Ghost ICP Builder: How to Define Your Ideal Customer Profile for the 600M Contact Database
TL;DR: Your ICP definition is the single biggest lever in LinkedIn outbound — better targeting beats better copy every time. This guide walks you through Ghost's six ICP dimensions, how to translate them into search filters against a 600M contact database, and how to iterate after your first campaign.
An ideal customer profile (ICP) for LinkedIn outbound is a precise, filter-ready definition of the person most likely to buy your product — built from firmographic, technographic, and role-level criteria that can be directly applied to a contact database. It is not a persona document. It is a search query.
Ghost is a LinkedIn GTM platform that connects content creation to intent-powered outbound. At the centre of that outbound engine is a 600M verified contact database — searchable by role, industry, company size, geography, and tech stack. But the database is only as powerful as the ICP you feed into it.
Why ICP Quality Is the Biggest Lever in LinkedIn Outbound
Most founders and sales teams treat copywriting as the primary variable in outbound performance. It isn't. A brilliant message sent to the wrong person converts at zero. A decent message sent to the right person converts at 3–5%.
According to Ghost's internal data from Q1 2025, campaigns where the ICP was narrowly defined (fewer than 5,000 contacts, at least four active filters) outperformed broad-list campaigns by 2.3x on reply rate — with no difference in message quality. The targeting did the work.
The problem is that most teams define their ICP once, in a slide deck, and never translate it into something a database can actually use. Ghost's ICP builder forces that translation. Every dimension you define becomes a live filter against the contact database — which means vague thinking produces vague lists, and precise thinking produces pipeline.
Founder's Take: I've reviewed hundreds of outbound campaigns inside Ghost. The ones that underperform almost always have the same problem: the ICP is too wide. "VP of Sales at a SaaS company" is not an ICP. "VP of Sales at a Series A–C SaaS company with 20–150 employees, selling into mid-market, using Salesforce, in the UK or US" — that's an ICP. The extra specificity feels like you're leaving people out. You are. That's the point.
The 6 ICP Dimensions Ghost Uses
Ghost structures ICP definition around six dimensions. Each one maps directly to a search filter in the contact database. Work through them in order — later dimensions refine the list that earlier dimensions produce.
Role and Seniority
Start with job title and seniority level. These two filters together define who you're actually talking to — not just the department, but the decision-making layer within it.
Be specific about title variants. If you're selling HR software, your buyer might be "Head of People", "VP HR", "Chief People Officer", or "People Operations Manager" — and these titles carry very different seniority signals depending on company size. In Ghost's database, you can search by keyword across title fields and layer a seniority filter (Director, VP, C-Suite) on top to catch the right level regardless of exact wording.
Avoid filtering by department alone. "HR" as a department filter will return everyone from a junior HR coordinator to the CHRO. You want the person with budget authority and the pain point your product solves.
Industry and Vertical
Industry filters in Ghost map to LinkedIn's standard industry taxonomy, which means you can get granular. "Software" is too broad. "HR Tech", "Fintech", "B2B SaaS", or "Professional Services" will produce a list that actually reflects your go-to-market.
If your product sells horizontally across industries, resist the temptation to select all of them. Pick the two or three verticals where you have the strongest case studies or the clearest pain point narrative. Run those as separate campaigns so you can personalise the messaging to each vertical's specific language.
Company Size
Company size (measured by headcount) is one of the most powerful ICP filters because it directly correlates with buying behaviour, budget cycles, and decision-making complexity.
A founder selling a £200/month project management tool to a 10-person agency has a completely different sales motion than one selling a £50K enterprise contract to a 500-person professional services firm. Ghost lets you filter by headcount range — typical bands are 1–10, 11–50, 51–200, 201–500, 500–1,000, 1,000+. Choose the band where your current customers cluster, not where you aspire to sell.
Geography
Geography matters more than most teams admit. Outbound conversion rates vary significantly by region — not just because of language, but because of market maturity, buying culture, and the density of your existing customer base.
In Ghost, you can filter by country, region, or city. If you're a UK-based founder with case studies from London-headquartered companies, lead with the UK market. The social proof lands harder when the prospect recognises the reference customer. Expand geographically once you've validated your ICP and message in a familiar market.
Tech Stack or Category
Technographic filters are where Ghost's ICP builder pulls significantly ahead of most outbound tools. You can filter contacts by the software their company uses — which means you can target, for example, every company running HubSpot CRM with between 50 and 200 employees in the SaaS space.
This matters for two reasons. First, tech stack is a strong proxy for budget, sophistication, and workflow maturity. A company using Salesforce, Gong, and Outreach is a different buyer than one using spreadsheets. Second, it enables integration-led positioning: "We integrate natively with the tools you're already using" is a far stronger opening than a generic value proposition.
Exclusion Criteria
Exclusion filters are the most underused dimension in ICP building — and often the most valuable. Defining who you don't want in your list is as important as defining who you do.
Common exclusion criteria include: companies that are too early-stage (pre-revenue or pre-product), direct competitors, existing customers, industries with regulatory constraints that make your product unsellable, and roles that look right by title but sit in the wrong function (e.g. "Operations Manager" who manages facilities, not revenue operations).
In Ghost, exclusions can be applied at the database filter level before you build the campaign — which means you're not wasting sequence steps on contacts who will never convert.
How to Translate Your ICP Into Ghost's Search Filters
The gap between "we sell to mid-market HR teams" and a deployable Ghost search is where most teams lose time. Here's how to close that gap systematically.
Take a SaaS founder selling an employee engagement platform. Their ICP statement might be: "Head of People or VP HR at a Series B or C SaaS company with 100–500 employees, using Workday or BambooHR, based in the UK or US, not a competitor or existing customer."
Translated into Ghost filters, that becomes:
- Title keywords: Head of People, VP HR, Chief People Officer, People Operations
- Seniority: Director, VP, C-Suite
- Industry: Computer Software, Internet, Information Technology
- Headcount: 100–500
- Geography: United Kingdom, United States
- Tech stack: Workday OR BambooHR
- Exclusions: Known competitors (by company name), existing customer domains
That filter set will typically return a list of 2,000–8,000 contacts — large enough to run meaningful campaigns, small enough to personalise meaningfully.
How to do this in Ghost: Navigate to Find People and open the advanced filter panel. Add your title keywords in the "Job Title" field, then layer seniority, industry, headcount, and geography filters. Use the "Technologies" field for tech stack filters — type the software name and select from the dropdown. Add exclusion domains in the "Exclude Companies" field. Review the live contact count in the top right before saving — aim for a list under 10,000 before launching your first AI campaign.
How to Tell If Your ICP Is Too Broad
There are four warning signs that your ICP definition needs tightening before you burn through your contact list.
Your filtered list returns more than 50,000 contacts. At that scale, you haven't defined an ICP — you've defined a market. Narrow by adding at least one more filter dimension.
Your message could be sent to anyone in the list without changing a word. If the personalisation fields in your sequence feel generic, your ICP isn't specific enough to generate relevant context. A well-defined ICP produces messages that reference the prospect's specific stack, stage, or vertical.
Your reply rate is below 1% after 200 sends. According to Ghost's internal benchmarks, a well-targeted ICP with a competent sequence should produce a 2–4% positive reply rate within the first 200 sends. Below 1% almost always indicates a targeting problem, not a copy problem.
You're getting replies from people who aren't buyers. If the people responding are too junior to make a decision or are in adjacent functions that don't have the pain point, your title or seniority filters need adjustment.
Iterating Your ICP After the First Campaign
Your first ICP definition is a hypothesis. The campaign is the test. Here's how to read the results and refine.
After your first 200–300 sends, pull the reply data from Ghost's outbound analytics and segment it by the filter dimensions you used. Look for patterns: which industry vertical replied most? Which seniority level converted to booked calls? Which company size range produced the most qualified conversations?
If 80% of your positive replies came from companies with 50–150 employees but your filter included up to 500, tighten the headcount band. If your best conversations came from Heads of People rather than VP HR, update your title keywords and deprioritise the VP filter.
Run ICP iteration as a monthly discipline, not a one-time setup task. Markets shift, your product evolves, and the contacts who were a poor fit six months ago may be a strong fit today. Ghost's saved search filters make it straightforward to update your ICP definition and regenerate a fresh contact list without rebuilding the campaign from scratch.
Founder's Take: The best ICP work I've seen happens after the third campaign, not the first. By then you have real signal: who replied, who booked, who ghosted, and who unsubscribed. That data is worth more than any persona workshop. Build your ICP in Ghost, run it, read the replies, and adjust. The database has 600M contacts — you're not going to run out of people to test against.
Frequently Asked Questions
What is an ideal customer profile in the context of LinkedIn outbound?
An ideal customer profile (ICP) for LinkedIn outbound is a filter-ready definition of the prospect most likely to buy your product, built from role, seniority, industry, company size, geography, and tech stack criteria. Unlike a buyer persona, an ICP is designed to be directly applied to a contact database to produce a targetable list. The more precisely you define your ICP, the higher your outbound reply and conversion rates will be.
How many filters should I use when building my ICP in Ghost?
For most B2B outbound campaigns, four to six active filters produce the best balance between list size and targeting precision. Fewer than four filters typically returns a list that's too broad to personalise effectively. More than eight filters can over-constrain the list and reduce it below the minimum viable size for statistically meaningful results — aim for a final list of 1,000–10,000 contacts.
How do I define ICP for outbound if I'm an early-stage founder with limited customer data?
Start with your best current customer — even if you only have three or four — and reverse-engineer their firmographic profile: their role, company size, industry, and tech stack. Use that as your first ICP hypothesis and run a campaign of 200–300 sends to validate it. Ghost's 600M contact database means you can find lookalike contacts at scale even before you have a large customer base to analyse.
What is the best company size filter for B2B SaaS outbound?
The best company size filter depends on your product's price point and sales motion. Products priced under £500/month typically convert best with companies in the 10–100 headcount range, where a single decision-maker can approve the purchase. Products priced at £1,000–£5,000/month tend to perform best with 100–500 headcount companies, where there's budget authority but not a lengthy enterprise procurement process.
Why does tech stack matter as an ICP filter?
Tech stack filters reveal budget, workflow maturity, and integration fit — three signals that strongly predict whether a prospect will buy. A company running Salesforce, HubSpot, and Gong is demonstrably investing in revenue infrastructure, which makes them a stronger prospect for a complementary tool than a company with no visible tech investment. Technographic targeting also enables integration-led positioning, which consistently outperforms generic value proposition messaging in cold outreach.
How do I know if my ICP is too narrow?
Your ICP is likely too narrow if your filtered contact list returns fewer than 500 people, or if you've exhausted your list within two campaigns without generating enough replies to draw conclusions. Another signal is when your exclusion criteria are eliminating the majority of contacts — this usually means you're excluding based on assumptions rather than data. Widen one filter at a time and monitor whether reply quality changes.
How often should I update my ICP definition?
Review and update your ICP definition after every three campaigns, or monthly — whichever comes first. Outbound data accumulates quickly, and patterns in who replies, books, and buys will become visible within 300–500 sends. Treat your ICP as a living document rather than a fixed definition, and use Ghost's saved search filters to update your targeting without rebuilding campaigns from scratch.
Can I run multiple ICPs simultaneously in Ghost?
Yes, and for most founders this is the recommended approach once you've validated your first ICP. Running two or three parallel campaigns targeting different verticals or seniority levels lets you identify which segment converts fastest, without waiting months to test sequentially. Ghost's campaign builder supports multiple simultaneous campaigns, each with its own ICP filter set, sequence, and analytics — so you can compare performance side by side.
Your ICP is the foundation everything else in outbound is built on. Get it right, and the database, the sequences, and the AI do the heavy lifting. Get it wrong, and no amount of clever copy will save you.
Ghost gives you a 600M contact database, six ICP filter dimensions, and an AI campaign builder that turns your ICP definition into a live outbound sequence in under 90 seconds. Start your 7-day free trial — no credit card required — and build your first ICP-driven campaign today.
