How to distribute enquiries between agents: three models and dispute-free rules

In an agency where leads are taken on a “whoever sees it first” basis, the same picture emerges within a couple of months: strong agents are overloaded and reply late, newcomers sit without clients, and the team meeting turns into arguments about whose client it is. Distributing leads isn’t bureaucracy - it’s a way to answer every client quickly and without conflict inside the team. Let’s look at three working models and the rules worth writing down once.
Why “first come, first served” doesn’t work
Free-for-all lead picking looks fair but rewards the wrong behaviour. Agents choose the “juicy” leads - a big budget, a clear request - and ignore the rest. Whoever spends more time in the team chat grabs more than they can handle. Nobody takes night-time and weekend inquiries. And when a client gets in touch again, two agents are each sure the client is theirs.
- An overloaded agent replies hours later - and the client goes to whoever answered sooner.
- Newcomers get no practice and leave the agency before they’ve really started.
- Inconvenient leads go unanswered, even though each one was paid for with advertising.
- Disputes over clients take up the manager’s time and sour the atmosphere.
Three distribution models
There’s no universal model: it depends on the team’s size, how specialised the agents are and where leads come from. Most agencies end up with a combination of two models.
Scroll the table sideways
| Model | How it works | Strength | Risk |
|---|---|---|---|
| Round robin | How it worksEach new lead goes to the next agent in turn | StrengthSimple, transparent, everyone gets an equal share | RiskIgnores workload and experience: a complex request may go to a newcomer |
| By specialisation | How it worksDistrict, segment, rental or sale, the client’s language | StrengthThe client goes straight to someone who knows their market | RiskImbalance: in a “hot” district the agent drowns, in a quiet one they idle |
| By workload | How it worksThe lead goes to whoever has the fewest active clients | StrengthNobody is overloaded, replies come faster | RiskRequires honest tracking of active clients, otherwise the model lies |
| Combination | How it worksSpecialisation first, then round robin or workload within the group | StrengthA balance of market knowledge and speed | RiskMore complex rules - they must be written down and shown to everyone |
Whichever model you choose, what decides the outcome isn’t the model but the rule that a lead can never be left without an owner.

A lead’s route: from arrival to owner
The model answers “to whom”. The route answers “what happens to the lead every minute”. Without a route, any model breaks on the first lead that arrives on a Saturday evening.
A lead’s path through the agency
A working scheme with checkpoints
The lead lands in a shared queueImmediately
A call, Telegram, the website, a listing site - everything in one place, not on personal phones.
A rule assigns an ownerAutomatically
In turn, by specialisation or by workload - as agreed in the agency.
The agent confirms they’ve taken itWithin the deadline
Not taken in the agreed time - the lead moves to the next agent, and the manager sees the reassignment.
First reply to the clientWithin the deadline
A real answer to the point, not “hello, please clarify”.
The request is recorded on the cardAfter the conversation
Budget, district, timing, source of money - so that matching and colleagues work with facts.
A next step is setWith a date
A viewing, a call, a shortlist - the lead always has a next task.
The key step here is the third - reassignment. If the assigned agent hasn’t confirmed the lead within the agreed time, it doesn’t wait for them to get back from a meeting; it moves to the next agent. It’s not a punishment but insurance for the client. Such a route can’t be run through personal messengers; it needs the agency’s Chats, where Telegram conversations go through the agency bot, are tied to the client card and the manager sees the team’s conversations.
Rules that defuse disputes
Most conflicts over clients repeat themselves. It’s easier to settle them once in writing than to sort them out at every team meeting.
- A client is assigned to the agent who first gave them a real answer and recorded the request.
- A repeat inquiry from the same client goes to their agent, not to the next in the queue.
- A client who came to a specific agent through a referral is that agent’s client, outside the queue.
- If the agent doesn’t contact the client for the agreed number of days, the client returns to the queue.
- During holidays and sick leave, an agent’s leads go to a pre-assigned stand-in.
- When an agent leaves, their clients and conversation history stay with the agency.
The first-reply deadline
A distribution model is pointless if, once assigned, a lead sits until evening. The first-reply deadline is the most important number in the whole scheme. It’s set for working hours and separately for nights and weekends: for example, night-time leads get an automatic message and a call first thing in the morning. We covered why an early reply is worth so much, with open research, in the article on response speed.
The model decides who gets the lead. The reply deadline decides whether there will be anyone to give it to next time.
The usual way
Leads are self-assigned
- Whoever saw it in the chat first gets the client
- Night-time and awkward leads wait until morning or longer
- A repeat inquiry goes to a different agent
- Disputes are settled at the meeting from memory
How it can be
A queue with rules
- A rule assigns the owner straight away
- Not taken in time - the lead goes to the next agent
- The client is assigned, the history is shared
- Fewer disputes: they were settled in advance
How to launch the rules in a week
New distribution rules fail not because of the model but because of a launch along the lines of “from Monday we all work the new way” with no explanation. A week of calm preparation pays off.
- List every channel leads come from and decide where they will land.
- Choose a model and write the rules on one page - no more.
- Discuss the rules with the team before launch and answer “what happens to my clients” honestly.
- Move current clients over: each gets an owning agent, no exceptions.
- Launch, and for the first two weeks review every disputed case at the team meeting.
- After a month, look at the numbers and adjust the rules, not the people.
What agents will say - and how to answer
Resistance to new rules is predictable, and it’s almost always about the fear of losing what’s yours. A strong agent fears the queue will hand “their” leads to newcomers. A newcomer fears experienced agents will grab the complex requests anyway. The answer is the same for everyone: the rules protect each of them. A strong agent keeps their clients and referrals, a newcomer gets a guaranteed flow, and the “whose client” argument is no longer won by whoever is loudest at the meeting. If the model takes specialisation and workload into account, experienced agents get complex leads not by seniority but because that’s faster for the client.
How to tell distribution is working
A month after launch it’s useful to look at a few numbers for each agent. Not to punish anyone, but to see where the model is skewed.
- Median time to first reply - by agent and by time of day.
- How many leads were reassigned, and whose more often.
- Share of leads with a completed client request.
- Share of leads that reached a viewing - by agent and by source.
- Active clients per agent: is any one person drowning.
If the numbers have to be pulled together by hand from spreadsheets, people stop looking at them within a month. In agency analytics the funnel by stage, deals and commission by month and where clients come from are calculated from working data automatically; what isn’t there, it’s enough to count by hand once a week at first. We describe how else team work fits into one system on the page for agencies.
This article contains no platform statistics: the models and rules are agency management practice. Each agency sets its own reply deadlines and client assignment rules.


