How to Build a B2B Prospect List from Scratch

Building a prospect list sounds simple until you're three hours in, staring at a spreadsheet full of job titles you guessed at, company sizes pulled from LinkedIn's fuzzy headcount bands, and email addresses you're not sure are real.

Most SDRs skip the architecture and jump straight to outreach. Then they wonder why their bounce rate is 12% and their reply rate is 0.4%.

This guide walks through a repeatable process for building a prospect list that holds up — starting with a tight ICP, filtering down to the right accounts, sourcing verified contacts, and ending with something ready to sequence without torching your sender reputation.

Step 1: Lock down your ICP before you touch any tool

Every bad list starts with a vague ICP. "Mid-market companies in the US" is not an ICP. It's a direction.

A usable ICP has at least six dimensions:

Industry (NAICS or SIC level) — not just "tech" but "B2B SaaS" or "IT Managed Services"
Company size — defined by headcount range AND revenue range, because a 50-person company can be $2M or $20M
Geography — country, region, or metro if your product has a local component
Tech stack — what tools do your best customers already use? (Salesforce shops vs. HubSpot shops behave differently)
Business motion — do they sell to SMB or enterprise? Do they run outbound? Do they have a field sales team?
Trigger signals — recent funding, new hire in a relevant role, product launch, expansion to a new market

Spend 30 minutes pulling your last 20 closed-won deals and look for the pattern. If you're pre-revenue or early-stage, look at competitors' customer logos or case studies. The ICP is a hypothesis you refine — but you need a starting hypothesis with real specificity before you build anything.

Practitioner note: Headcount from LinkedIn is notoriously unreliable — companies game it, and LinkedIn counts contractors and alumni inconsistently. Cross-reference with a data provider's revenue estimate or Crunchbase funding rounds for more signal.

Step 2: Build your target account list first

Contact-level prospecting is a trap if you skip account-level targeting. You end up with 500 contacts spread across 400 companies — too thin to run any meaningful account-based motion.

Start with accounts. Layer contacts on top.

How to source target accounts

LinkedIn Sales Navigator is the default starting point. Use the Account filter — not the People filter — and set industry, headcount range, geography, and keywords (look for product or market descriptors in company descriptions). Export or save up to 1,000 accounts per search. You'll cull later.

Industry databases like Crunchbase, G2, Capterra (for software companies), or trade association directories surface accounts that LinkedIn misses — especially sub-$10M companies that don't maintain a strong LinkedIn presence.

Job boards as a signal — companies posting for SDRs, BDRs, or VP of Sales are building out a sales function. Companies posting for roles that use your product category are signaling active investment. Boolean search on Indeed or LinkedIn Jobs: "HubSpot" AND "Sales Development Representative" surfaces companies where you know both the tech stack and that they're scaling sales.

Your CRM's closed-lost list — if you're not re-prospecting churned opportunities from 12+ months ago with a new message and a new champion, you're leaving money on the table. These accounts already know your category exists.

Account scoring: keep it simple

Score each account on two dimensions — fit (how closely they match your ICP) and timing (any signal they're in-market right now). A 2x2 matrix works fine:

	High timing signal	Low timing signal
High fit	Tier 1 — prioritize now	Tier 2 — sequence, monitor
Low fit	Tier 3 — opportunistic	Discard

Tier 1 accounts get multi-channel, multi-touch sequences. Tier 2 goes into a longer nurture cadence. Tier 3 either gets a single touchpoint or gets cut entirely.

Step 3: Define the buying committee, not just a title

For any deal above ~$5K ACV, you're rarely selling to one person. Map who's involved before you pull contacts.

A typical B2B buying committee has three to five roles:

Economic buyer — approves budget, often VP or C-level, typically gets involved late unless you hook them early
Champion — the internal advocate who feels the pain most acutely and will sell on your behalf
Technical evaluator — validates implementation, security, or integration fit
End user — the person who'll actually use the tool; their buy-in affects adoption and renewal
Procurement/legal — enters late but can kill deals

For each account, you want at least two to three contacts: the champion (usually a director or manager), the economic buyer (VP or above), and one technical or operational contact if your product has an IT component.

Searching only for VPs misses champions. Searching only for managers misses the person who signs. Map the committee first, then source contacts for each role.

Step 4: Source verified contacts

This is where most lists fall apart. You have 200 accounts and the right titles in mind — now you need actual contact data that won't bounce.

The B2B data industry has a dirty open secret: most providers sell you an email address that was verified six months ago, at best. Email addresses decay at roughly 2–3% per month — that's 25–30% of any static database going stale in a year. If your provider refreshes on a monthly or quarterly schedule, you're prospecting with a meaningful percentage of dead addresses already baked in.

Target a bounce rate below 3%. If you're seeing 5%+, your data source is either stale or low quality. Above 8%, you're going to start damaging your sending domain.

When evaluating a contact data provider, ask:

When was this contact last verified? Not "when was it added to the database" — when was the email last confirmed deliverable?
What's the verification method? SMTP handshake verification catches bad addresses without sending a real email. Providers that rely only on format validation are giving you false confidence.
What's your bounce rate guarantee or refund policy? A provider confident in their data will stand behind it.

Tools like LeadsApp verify contacts at the moment you reveal them — rather than pulling from a cache verified months ago — and refund credits for contacts that can't be verified. That point-of-use verification model matters more than a flashy database size number.

For a deeper look at verification timing and what it means for deliverability, see our breakdown of email verification approaches.

Multi-source stacking

No single provider has 100% coverage. For high-priority Tier 1 accounts, source contacts from two providers and cross-reference. If both return the same email, your confidence goes up. If they diverge, run the address through a verification tool (NeverBounce, Zerobounce, or Millionverifier) before sending.

Step 5: Enrich and structure your list

Raw contact data — name, title, email, company — isn't enough to run a personalized sequence. You need context.

Enrichment fields worth adding:

Direct dial or mobile — improves call connect rates significantly. Mobile number connect rates run 3–5x higher than office lines in 2025.
LinkedIn URL — for social touches and to verify the contact is still at the company
Seniority level — not just title, but actual seniority (Director vs. Senior Director matters for tone)
Tech stack — pull from Builtwith, Clearbit, or your data provider if available. Knowing they use Salesforce changes your opening line.
Recent company news — funding rounds, leadership changes, product announcements. Set up Google Alerts for Tier 1 accounts.
Persona tag — map each contact back to their buying committee role (champion, economic buyer, etc.)

Structure your spreadsheet or CRM with these fields before import. Retrofitting enrichment after you've started a sequence is painful.

Step 6: Clean and validate before you import

Before this list touches your sequencing tool, run a final hygiene pass.

Deduplication — check for the same contact appearing twice with slightly different name formatting. Salesforce and HubSpot both have native dedupe tools; use them.

Email validation pass — if you're not using a point-of-use verification provider, run the full list through a bulk verification tool. Accept "valid" results. Flag "catch-all" domains for manual review — catch-all means the domain accepts all email regardless of whether the mailbox exists, so you can't verify automatically. Treat risky or unknown results as uncontactable until you can validate another way.

Suppression check — pull your existing customer list, your opt-out list, and your recent-contact list. Suppress anyone who's heard from you in the last 90 days or who's already a customer. Hitting a current customer with a cold prospecting email is a relationship-damaging mistake that happens more often than it should.

Title normalization — "VP, Sales" and "Vice President of Sales" and "VP Sales" should all map to the same field value, or your segmentation breaks down.

Step 7: Segment before you sequence

A list isn't ready to work until it's segmented. Sending the same email to a founder at a 10-person startup and a VP of Sales at a 500-person company wastes everyone's time.

Minimum segmentation cuts:

By persona (champion vs. economic buyer — these get different messages with different angles)
By company size (SMB vs. mid-market vs. enterprise — tone, pricing anchors, and pain points all differ)
By industry vertical (if your messaging can be industry-specific, it should be)
By timing signal (Tier 1 accounts with recent triggers get a different opening than Tier 2 evergreen accounts)

For a list of 500 contacts, you might end up with 6–8 segments. Each gets its own sequence variant. The personalization doesn't have to be manual — dynamic fields and conditional blocks in your sequencer (Outreach, Salesloft, Apollo, Smartlead, Instantly) let you scale it. But the segmentation logic has to be built before you import, not after.

Step 8: Set baseline metrics and iterate

A prospect list is a hypothesis. The sequence is the test. Know what success looks like before you run it.

Typical benchmarks for cold outbound (email):

Metric	Weak	Acceptable	Strong
Bounce rate	>5%	2–5%	<2%
Open rate	<25%	25–40%	>40%
Reply rate	<2%	2–5%	>5%
Positive reply rate	<0.5%	0.5–1.5%	>1.5%
Meeting booked rate	<0.3%	0.3–1%	>1%

High bounce rate? Your data source is the problem. Low opens? Subject line or sender reputation. Acceptable opens but low replies? Your body copy or call-to-action needs work. Each metric points to a different lever.

Track these by segment, not just overall. A failing segment pollutes your numbers and hides what's working. If your mid-market champion segment is booking at 1.2% but your enterprise economic buyer segment is at 0.1%, that's directional information about where to double down.

Keep a version log of your lists — which accounts, which contacts, which sequence, what time period. When you find a combination that works, you want to be able to replicate it.

The build-out checklist

Before you call your list ready to sequence:

ICP defined with at least 6 dimensions
Target accounts sourced from 2+ channels
Accounts scored by fit and timing signal
Buying committee roles mapped per account
Contacts sourced with recent verification date
Enrichment fields populated (LinkedIn, tech stack, news)
Full list deduplicated
Email addresses validated (< 3% expected bounce)
Suppression list applied
List segmented by persona, size, and vertical
Sequence variants built per segment
Baseline metrics defined

If you want to see what contact search looks like when verification is built into the workflow rather than bolted on after, explore LeadsApp's features — it's built for exactly this kind of structured list-building.

Frequently asked questions

How many contacts should a B2B prospect list have?

Size depends entirely on your capacity to work the list. An SDR running structured multi-channel sequences can realistically work 150–200 active contacts per month at a quality level that drives results. A list of 2,000 unworked contacts isn't worth more than a list of 300 well-researched ones. Start with enough contacts to fill 4–6 weeks of sequences, then replenish as you move accounts through the funnel.

How often should I refresh or rebuild my prospect list?

At minimum, run a fresh verification pass every 90 days on any contacts you haven't yet sequenced. Contacts you're actively sequencing should have been verified within 30 days of first send. For accounts you're not actively working, expect 25–30% contact-level decay per year — meaning a list you built 12 months ago has meaningful spoilage even if you haven't touched it.

What's the difference between a prospect list and a contact database?

A contact database is a pool of possible contacts matching broad criteria. A prospect list is a curated, prioritized, verified subset of that database — segmented and ready for outreach. Most teams confuse the two and try to sequence their entire database. The process above is specifically about converting a broad database into an actionable, sequenceable prospect list.

Is it better to buy a list or build one manually?

Neither, on its own. Buying a static list gives you volume but poor targeting and stale data. Building purely manually gives you high quality but doesn't scale. The right workflow combines a data provider for contact sourcing and verification with your own targeting logic — ICP, account scoring, buying committee mapping — applied on top. You're buying the contact data, not the strategy.

How do I handle catch-all email domains?

Catch-all domains accept all incoming email at the server level, so automated verification tools can't confirm whether a specific mailbox exists. You have a few options: skip them and accept lower coverage, send to them in a low-volume test batch from a warmed secondary domain and watch bounce rates, or find an alternate contact method (LinkedIn, direct dial) for those contacts. For Tier 1 accounts, the manual verification effort is usually worth it. For Tier 2 and below, the risk often isn't.

What tools do I actually need to build a prospect list?

At minimum: a contact data provider with recent verification (not a stale database), LinkedIn Sales Navigator for account sourcing and buyer research, a CRM or structured spreadsheet for list management, and a bulk email verification tool if your provider doesn't verify at point of use. Optional but high-value additions include a tech stack enrichment tool (Builtwith or Clearbit), a sequencing platform, and intent data if your budget allows. Don't buy intent data before you've nailed ICP definition — you'll misinterpret the signals. See LeadsApp's pricing for a cost-effective starting point on the contact data layer.