Toshad Sambhare- Every account on a list shows at least one live buying signal. Two or more puts it in Tier 1.
- We rate each signal on volume, effort and whether it tells us when to reach out, or only who.
- We find the first 20 accounts by hand, then automate what worked.
Fit tells you who. Signals tell you when.
Most lists are built on fit: industry, company size, location, job title. Fit is necessary, and it isn't enough. A company can match your ICP perfectly and have zero reason to talk to you this quarter.
A signal is evidence that something changed. A new leader, a hiring spree, a funding round, a bad review of the tool they use today. Something that makes the problem you solve more urgent right now. That's what we build lists on.
The difference shows up in replies. Accounts with two or three live signals usually reply at 7 to 8%. One signal, 2 to 3%. A plain ICP list, under 1%. (Typical ranges from our campaigns, not a promise.)
Start with the job your product does
Before we look at a single data source, we write down the job the product does for the buyer, in plain words. Then we ask what would make that job harder, more expensive or more visible this month.
For a payroll product, the job is paying people correctly. It gets harder when a company hires fast, opens a new state or still runs payroll in Excel. For an ERP partner, the job is closing the books. It gets harder with a second entity, a new plant or a board that wants cleaner numbers after a funding round.
Each of those is a signal we can look for. Most markets give us ten or fifteen candidates. Our library has more than 75. The work is picking the few that matter.
Rate every signal on three things
Every candidate signal gets three questions:
- Volume. How many accounts show it each month, realistically, in the client's market?
- Effort. What does it take to find and keep finding? A filter, a scraper, a manual check?
- Timing. Does it tell us when to reach out, or only who to reach?
Here's how that looked in one plan we wrote for a manufacturing-focused client:
| Signal | Volume a month | Effort | Says when? |
|---|---|---|---|
| Relevant job posts | 100 to 200, pulled weekly | Low, a scraper on job boards | Yes |
| Leadership moves | 40 to 80 | Almost none, a Sales Navigator job-change filter | Yes, the cleanest opener on the list |
| Expansion news and filings | 20 to 40 | Medium, news and company filings | Yes |
| Followers of a competitor | Large, one-off | Low | No. It says who, not when |
(Volumes from one client plan. Yours will differ.)
Signals that only say who, like competitor followers, aren't wasted. We use them to raise the score of an account that already shows a timing signal.
Signal scorecard| Signal | Volume | Effort | Tells us |
|---|---|---|---|
| Hiring for a role you replace | High | Low | When |
| New leader, last 90 days | Medium | Low | When |
| Bad review of a rival | Low | Medium | When |
| Company size and industry | High | Low | Who only |
Stack them, then tier the accounts
One signal can be noise. A job post might be a backfill. A funding round might go to product, not sales. So we stack two to four signals per account and tier on the count:
| Tier | What it means | What it gets |
|---|---|---|
| Tier 1 | Two or more live signals | Email, LinkedIn and ads |
| Tier 2 | One live signal | Email and ads |
| Tier 3 | Fit only | A small test audience |
On top of that, each account gets a score out of 10 from company size, current setup, buying signals, intent, whether we found the decision maker and geography. The weights are set with the client in onboarding. Eight or more is Tier 1, whatever the signal count says.
Signal sourcesJob boards, news, reviews
Contacts and emails27+ providers, verified
Tier the accountsTwo or more signals is Tier 1Tier 1
SequenceEmail, LinkedIn, adsResearch by hand before you automate
The temptation is to build the scraper first. We do the opposite. For a new ICP, someone on the team finds the first 20 good accounts by hand, reads why they look good and writes down where the evidence came from.
That manual pass does two things. It shows which signals actually turn up in this market, and it gives us the buyer's own words. We read 1 and 2 star reviews on G2, Capterra and Reddit for the language people use about the problem, and that language ends up in the copy.
Only then do we automate: Apify or PhantomBuster for the scraping, Clay to enrich and score, a waterfall of 27+ data providers to find a verified email, and a weekly refresh so new accounts arrive as fresh signals appear.
The signals we reach for most
These come up across almost every market we work in:
- Hiring for a role the product helps or replaces
- A new leader in the last 90 days
- Fresh funding, M&A or IPO plans
- A change in the tech stack, or a legacy system that is reaching end of life
- 1 and 2 star reviews of a competitor
- Followers and engagers of competitors, used with the client's permission
- A new plant, branch, entity or state
- Regulatory registers and deadlines
- Event speakers and attendees
- Website visitors, named and followed up the same day
- Job posts that mention Excel or manual processes, a sign there's no proper system yet
Which ones make the cut depends on the three questions above. Usually three to six signals per campaign survive.
