How to find the hot enquiry among the junk

Why hot enquiries drown among the junk ones

When there are dozens of enquiries a day, a salesperson physically cannot go through each one at once. They open them in the order they arrived rather than by importance, because the order is the only thing visible in the general list. While they dig through obviously empty enquiries, right next to them lies one from a person ready to buy right now.

The problem is not laziness or inattention. The problem is that the salesperson physically has no focus left for sorting through dozens of rows, each of which looks the same until you open it. The most valuable enquiry simply gets scrolled past with the rest — not because it went unnoticed, but because there was no time.

The outcome is predictable: the hot lead cools down while its turn comes, and goes where the reply was faster. Meanwhile in the CRM it will have an overdue task, exactly like the dozen junk enquiries next to it. From the outside this looks like an ordinary working day in the sales department, but in fact it is lost money.

What has to happen in the first minutes after an enquiry

Until an enquiry is labelled, the salesperson sees just a row in a list, with no idea whether to open it first. The first task is to make sure the labelling happens before the enquiry reaches the salesperson at all, not after. That means the reaction has to be automatic, with no human involved at the first step.

In the first minutes it should not be a person who acts but the system: it replies to the client, asks clarifying questions and understands from the answers what this is actually about. It is at this stage that it is decided whether the enquiry becomes a priority for the salesperson or goes into the general stream for later. The later this happens, the higher the risk that a hot lead simply gets lost among the rest.

While this initial processing goes on, the salesperson deals with what has already been selected and passed to them as important. They do not spend time working out whether the next enquiry is worth opening at all — that has already been done for them.

A qualifying robot is the first filter before the salesperson

Sales robot replies to the client in every channel where enquiries arrive: Telegram, WhatsApp, Instagram, Avito, the website. It recognises voice messages and photos, remembers the conversation history and does not start the dialogue over at every new message from the same person. This lets it talk to the client the way a salesperson would — but straight away, without delay.

Its task at this stage is not to sell but to understand: what the person needs, how ready they are to move on, whether the enquiry should be passed to a salesperson right now at all. Based on the conversation it assigns the enquiry a status — hot, warm or cold — and only after that does a hot client reach a salesperson.

In the live GetGate project 27% of enquiries arrive in the evening and at weekends — they used to be answered only the next morning. With the robot the reply goes out immediately, whatever the hour, and the enquiry does not cool down while waiting for the working day to start.

Scoring and routing in amoCRM: how it is set up

The “hot / warm / cold” labelling is not an abstract system in the CRM but specific fields and tags in the deal card. Each attribute — the source of the enquiry, the type of request, the client's readiness to reply — gets its own field, and the digital pipeline moves the deal through the stages automatically, without dragging cards by hand.

A webhook passes the data from a chat or a call into CRM the moment they appear, not when a salesperson opens the card and fills the fields in by hand. The status is set by the robot right after the conversation — before the enquiry even reaches the salesperson's task list.

The pipeline is built for the specific client's process rather than from a template. The salesperson gets an already filtered list: hot enquiries at the top, the rest as they become ready. That is the difference between a system that merely stores enquiries and one that sorts them.

How to tell a hot enquiry from a junk one

For B2B and services the trigger is usually not the fact of the enquiry but what the person says: specific deadlines, a named task, readiness to discuss terms. The phrase “tell me more” without details is one thing; “we need this sorted by the end of the week, here is the budget” is quite another, even if both came from the same website form.

A junk enquiry is a question with no substance, a duplicate of one already handled, spam or an accidental click on the form. It is not always possible to tell it from a hot one by eye, especially when there are many enquiries and they are worded alike. That is exactly why labelling cannot be trusted to the outward look of an enquiry — a conversation is needed to bring out the substance.

The criteria for “hot” and “junk” are specific to each niche and cannot be described in general phrases. They are written down together with the client when the robot is set up — as a concrete checklist the conversation follows, not as an abstract rule to “answer politely and to the point”.

What the salesperson sees once the enquiry is labelled

Instead of a list of dozens of identical rows the salesperson gets an already ranked order: hot enquiries first, warm ones next, cold ones not urgent. They do not spend time working out where to start the day — that was decided for them before they even opened the CRM.

Sales analytics shows plan against actual in real time, broken down by channel, service and salesperson. On the GetGate project, for example, five product groups are counted automatically — with no manual recalculation in spreadsheets that quickly go stale.

If in some channel enquiries stop turning into deals, an alert about the dip arrives — before it becomes a problem for the whole department. This makes it possible to notice not only a lost enquiry but also the systemic reason why such enquiries are lost.

Control: a hot enquiry must not get lost at the salesperson either

Scoring at the entrance solves only half the task. Even with correct labelling the enquiry can be lost further on — if the salesperson saw the “hot” status but did not call in time or put the call off. Here a separate loop is needed, one that checks not the enquiry but what the person did with it.

Quality control — is when AI listens to calls and reads chats, scores them against a checklist and checks whether the salesperson reacted to a hot enquiry the way they should have. This is also where blown deals are looked for — cases where an enquiry is formally handled but in substance lost.

In our check on real correspondence, in 80% of cases the assistant's reply was no worse than a human's in a blind comparison. That is not a guarantee of quality for a particular conversation but a benchmark: an automatic reply at this stage can be compared with a salesperson's work rather than assumed to be worse.

Where to start if enquiries are already being lost

The first step is to review the client's current pipeline and find at which stage exactly a hot enquiry is lost: at the entrance while nobody has replied, in the queue among the junk, or already at the salesperson who forgot to call back. That determines what to fix first.

Next the CRM is assembled for the process, the robot is trained for the specific business's niche and the necessary channels are connected — where enquiries actually arrive, not where the integration is more convenient. This is described in building a sales department — together with scripts, rules and an analysis of why clients say no.

After that the robot replies and labels enquiries, salespeople work under control, and the manager looks at plan against actual and tunes the scoring rules as they go. In the live GetGate project a salesperson frees up around an hour of working time a day — because they no longer spend it sorting junk enquiries by hand.