What Is Signal-Based Selling?

Signal-based selling is a way of prospecting where you contact a person because something changed at their company, not because they sit on a list. This page defines the term, walks through a worked example, and shows which signals LeadOcean holds as filters.

Key takeaways

• A signal is an observable event or fact that makes a purchase more likely. Funding, hiring, a new tool and a job change are common ones.
• Signals decide who to contact and when. They do not replace a good list or a good message.
• LeadOcean holds static signals as filters: funding year, detected technology, headcount, title and more. It does not push alerts.
• On 2026-10-01, US VPs of Sales at companies with 50 to 500 employees that run HubSpot: 7,682 mailable people.

What it is

Signal-based selling is a prospecting method where outreach starts from a buyer signal, an event or fact that suggests a company has a need right now.

Traditional outbound picks a segment and emails everyone in it on a schedule. Signal-based outbound picks the same segment, then narrows it to the accounts where something just happened.

The signal does two jobs. It filters the list down to better-timed accounts, and it gives the first line of the message a reason to exist.

How it works

The method is a loop of five steps. You can run it by hand or wire it into a script or an agent.

1. Define the ideal customer. Fix the title, function, company size and country first. Signals narrow a good segment. They do not rescue a bad one.
2. Pick the signals that predict your sale. A payroll tool cares about headcount growth. A security tool cares about a new cloud vendor. Choose two or three, not twenty.
3. Detect them. Some signals are fields you can filter on. Others, like a job change or a press mention, come from monitoring feeds or manual research.
4. Find the right person. Match the signal to the buyer. A funding round points to the founder or a VP of Sales, not the whole company.
5. Send a message that names the signal. "I saw you raised a round last year" beats a generic opener. Then measure replies per signal and drop the ones that do not convert.

Types of signal

Signals fall into three groups. Knowing which group you are using tells you how much to trust it.

• Company facts: funding stage, headcount, founding year and detected technology. These are stored fields, so you can filter on them and rerun the same query.
• Person events: a new role, a promotion, a post on LinkedIn. These are fresh but need monitoring, and they go stale in weeks.
• Behavior: a site visit, a content download, a pricing-page view. These are the strongest, but you need your own tracking to see them.

Most small teams start with company facts because they are cheap to query and easy to repeat. Add person events once the first loop works.

A worked example

Say your company, Acme (a placeholder), sells sales-engagement software, and your best accounts run HubSpot. You want VPs of Sales at mid-sized US companies.

The technology signal cuts the list to under a third of its size. Every person left has a company with a reason to be interested. Both counts were run on 2026-10-01, are free, and use the default mailable statuses (verified, catch_all_valid, catch_all).

Signal-based selling vs account-based selling

People mix these up because both start from accounts. They answer different questions.

The two work together. Build the target list the account-based way, then use signals to decide which accounts to contact this month. See account-based selling and buying signals for the definitions.

When it matters

Cold outbound with a small team

A small team cannot email 20,000 people with care. Signals let you pick the 500 who are most likely to reply, and write one specific line to each.

Selling a product tied to a company event

Some products sell on events. Funding rounds, a new hire in a role, or a move to a new platform all open a buying window. If your sale depends on one of these, signals are the whole method.

Running an agent or a script

An AI agent that searches, scores and drafts needs structured inputs. Fields like funding year and detected technology are easy to filter. Vague intent scores are not.

When your list is already good but replies are low

If the right people already get your email and few answer, timing is the likely gap. Add a signal filter before you rewrite the copy.

How LeadOcean handles it

LeadOcean exposes signals as filters on the people and company search endpoints, not as a feed. Three are useful for signal-based selling today.

• Funding: fundingType and minLastFundingYear find companies by their last round.
• Technology: technologies matches tools detected on the company website. technologyCategories matches what a tool is for, such as Analytics. Both use LeadOcean's own crawl data.
• Size and role: minEmployees, maxEmployees, title, jobLevel and jobFunction pick the buyer.

Funding filters work in search but not in free counts, so size the audience without them. The real list will be narrower than the count.

LeadOcean does not send alerts when a signal fires. It also has no job-change or hiring feed. The dataset is refreshed monthly, and each record carries its own fetched_at date, so rerun your search on a schedule to catch new matches.

Send count=true as a query parameter with limit 1 to size a segment for free. Filters go in the JSON body.

curl -X POST "https://api.leadocean.io/v1/people/search?count=true" \
  -H "x-api-key: $LEADOCEAN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "jobLevel": ["VP"],
    "jobFunction": ["Sales & Business Development"],
    "country": ["US"],
    "minEmployees": 50,
    "maxEmployees": 500,
    "technologies": ["HubSpot"],
    "emailStatus": ["verified", "catch_all_valid", "catch_all"],
    "limit": 1
  }'

The response returns an empty data array and meta.countOnly set to true. Remove count=true and raise limit to pull rows. Each person returned counts one record.

To move a segment into a spreadsheet, use POST /v1/exports or the Exports page in app.leadocean.io. The page has a filter builder, a column picker, the record price before you start, and a CSV download. Free gives you 1,000 records, one-off, no card. Pro is $499 a month, flat. See pricing.

If your list is built around personas, read persona-based marketing next. More guides live in the blog.

FAQ