``You're approaching the wrong people.'' This is the biggest cause of decreased sales productivity. No matter how good your sequence is, no matter how polished your talk script is,Approaching someone who doesn’t fit your ICP is a waste of time and energy.It's just that. In this chapter, we will systematically explain the technique of ``approaching the right person, at the right time, and with the least amount of effort'' using HubSpot's Prospecting Agent, Buyer Intent, and lead scoring.
The starting point of prospecting ICP（Ideal Customer Profile） It is the verbalization of ICP refers to the image of a customer who can deliver the most value to a company's products and services, has a high order probability, and has a high LTV (lifetime customer value). ``Selling to anyone'' is synonymous with ``not hurting anyone.'' Effective prospecting is impossible without a clear ICP.
Once you have defined your ICP, you can filter your company records in HubSpot.Target company listCreate. From "Add filter" on the company list screen, combine multiple conditions such as industry, number of employees, lead status, tags, etc. and save. The saved list is always kept up to date as an ``active list,'' and new companies that meet the criteria are automatically added.
| Examples of ICP conditions | Filter settings in HubSpot | Usage properties |
|---|---|---|
| Number of employees: 50-500 | Number of Employees is between 50 and 500 | Number of Employees (Standard) |
| SaaS/IT industry | Industry is any of "Internet Software and Services", "Computer Software" | Industry (standard) |
| Currently using Salesforce | Technologies contains "Salesforce" | Technologies（Breeze Intelligence） |
| recently raised funds | Recent Funding Event is known (after 2025) | Recent Funding（Breeze Intelligence） |
| ICP Tier A configured | ICP Tier is "Tier 1" | ICP Tier (custom properties) |
HubSpot Breeze Intelligence(formerly Insights, purchased separately with credits) allows you to automatically fill in blank properties in contact and company records with external data. It automatically fills in information such as number of employees, sales, industry, technology stack, LinkedIn URL, etc., greatly improving the quality of imported lists. Consumes 1 credit for enriching 1 record.
Breeze Prospecting Agent is an AI sales agent that HubSpot began full-scale operation in the spring of 2025. It autonomously performs tasks that take up the most time for salespeople, such as researching prospects and generating and sending personalized emails. From September 2025, a credit consumption system will be introduced (1 action = approximately 100 credits), making it necessary to manage costs according to the frequency of use.
The output quality of Prospecting Agent is Selling Profile This is directly connected to the setting accuracy. The Selling Profile is a collection of prompts that teach the AI ``who you are, who you are, and what value you provide.'' Settings can be made from "Settings → AI → Selling Profile".
①Brush up your Selling Profile monthly——Check the response rate of emails actually sent and continue to modify ineffective value propositions. ② Be sure to use “Send after review” mode at first.——Understand the quality of AI production before expanding automation. ③ Control the number of actions per day——Credit consumption is high, so use it only for high-priority contacts and deal with the rest in sequence.
The most difficult aspect of prospecting is timing. Even for companies that meet the ICP, the response rate is extremely low unless now is the time to consider it.Buyer Intent SignalBy using this, you will be able to preferentially approach companies that are about to start considering them.
Don't just "see" the signals;Use as a trigger for automated actionsThis is how we demonstrate our true value. With the December 2025 update, you can now use these signals as workflow triggers.
Lead scoring is a system that quantifies how likely this contact is to turn into a business negotiation. At HubSpotManual scoring (rule-based)andPredictive scoring with AI (Enterprise only)Two types are available: Each has different characteristics, and it is important to use them properly.
Manual scoring, available for Professional and above, is a format in which you set ``rules for adding and subtracting points for specific actions and attributes.'' Configure settings from "Contacts → Scoring → HubSpot Score".
| manual scoring | AI predictive scoring | |
|---|---|---|
| Usage plan | Professional or above | Enterprise only |
| structure | Add and subtract points based on rules set by humans | Machine learning of past order/loss data and scoring the probability of winning an order from 0 to 100 |
| Setting time | systematize | text |
| Strengths | Highly transparent (you can see why the score is given) / Easy to adjust rules | Discover correlations that humans are unaware of; accuracy improves as more data accumulates |
| weakness | Incurs designer bias/inaccuracy when there is little data | Difficult to understand "why this score" - Does not work if there is little data |
| Recommended use | In the early stages of implementation, while data is being accumulated, and when the rules are clear | Mutual complementation in combination with stages and manual scores when sufficient data has been accumulated |
Even if you create a scoring model, it won't work unless you decide how many points to pass for sales as an MQL. At firstThreshold = 60 pointsStart operation with a temporary setting, and then check the "business negotiation rate of leads passed with a score of 60 or higher" every three months and adjust the threshold. Ultimately, the goal is to automate things up to ``automatic notification and automatic task creation in the workflow the moment a threshold is exceeded.''
Once you have the ICP, intent signals, and scoring, the next step is to prioritize who to approach first each morning.Sales Workspace Prospecting Viewis a dedicated UI for managing this "today's priority list" on a single screen.
| time zone | action | HubSpot features |
|---|---|---|
| First thing in the morning (15 minutes) | Check out Intent signals for today's Tier 1 targets. Check the changes since yesterday (funding, site visits, rapid increase in recruitment) | Sales Workspace → Target Accounts tab → Intent signal feed |
| Morning (30-60 minutes) | cause | Sales Workspace → Prospecting Tab → Agent Review Queue |
| Afternoon (30 minutes) | Sequence confirmation and follow-up call execution to Tier 2 and 3 of task queue. Make a one-click call from HubSpot | Sales Workspace → Today tab → Task queue (call/email) |
| Evening (10 minutes) | Check the day's outreach results. Check replies, opens, and clicks, and follow companies that respond with top priority. | Sequences Tab → Engagement View / CRM Timeline |
``Making high-quality approaches to 10 Tier 1 emails a day'' will often result in more business negotiations than ``sending 100 emails a day.'' A high-quality personalization approach with Prospecting Agent reduces the response rate of generic templates.3-5x response rateIn some cases, it is recorded. Rather than maximizing the number of outreach“Maximizing the contact rate to Tier 1”Let's set it as a KPI.
Technical evaluation (PoC/demonstration added)
The output quality of the AI agent is directly linked to the accuracy of the Selling Profile settings. Review value propositions, target buyers, and points of differentiation on a monthly basis to keep them up-to-date and of the highest quality.
By setting four signals as workflow triggers: funding, recruitment trends, technology changes, and site visits, we automatically detect and automatically approach the moment when the intention to buy increases.
Manual scoring is effective in the early stages of implementation and when transparency is important. AI predictive scoring is performed on Enterprise, where data is accumulated. In both cases, it is necessary to set a threshold for ``how many points are needed to pass sales.''
Concentrate most of the man-hours on Tier 1 (ICP complete match + high intent), and leave Tiers 2 and 3 to automated sequences. The KPI is to maximize the contact rate to Tier 1 rather than volume.
Prospecting becomes a habit by centrally managing the daily routine of checking signals first thing in the morning → reviewing agents in the morning → digesting the task queue in the afternoon → checking replies in the evening.