread can show more than one pattern.
Turn “thanks for connecting” into a question
The agent answered quick replies with a question about the prospect’s own clients: “When they ask about second passports or international residency, how do you typically handle those conversations?”
39 booking threads the agent opened, two in five, began with a pleasantry.
Let the prospect ask for the call
Advisors often proposed the call themselves, before the agent asked. The agent answered in about two minutes with two named days or a link.
55 of 116 booking threads.
Send the link, then nudge
A link alone rarely closed. A short follow-up offering “the link or new times” did, often weeks later, when the prospect resurfaced.
48 of 111 linked threads booked only after a nudge.
Accept the reciprocity frame
Advisors pitched back or proposed mutual referrals. The agent took the frame and moved to a call: “I’m happy to explore how our client needs might align for potential referrals.”
40 booking threads use the prospect’s own collaboration language.
Hand the call to a person
The agent booked; the network’s team took it from there. Staff named “my business partner” as the host, sent video links and handled reschedules.
25 booking threads hand the call to a named colleague.
Low-pressure openers beat questions
From mid-April, openers that only complimented the prospect’s work and asked to connect beat openers built on a question.
27.5% vs 20.2% replied to the opener; 2.10% vs 1.11% booked (3,477 vs 990 prospects, observational).
“Sounds like we do the same for our respective counties [...]. Lets keep in touch then and maybe we can form a reciprocal introduction relationship between us.”
“[...] pleasure connecting with you. Let’s schedule a call to understand a bit more”
“Sounds like a deal, should we book a meeting ?”
Inside the conversations
Three real threads, condensed. Prospect names are changed and the people behind the network’s profiles are described by role; otherwise the words are verbatim.
AI illustration“Lovely to connect. Believe the best way to know more would be to have a video call.”
Director at a private bank
Migration agent, Australia · agent for a member firm in UAE investment migration
From “thanks for connecting” to a booked call in 13 minutes
- Agent (connection note)Day 0Hi Dana, I saw your profile as a Registered Migration Agent in Australia. I'm also in the immigration advisory space and would be interested in connecting.
Peer framing in the note. No pitch.
- Dana, migration agentDay 11Thank you for connecting [name]!
A pleasantry, 11 days later.
- Agent+1 minGreat to connect, Dana. As a Registered Migration Agent, you likely have clients exploring options beyond Australia. When they ask about second passports or international residency, how do you typically handle those conversations?
Turns the pleasantry into a question about her clients.
- Dana, migration agent+4 minI can only give advice regarding Australia, so If they are no longer interested or do not have the requirements for Australia I just let them go
She names the gap herself.
- Agent+2 minThanks, Dana. Many advisors face that gap. [The firm] partners with agents like you to guide those clients to the right second passport or residency, so they don't have to be turned away.
Her words back, and the offer in one sentence.
- Dana, migration agent+1 minsounds interesting
- Agent+2 minGreat. Would Tuesday or Wednesday next week work for a quick 15-minute call? I can walk you through how our partnership works and what we offer your clients. You can also book directly here: [booking link]
Two named days plus a self-serve link.
- Dana, migration agent+3 mingreat, I've done it, see you on Tuesday!
Booked.
- Agent+2 minPerfect. Looking forward to it.
Why it worked: The agent never pitched before she described her own problem, clients she has to turn away. It answered every message within one or two minutes, then asked for a small, specific next step the moment she showed interest.
Senior wealth manager, global private bank · advisor-recruitment profile
A compliance boundary, turned into “you can definitely count on me”
- Rui, senior wealth managerDay 0Great to hear and happy to connect! What's your experience working with Portuguese or international UHNWIs looking to relocate to Portugal [...]?
Replies within 3 hours to a peer-to-peer opener with no ask, and tests the sender.
- Agent+2 minMy experience with HNW clients often involves cross-border solutions, including Golden Visas and relocation. How do you manage introductions for clients needing specialized advice in these areas?
Answers from the stated background, then turns the question to his referrals.
- Rui, senior wealth manager+6 min[...] I connect them with the right specialists within my network (tax, legal, relocation, structuring,etc)
- Agent+<1 minThat's a common and effective way to support your clients. How do you ensure those introductions are always tracked and that your role is protected, especially for cross-border deals?
The protection question behind his next answer.
- Rui, senior wealth manager+5 minI work for one of the world's top banking groups, so trust, compliance and long-term relationships are at the core of everything we do. Cross-border situations are part of our daily activity and are not an issue for us… except in situations where, for regulatory or compliance reasons, we simply do not want to be involved.
Objection: we have this covered, within compliance limits.
- Agent+2 minThat makes sense. Even with robust internal processes, I've found that external introductions for specialized cross-border needs can benefit from infrastructure that formalizes and tracks these arrangements, protecting everyone involved.
Accepts the boundary and positions the offer inside it, as protection.
- Rui, senior wealth manager+4 minAbsolutely! Within the framework I mentioned, you can definitely count on me whenever there is a situation where I can add value or support the right clients. [...]
Objection turned.
- Agent+3 minIt sounds like we share a similar philosophy. [...] Would you be open to a quick 15-minute call sometime next week?
One ask.
- Rui, senior wealth managerDay 1Yes, for sure Send me two slots of your availability
Asked for two slots, got only a link (a miss). Nudged on day 6 and 13.
- Rui, senior wealth managerDay 13Just booked to next week
Booked after the second nudge. The booking email matches him.
Why it worked: The agent let a regulated banker set his own limits, framed the offer as protecting introductions he already makes, and nudged twice.
Founder, UK property advisory for private clients · agent for a member real-estate firm in Georgia
Agreed on day 2, booked on the second nudge, 23 days in
- AgentDay 0Hey Rob. With your focus on [his niche], how do you approach identifying unique international real estate opportunities for your HNW clients?
A question tied to his headline.
- Rob, founderDay 2[...] great question. For me, it's all about relationships rather than listings. [...] Discretion and trust are everything when you're dealing with HNW individuals. [...]
- Agent+2 minThanks for sharing your insights, Rob. I resonate with your focus on relationships, discretion, and working directly with developers for HNW clients. My team [...] also builds those deep connections here in Georgia to find unique opportunities for international investors. I'd be glad to connect for a brief call to explore potential synergies.
Mirrors his values, proposes a peer call.
- Rob, founder+2 min[...] I'd be happy to arrange a short call. Feel free to send over a few times that work for you next week, and we'll make something happen.
- Agent+3 minGreat, Rob! I'm glad we're aligned. How about next Tuesday, July 28th, at 10 AM or 3 PM [local] time? You can also pick a time that suits you best here: [booking link]
The two slots he asked for, plus a link.
- AgentDay 7Hey Rob, just following up on our chat about connecting for a call. Did you have a chance to look at the calendar link, or should I suggest some alternative times for our chat this week?
Nudge 1: an easier path, no pressure.
- AgentDay 23Hey Rob, I hope you had a good week. Did you find a moment to check the calendar link, or would you prefer I suggest some new times?
Nudge 2, the same low-pressure choice.
- Rob, founder+5.5 h[...] Apologies i have been traveling and also taken on new consultancy roles. I will check your calendar link and book a meeting. Very best, Rob
He resurfaces the same day.
- Rob, founder+4 min[...] I just booked for this Friday, you should have got a notification. Look forward to chatting. Very best, Rob
Booked. He later postponed and promised to rebook.
Why it worked: The interest was real, but he went quiet for three weeks. Two short, polite nudges, each offering “the link or new times”, brought him back the day he resurfaced. 48 of 111 linked booking threads needed at least one such nudge.
How the agent handled what wealth professionals push back on
Thread counts are approximate (keyword-matched across all replied threads). Quotes are the agent’s own words; names are changed.
| Objection | Threads | What the agent said | What happened |
|---|---|---|---|
| “What exactly is this?” | ~168 | “My apologies for not being clear, Gabriel. I meant that in global mobility and international recruitment, finding vetted local partners for your clients can be challenging. We help professionals connect with trusted advisors for those cross-border needs. [...]” | Plain, full-sentence answers kept these threads alive (10 bookings across them); fragments like “Earn on referrals.” lost them. This prospect did not reply. |
| Cost and fees | ~75 | “Absolutely, Neil. For advisors, it’s completely free, forever. No upfront costs. Would you be open to a quick chat to explore further?” | The best-converting objection: 14 bookings. Deferring firm pricing to a call annoyed some who asked twice. |
| Already covered, or limited by the employer | ~82 / ~11 | “Even with robust internal processes, I’ve found that external introductions for specialized cross-border needs can benefit from infrastructure that formalizes and tracks these arrangements, protecting everyone involved.” | Usually a polite exit; most bank-employed advisors declined. Reframed as protection, it booked the private-banker conversation above. |
| Timing: travel, Eid, Easter | ~273 | “Farid, I completely understand. Enjoy your Eid celebrations. I’ll follow up with you next week to arrange a time for our discussion.” | The most bookings of any objection: 31. The agent came back when it said it would. |
| “Not my area”, “my clients go elsewhere” | ~69 | Usually thanked them and stepped back; sometimes it kept probing after a clear no, which did not help. | Rarely converted (4 bookings); the main way member-firm threads ended. |
negative first replies to 9,270 opening messages
scored agent replies in the prospect’s language; it also wrote Portuguese, Spanish and Greek when prospects did
bookings after a timing objection: acknowledge it, then come back when promised
“Since my work is highly relationship and network driven within the luxury real estate sector, I definitely understand the importance of trusted referrals and structured partnerships.”
What did not work
- Member-firm profiles. 27.4% replied and under 1% booked; Cyprus and Greece real estate produced essentially no external meetings.
- Too many follow-ups. From the fourth message on, 16% of replies were negative.
- One-line replies. 6.8% of agent replies were three words or fewer (“How so?”), and prospects noticed. Full sentences won confused prospects back.
- Holding back the price. Firms that asked for the price and were told it would be discussed on a call were irritated.
- Unstaffed calls. About 14 prospects waited in a booked call that nobody joined. The booking is only half the job.
- Language slips. One profile answered some English-writing prospects in Russian.
“[...] the way you have kept taking followbacks is the only reason I’m replying right now, kudos for your effort. But [...] such highly chatgpt’d message structures will dissuade people from responding.”
What made it work
- 1
Choose the audience before the copy. Advisor recruitment to private bankers and golden-visa advisors replied at 49.6%; profiles run for individual member firms replied at 27.4%.
- 2
Treat “thanks for connecting” as the start. 44% of first replies were pleasantries. A question about the prospect’s clients turned 59% of them into real conversations.
- 3
Answer in minutes. The agent’s median reply was 2.5 minutes, and 25 booking threads closed within an hour of the first reply.
- 4
Send the link, then follow up on it. 48 of 111 linked threads booked only after a nudge, and 24 bookings came more than two weeks after the first reply.
- 5
Keep the ask small and peer-to-peer. A 15-minute call on “Tuesday or Wednesday”, handed to a partner from the network.
- 6
Come back when you said you would. Timing was the most common objection (about 273 threads) and produced the most bookings: 31 followed one.
How we measured this
- Source: Sellinger production data for the client’s workspace, read-only, snapshot 10 October 2026. LinkedIn, 14 November 2025 to 18 September 2026; outreach has been stopped since.
- Messaged = received at least one message after connecting (9,808). Replied = wrote back after our first message, auto-replies excluded; LinkedIn quick replies count as replies.
- Replied with substance = at least one reply that is more than a pleasantry, reaction or quick reply (keyword-based classification), measured on the 9,264 threads the agent opened.
- Touch = the n-th message before the first reply. Reply rate per touch = of people who got that touch without having replied, the share who replied before the next one (base 9,270).
- Response time = prospect message to the agent’s next message; replies typed by the team are excluded from agent figures. Messages carry an AI or human tag only from 21 May 2026.
- Booked = the prospect booked a slot on a calendar connected to Sellinger. The calendar was connected in April 2026, so calls exist only for April to August 2026; calls agreed in chat before then are not counted.
- 117 leads carry a calendar booking; one is an internal test, leaving 116 external threads, all read in full. 95 are corroborated: in 66 the booking email matches the prospect; in 29 the prospect says in the thread that they booked, but the platform’s booking matcher had attached another person’s booking to the lead. The other 21 are not counted. 75 of the 95 came from threads the agent opened, 20 from prospects who wrote first.
- Attendance was not recorded (115 of 117 bookings have no outcome), so this study reports calls booked, not calls held. It does not claim deals closed, members won or revenue.
- Pattern counts use the 116 external booking threads. Booking rates by program and profession use all calendar-flagged bookings for prospects first messaged 15 April to 31 August 2026, and are for comparison only.
- Reply types, objections and patterns are keyword-based classifications, checked by reading the threads. Transcripts are condensed: some turns are left out, cuts inside messages are marked [...], and emoji are dropped.
- The client, its team and its member firms are described by role, not named. Prospect names are changed and companies are described.
- Images are AI illustrations, not photos of the client, its members or its prospects.
More case studies

Lamothermic
One in three engineers replied to a foundry president’s AI agent

Green Shield Logistics
18 days, one letter, a meeting and an RFQ for 81 pallets

Uberwood Agency
Three in ten local business owners wrote back

ShortVibe
A solo founder booked 19 discovery calls while still building his products

Sellinger
We sell Sellinger with Sellinger: a year of our own LinkedIn outbound
Want numbers like these for your pipeline?
Book a 15-minute walkthrough of the agent, the playbook and the numbers behind these case studies.

