How to build a lead list from company signals with PredictLeads and Crustdata

You build a lead list from company signals by starting with events instead of a static filter. First you find companies where something changed recently, then you check whether each one fits your ideal customer profile, and only then do you find the people to contact. PredictLeads is built for the first step. It tracks news events, funding rounds, and job openings across more than 120 million companies, and its discovery searches let you ask "which companies just raised money?" or "which companies are hiring a VP of marketing?" across that whole set. Crustdata is built for the second and third steps. Its company search filters on industry, headcount, headcount growth over one, three, six, and twelve months, funding, and location, and its person search covers more than a billion professional profiles by title, seniority, department, and current employer.

The two products don't talk to each other. Both are data providers that sell access through an API, and PredictLeads also offers flat files and webhooks, so the usual manual route is to export a batch of signals from PredictLeads, collect the company domains, run those domains through Crustdata to check size and growth, then search Crustdata for the right titles at each survivor and look up their emails one profile at a time. That works, but it takes an afternoon per batch, and most of that time goes to the decisions in between: which events still matter, which companies are a real fit, and which person at each company owns the problem you solve.

A good result is a short list you can act on this week. Every company on it fits your profile and had something happen recently, every contact still holds the role, and every row says what happened, when, and where it was reported. That event is what makes the first message specific instead of generic.

Kite does this work for you. It pulls the signals, judges which ones still count, checks each company's fit, and finds and scores the people, then delivers the list in Slack or writes it to your CRM.

Start by telling Kite in Slack what kind of moment you want to catch:

"Build me a list of heads of marketing at US B2B software companies with 50 to 500 employees that raised a round in the last two months."

Or you can describe the signal more loosely:

"Find companies that are hiring their first demand-generation team and tell me who to talk to."

Kite is an AI marketer in your Slack. It researches your business, builds a growth strategy, and ships the work that brings you new customers. Because it has already researched your company, it knows your ideal customer profile, your current customers, and your competitors. It proposes which signals to watch for, the company size and locations to target, and the titles to look for, and you confirm or adjust them before it starts.

You don't need a PredictLeads or Crustdata account. Both come built into Kite, along with Apollo for additional company and contact data. If you want the finished list in your CRM or a spreadsheet, connect that account through OAuth so Kite can write to it.

Kite picks the signal that matches what you sell and why a company would need it. A funding round usually means new budget. A new marketing or sales leader usually means the tools and vendors are about to be reviewed. Headcount growth or a new office means a team is scaling and running into problems it didn't have last quarter. A cluster of open roles in one department shows what the company plans to build next.

PredictLeads sorts news into categories, and Kite chooses among them on purpose rather than pulling everything. The categories it uses most for prospecting include these:

• A company raised a funding round.
• A company hired a new executive or leader.
• A company grew its headcount.
• A company opened an office in a new location.
• A company launched a new product.
• A company signed a new client.

Kite can also read a signal from Crustdata directly, such as companies whose headcount grew more than 20 percent in the last six months or whose last funding round was recent. It uses whichever source expresses your signal most precisely, and often both.

Kite uses PredictLeads's discovery searches to find companies by what just happened to them, narrowed by company size and location. For a funding signal, it searches news events in the financing category. For a hiring signal, it searches job openings by seniority, such as C-level, vice president, head, or director, and by occupation, such as marketing managers, because PredictLeads can't search job titles as free text.

Each event comes back with its category, the date PredictLeads found it, a one-sentence summary, and the article that reported it. For funding, Kite also reads PredictLeads's structured financing record, which gives the round type, the amount, and the investors, along with the source articles. For hiring, each job opening carries its title, seniority, location, the date it was first seen, and a link to the posting.

Kite throws out any event older than about 90 days, because a signal that old is no longer a reason to reach out today. PredictLeads's news discovery doesn't accept a date range, so it can return events from years ago. Kite reads each event's date and drops the stale ones before it spends anything on enrichment.

It also reads what the event is actually about. Some PredictLeads events describe a relationship mentioned in an article about another company, such as one firm being named as a competitor of another. Kite keeps the events that are news about the company itself and sets the rest aside. For job openings, it keeps roles that are still open and reads the titles, because a company hiring three demand-generation managers tells a different story from one hiring a single marketing coordinator.

Kite checks every company against your ideal customer profile before it looks for a single person, because a signal is a reason to reach out, never a substitute for fit. It looks each company up in Crustdata by its domain and reads the industry, total headcount, headcount growth over recent months, estimated revenue, funding history, and headquarters location.

Companies that fail the profile are dropped, along with your existing customers and your competitors. When a company's headcount or funding numbers look off, Kite compares them with PredictLeads's financing record and the company's own pages before trusting them. Data providers sometimes attach one company's numbers to another with a similar name, and a wrong headcount can put a 40-person startup on an enterprise list.

Kite uses Crustdata's person search to find the people at each company who own the problem you solve. It searches by the company's domain, the department, such as Marketing or Growth, and seniority, such as Director, Vice President, or CXO. When the signal names a person, such as a newly hired head of marketing, Kite looks for that person first.

A Crustdata profile can lag behind a recent job change. Kite confirms that each person still holds the role at that company, using their current professional profile or the company's own team page, and drops anyone who has moved on. It notes recent title changes too, since someone who started in the job last month is often the one reviewing vendors.

Kite enriches each confirmed contact through Crustdata's profile enrichment, which returns the person's current and past employment, headline, location, and their work email when Crustdata has one. When Crustdata has no address, Kite tries its other built-in sources, including Apollo.

Kite keeps each email's verification status exactly as the provider reports it, whether that's verified, catch-all, risky, or unverified. A missing address is marked as not found rather than guessed. Kite includes only professional details, so personal email addresses and phone numbers stay off the list.

Kite gives every contact a fit score from one to five with a one-line reason that joins the fit and the signal. A reason might read, "VP of marketing at a 140-person data-infrastructure company that raised a Series B six weeks ago and has posted four marketing roles since." When the list runs past about fifteen people, Kite splits it into primary and secondary groups so you know where to start.

If fewer companies qualify than you asked for, Kite tells you how many it found, which filters it used, and why the rest fell short. A short list of companies that fit and have a fresh reason to talk is worth more than a long list padded with old news.

Kite posts the finished list in the Slack thread where you asked, with a summary of how many signals it pulled, how many it dropped as stale or off-profile, and how many contacts have a work email. With a CRM or spreadsheet connected, it writes the list there as well.

From there, Kite can draft a first email for each primary contact that opens with the event that put them on the list, and it can send those emails once you approve them. Signals keep arriving, so you can also ask Kite to run the same search every week. Each run looks for new events and posts its batch to the thread you choose.

When Kite finishes, you have a list of people at companies that fit your profile and had something happen recently, with everything you need to write a first message that isn't generic.