A sales plan in CRM rests on deals
To plan sales automatically, set a goal, record leads and deals in CRM and mark the next step for every lead. Then the plan can be compared with actual results and the forecast built from the state of the pipeline. If deal data is not kept current, an automatic calculation only hides the errors.
What CRM needs to know to build a plan
An automatic plan starts not with a report but with a clear sales process. You need to define what counts as a lead, when it becomes a deal and which event means a sale. If your people use one word for different events, comparing their results is hard. Agree on the recording rules first, then move them into CRM.
Set the goal you will compare actual results against separately. It can be revenue, the number of closed deals or any other figure the team really uses to manage the work. Do not add enquiries, bookings and payments into one metric: they describe different parts of the customer journey. For every metric, state where the data comes from.
A plan is useful when it can be tied to what the team does. If the goal is missed, the manager needs to see where the gap appeared: too few leads, deals that stalled or customers who said no. That is why CRM needs not only totals but the state of work on every lead. Otherwise the report shows the problem without helping to take it apart.
Before configuring anything, write down the questions the report has to answer. For example: which channels bring enquiries, which deals are waiting for a reply and where a rep's next step is needed. These questions will determine the record fields, the pipeline stages and the filling rules. Build the report once it is clear what decisions you will make from it.
How to choose the planning unit
A single revenue target is convenient for the manager but explains little about what happens inside the pipeline. Split the goal along the lines you actually steer: services, channels or reps. That split should reflect your process, not add rows for the sake of a detailed report. A metric nobody uses for a decision only complicates the bookkeeping.
A breakdown by channel helps separate a shortage of enquiries from trouble in handling them. A breakdown by service shows which part of the offer produces the expected result. A breakdown by rep is there to discuss workload and work on deals. When you compare people, take into account which leads and tasks were handed to them.
Decide in advance who gets credit when several people worked with the customer. Also define how to record a repeat enquiry and a deal that moved across periods. Without these rules the same result can land in different reports. Automating the calculation will not remove the ambiguity built into the source data.
Do not split the plan finer than the quality of your records allows. If the source of an enquiry is often unknown, a detailed plan by channel will look precise only on screen. Better to get records filled in properly first and check what information actually reaches CRM. Then add the breakdowns you need to plan and actual.
What to keep in the deal record
The record should help you see what is happening with the customer right now. For planning, the enquiry source, what they are interested in, the owner, the current stage and the expected deal amount matter most, provided the amount can already be stated with reason. The details should be updated as the conversation goes on, not only when the lead is created. Otherwise the forecast rests on a state that is out of date.
Do not make the rep guess the amount at the customer's first message. Until the need is clear, such a figure creates false confidence in the forecast. Better to agree at which stage the amount appears and by what rule it is refined. The same goes for the expected closing date.
The «next step» field ties the report to everyday work. It should make clear what action is required and who is responsible for it. A deal with no next step can stay open even though work on it has actually stopped. The manager needs to see such records separately from deals that are moving.
The conversation history is worth keeping next to the deal as well. When chats and enquiries are linked to CRM, it is easier to restore the context and see the reason for a delay. This matters especially when a customer is handed to another person. At AIBP, CRM implementation includes setting up pipelines for the client's process, connecting channels and training the team.
How pipeline stages affect the forecast
Pipeline stages should describe a verifiable state of the deal, not the rep's impression. A stage called «the customer is interested» leaves far too much room for interpretation. A rule that tells the employee which event allows the record to move on is far more useful. Then identical deals will not end up at different stages without explanation.
For every transition, describe the required action and its result. State what to do if the customer stopped replying, said no or came back after a pause. Lost deals cannot be kept among the active ones indefinitely for the sake of a nice forecast. The reason for a refusal helps separate lost demand from a mistake in handling the enquiry.
The forecast can be built from deal amounts and their current position in the pipeline. But the calculation is only useful when stages are updated in time. If the arrangement has changed while the record stays where it was, the report shows old expectations rather than the current picture. That is why the rules for keeping the pipeline matter more than the complexity of the formula.
When you check a forecast, do not look only at the total. Open the deals that make up a noticeable share of the expected result and check what the expectation rests on. Is there contact with the customer, is their request clear, is the next action scheduled? That is how a manager separates confirmed arrangements from assumptions.
How to connect plan, actual and forecast
The plan shows the desired result, the actual shows what has already happened, and the forecast shows the expected outcome of work in progress. These three cannot stand in for one another. An open deal is not yet a sale, even if the rep is sure of the customer. The report must make clear where a result is finished and where there is only an expectation.
Compare plan and actual in the same breakdowns you use to make decisions. If the goal is set by service, actual sales must be allocated by the same rules. If the manager discusses how channels perform, the origin of an enquiry must be kept in the record. When the recording rules do not match, a deviation becomes hard to explain.
A forecast helps you see a possible gap before it becomes the result of the period. But on its own it does not answer the question of what to do next. To decide, you need the records of specific deals, the reasons things stalled and the team's workload. The report should lead from the overall picture to the enquiries that need attention.
The AIBP sales analytics provides plan and actual in real time, breakdowns by channel, service and rep, a revenue forecast and alerts when results dip. To use such a picture for management, first define the rules for filling in CRM. Then a discussion of a deviation will rest on deals and actions rather than guesses about causes.
Which actions are worth automating
Automate above all the transfer of information that people would otherwise move by hand. An enquiry should land in the right record, and the conversation history should stay available to the owner. Tasks and reminders help you not to lose the next step. At the same time an automatic action must match your process, not replace the rep's judgement where a conversation is needed.
Check separately the path a new enquiry takes to the person responsible. If the customer wrote in one channel and continued in another, the team needs a shared context. Scattered chats make it hard to see who replied and what has already been promised. Linking the channels to CRM makes the state of an enquiry visible to everyone working on it.
The AIBP sales robot answers and qualifies customers through AmoCRM, keeps the dialogue history and hands a hot customer to a rep. It can work with enquiries from Telegram, WhatsApp, Instagram, Avito and the website. For planning, what matters is not only the speed of the reply but keeping the details of the request. The rep must see what the customer came with and at which step their help is needed.
Do not ask automation to fix unclear working rules. If your people understand lead qualification differently, the robot and the reports will inherit that discrepancy. First describe what information must be obtained and when the conversation is handed to a human. Then check whether these events are reflected in the record and available for analysis.
What to do with enquiries outside working hours
Planning easily misses the enquiries that arrive after the team has left. If they are only recorded at the next look through the mail or the chats, the picture of demand will be incomplete. In the live GetGate project, 27% of enquiries arrive in the evening and at weekends; they used to wait until morning. This observation applies to that particular project, not to every business.
Look at when enquiries arrive in your own case and what happens to them before a rep replies. What matters is not only the number of enquiries but the further path of each one. Did the customer get an answer, was their request clarified, did a record appear in CRM? This information will show whether you need to handle enquiries outside working hours.
If such enquiries are rare and call for a complex decision, the workflow can be built around the rep. If the questions repeat often, part of the opening conversation can be handed to a sales robot. It answers, clarifies the request and passes a hot customer to a person. Base the decision on the nature of the enquiries, not on the assumption that every channel needs the same scenario.
For the report it is important to keep the original time of the enquiry and everything done with it afterwards. Otherwise a lead that came in the evening can look like new work of the next day. That makes it harder to judge the workload and the reasons for delays. Check that automatic handling does not erase the contact history the rep needs.
How to use deviations from the plan
An alert about a dip is useful if it makes clear where to look. Start with the question of where the deviation appeared: fewer enquiries, deals that are not moving or too few closings. Each case calls for a different action. A blanket demand to «sell harder» does not tell your people what exactly to fix.
If there are few enquiries, check the channels and service types in the report. If enquiries come in but deals are stuck, open the records and look at the next steps. If customers are lost after the conversation, study the reasons for refusal and the quality of handling. This order turns the plan from a verdict on the team into a tool for finding a specific problem.
Quality control adds the substance of the conversation to the pipeline figures. At AIBP an AI does this by reviewing calls and chats against a checklist, looking for lost leads and scoring the reps' work. Such a review helps you see what happened in the conversation, not only at which stage the deal closed. The reasons are worth discussing together with the customer record and the history of actions.
Do not change the plan every time the actual result differs from expectations. Check the data first and find the reason for the gap. Sometimes the problem is in the way the work runs, sometimes in how the team records the result in CRM. Changing the goal without that review hides the question the report was meant to answer.
How to check that the system helps you decide
Take a few current deals and try to walk the path from the report to the action. Can you see why a deal entered the forecast and what the expected amount rests on? Can you tell who is responsible for the contact and what should happen next? If you have to look for the answer in an employee's private chat, CRM has not yet become the working basis of the plan.
Check the reverse path too: from a single enquiry to the overall picture. Does the deal land in the right channel, service and period? Do a change of owner or a repeat contact create confusion in the reports? Such checks are more useful than admiring one neat summary screen.
Ask the manager and the reps to name the decisions they make from the plan and the forecast. If the report is only needed to sum things up, it may lack information about current deals. If people do not understand the rules for moving between stages, start with the pipeline. The setup should help the real work, not demand a separate set of records for the sake of reporting.
At AIBP the work starts with a review of the sales department and a search for the places where money is lost. Then the CRM is assembled, the robot is trained for the niche and the channels are connected; after launch, plan and actual are watched and the process is refined. If you are choosing a solution, do not discuss only the look of the report at the meeting. Walk through the path of a lead and ask how a deviation and the next step will be found along it.
Frequently asked questions.
Are an AI agent and a chatbot the same thing?
No. A chatbot is given a script, an agent is given a goal. The bot follows a branch drawn in advance and gets lost at any step aside; the agent understands what is written in words and chooses what to do next itself.
Will the client realise they are not talking to a person?
Most likely yes, and passing the agent off as a human is not worth it: it comes out quickly and spoils the impression. In practice people are annoyed not by a robot but by silence.
What if the agent answers wrongly?
A properly configured agent answers from your materials and hands everything it doesn't know to a person. For the first weeks after launch you have to read real conversations and refine the answers — without that step the launch cannot be considered finished.
Does an AI agent need a CRM?
Formally no, it works without one. But then most of the value is lost: the correspondence stays in the messenger, no history accumulates and there is nothing to measure the result with.
Will an AI agent replace salespeople?
No, and there is usually no such goal. It takes over the first contact and the routine so that a person can deal with what needs a person: negotiation, difficult cases and money.
How much does an AI agent cost?
The price depends on the number of channels, the complexity of the scenario and the integrations needed. A sensible order of operations is to first count how many enquiries are being lost now, and decide on payback with that figure in hand.
See how it works.
Write a couple of lines about your business to our agent — it will reply and book you a free review. That is the demonstration.