Custom Audiences Masterclass

A practitioner-level playbook for building, managing, and exploiting custom audiences across Meta and LinkedIn to drive pipeline, warm prospects before outreach, and dramatically improve email performance through multi-channel sequencing.

---

1. Platform-Specific Upload Mechanics

Facebook Custom Audiences

Facebook accepts CSV or TXT uploads with the following identifiers (the more you provide, the higher your match rate):

• Email address (primary matcher -- typically yields 60-75% match rate on B2B lists)
• Phone number (include country code)
• First name, last name
• City, state, zip, country
• Date of birth
• Mobile advertiser ID (IDFA or GAID)

Upload process:

1. Navigate to Audiences inside Ads Manager
2. Select "Create Audience" then "Custom Audience"
3. Choose "Customer list" as the source
4. Map your columns to Facebook's accepted fields
5. Upload and wait for processing (typically 30 minutes to 2 hours)

Key detail: Facebook hashes all PII data using SHA-256 before matching. You can pre-hash your data before uploading for added security, but it is not required -- the platform handles it automatically.

Match rate optimization tips:

• Normalize email addresses (lowercase, trim whitespace)
• Remove duplicate rows before upload
• Include as many identifier columns as possible -- adding phone number to an email-only list can boost match rate by 10-15 percentage points
• Remove obviously invalid emails (info@, noreply@, test@)
• Facebook is flexible on minimum audience size -- you can upload lists as small as 100 contacts and still create audiences, though ad delivery may be limited

LinkedIn Matched Audiences

LinkedIn offers several matched audience types:

• Contact targeting: Upload email addresses (requires a minimum of 300 contacts to activate the audience)
• Company targeting: Upload company names or domains (requires minimum of 300 companies)
• Retargeting: Website visitors, video viewers, Lead Gen Form openers, event attendees, company page engagers

Upload process:

1. Go to Campaign Manager then Plan then Audiences
2. Click "Create audience" then select "Upload a list"
3. Choose contact list or company list
4. Map fields and upload CSV

Critical threshold: LinkedIn requires a minimum of 300 matched contacts before the audience becomes usable. If your list is 500 contacts but only 250 match, the audience will not activate. Always upload at least 400-500 contacts to account for match rate attrition.

Processing time: LinkedIn matched audiences take a full 24-48 hours to populate. Do not launch campaigns targeting a freshly uploaded audience -- wait at least a full day for the audience to finish building. Running ads against a partially populated audience wastes budget and skews early performance data.

Match rate reality: LinkedIn match rates on email uploads typically run 30-50% for B2B lists. Company list matching is higher (60-80%) when you use exact domain URLs rather than company names.

---

2. Lookalike Audience Creation

Facebook Lookalike Audiences

Lookalike audiences find users who statistically resemble your source audience based on demographics, interests, and behavioral signals.

Best practices:

• Source audience quality matters more than size. A 500-person list of actual buyers will outperform a 10,000-person list of newsletter subscribers as a lookalike seed.
• Start with 1% lookalike in your target country -- this represents the top 1% of users most similar to your source. In the US, 1% equals roughly 2.1 million people.
• Layer lookalike percentages for testing: Run 1%, 2-3%, and 4-5% as separate ad sets to compare performance at each similarity level.
• Refresh your source audience quarterly. Customer behavior profiles shift, and stale seed audiences produce progressively less relevant lookalikes.
• Exclude your source audience from the lookalike to avoid paying to reach people you already have.

Stacking strategy: Combine a 1% lookalike with a relevant interest layer to narrow targeting. For example, a 1% lookalike of your buyers intersected with "Business Decision Makers" interest targeting. This produces smaller but higher-intent audiences.

LinkedIn Lookalike Audiences

LinkedIn lookalike audiences expand your reach based on professional attributes similar to your source audience.

• Source audience must contain at least 300 members
• LinkedIn creates a lookalike that is typically 15x the size of your source
• You cannot adjust the similarity percentage like you can on Facebook
• Lookalike audiences update dynamically as LinkedIn identifies new similar members
• Best seed audiences: website retargeting lists, customer email lists, Lead Gen Form submitters

---

3. Retargeting Architecture

Engagement-Based Retargeting

One of the most powerful and underutilized retargeting strategies is building audiences from people who engage with your social content organically.

Facebook engagement retargeting options:

• People who engaged with your Facebook Page (any interaction)
• People who engaged with any post or ad (likes, comments, shares, clicks)
• People who watched a percentage of your video (25%, 50%, 75%, 95%)
• People who opened or completed a Lead Gen Form
• People who interacted with your Instagram business profile
• People who clicked any CTA on your Page

The comment tagging technique: When someone tags a colleague or connection in the comments of your content, that tagged person now appears in your engagement audience -- even though they did not originally follow you or interact with your brand. This is an organic audience growth mechanism that compounds over time.

To exploit this:

1. Create content that encourages tagging ("Tag a colleague who needs to see this")
2. Build a custom audience of "People who engaged with any post or ad" over a 180-day window
3. Run retargeting campaigns to this audience with next-stage content
4. The tagged individuals now see your retargeting ads, warming them before any direct outreach

Separate retargeting campaigns for social engagers: Do not lump social engagers into the same retargeting campaign as website visitors. Social engagers are earlier in the funnel and need different messaging. Create dedicated retargeting campaigns for:

• Social engagers (serve educational and value-driven content)
• Website visitors (serve case studies, testimonials, demos)
• Pricing page visitors (serve urgency and social proof)

Customer List Retargeting with Testimonials

Upload your existing customer list and target them with testimonial content and case study ads. This serves multiple purposes:

• Increases engagement from your existing customer base (they share and comment, expanding organic reach)
• Builds social proof that prospects in overlapping audiences will see
• Keeps your brand top-of-mind for upsell and cross-sell opportunities
• Creates engagement signals that feed back into your retargeting audiences

---

4. Engagement vs. Conversion Campaign Objectives

Why Engagement Campaigns Outperform Conversion Campaigns for Niche B2B

This is counterintuitive to most media buyers, but for small, niche B2B audiences, engagement campaign objectives consistently deliver higher ROI than conversion campaign objectives.

The mechanics behind this:

Conversion campaigns require the algorithm to find people within your audience who are most likely to convert. The algorithm needs a minimum volume of conversions (typically 50 per week per ad set) to exit the learning phase and optimize effectively. For a niche B2B audience of 5,000-20,000 people, you will rarely generate enough conversion volume to train the algorithm.

What happens instead:

• The algorithm thrashes during the learning phase, never stabilizing
• CPMs spike as the algorithm narrows delivery to an increasingly small subset
• You burn budget on a poorly optimized delivery model
• Results are erratic and non-repeatable

Engagement campaigns with niche audiences:

• The algorithm optimizes for engagement signals (likes, comments, shares, link clicks) which occur at much higher volume
• It exits the learning phase faster and delivers ads more efficiently
• CPMs stay lower because the algorithm has sufficient signal to optimize
• The engagement itself creates retargeting audiences you can use downstream
• Total cost per qualified pipeline opportunity drops significantly

When to use conversion objectives: Only when your target audience is large enough (100,000+) and you can realistically generate 50+ conversions per week per ad set. For most B2B campaigns targeting specific job titles in specific industries, this threshold is unreachable.

Statistical Significance with Small Audiences

Small audiences (under 10,000) will rarely produce statistically significant results in traditional A/B testing frameworks. Do not let this stop you from running campaigns. The goal with small audiences is awareness and familiarity, not statistical optimization. You are building recognition so that when you reach out via email or sales, the prospect already knows your brand.

---

5. Audience Warming Before Email Outreach

This is the highest-leverage technique in the custom audiences playbook: running awareness and engagement campaigns to your target list on Facebook and LinkedIn before sending your first email.

The Warming Sequence

Week 1-2: Social awareness phase

1. Upload your email outreach list to both Facebook Custom Audiences and LinkedIn Matched Audiences
2. Run low-budget awareness/engagement campaigns showing helpful, non-promotional content
3. Use video content when possible (15-30 second thought leadership clips)
4. Frequency target: 3-5 impressions per person across platforms
5. Budget: Even small daily budgets of a few dollars per platform are sufficient for lists under 5,000

Week 3: Email launch

1. Send your first email to the same list
2. The recipient has already seen your brand and face 3-5 times on social media
3. When they see your email in their inbox, they recognize you -- you are not a stranger

Results from Audience Warming

Running social awareness campaigns before the first email campaign produced higher-than-benchmark open rates, click-through rates, and click-to-open rates, with zero unsubscribes on the first send. The audience warming approach delivered meaningfully better results than cold outreach across all key email metrics.

Why this works psychologically:

• Mere exposure effect: repeated exposure to a stimulus increases preference for it
• Pattern interruption: instead of being yet another cold email, you are a recognized entity
• Trust transfer: seeing someone on social platforms (especially with engagement from others) creates implicit credibility
• The prospect feels like they chose to engage rather than being targeted

Ongoing Warming Cadence

Do not stop social campaigns after the first email. Maintain a continuous low-budget presence on social platforms targeting your email list. This keeps warming active throughout your entire email nurture sequence and ensures every subsequent email arrives to a recipient who continues to see your brand across channels.

---

6. Deep Target Market Persona Understanding

Custom audiences are only as effective as your understanding of who belongs in them. Improving CTR on custom audience campaigns requires going beyond basic demographic targeting.

Building Actionable Personas for Custom Audiences

For each audience segment, document:

• Job titles and variations: The same role can have 20+ title variations across companies. A "VP of Marketing" might also be "Head of Growth," "Director of Demand Generation," "Chief Marketing Officer," or "Growth Lead." Miss these variations and your matched audience shrinks.
• Content consumption patterns: What podcasts do they listen to? What newsletters do they read? What conferences do they attend? This informs interest targeting layers.
• Pain points by seniority: A Director-level person cares about execution efficiency. A VP-level person cares about pipeline metrics. A C-level person cares about revenue impact. Your ad creative must match the seniority within each audience segment.
• Platform behavior: Some audiences are more active on LinkedIn during business hours and Facebook in evenings. Test dayparting accordingly.
• Buying triggers: What events cause this persona to enter a buying cycle? Budget cycles, new hires, competitive pressure, regulatory changes? Time your campaigns to these triggers.

CTR Improvement Through Persona Alignment

When your ad creative speaks directly to the specific pain point of the specific persona in your custom audience, CTR improves dramatically. Generic messaging to a custom audience wastes the precision that custom targeting provides.

Example structure:

• Audience segment: Directors of Operations in healthcare
• Pain point: Compliance documentation taking 15+ hours per week
• Ad creative: Lead with the specific time-waste metric, show how the solution reduces it
• CTA: Tailored to their evaluation process (they need to present to a VP, so give them a shareable resource)

---

7. Operational Best Practices

Audience Refresh Cadence

• Customer lists: Re-upload quarterly to capture new customers and remove churned ones
• Email lists: Re-upload monthly if your list is growing actively
• Engagement audiences: Set rolling windows (30, 60, 90, 180 days) and test which window produces the best downstream metrics
• Lookalike audiences: Rebuild from updated source audiences quarterly

Audience Exclusions

Always layer exclusions to prevent wasted spend:

• Exclude existing customers from acquisition campaigns (unless running upsell)
• Exclude people who already converted from retargeting campaigns
• Exclude employees of your own organization
• On LinkedIn, exclude your competitors' employees if they are likely to click out of curiosity (this inflates CTR without producing pipeline)

Budget Allocation Across Audience Types

A general allocation framework for B2B custom audience campaigns:

Audience Type	Budget Share	Objective
Lookalike (cold)	40-50%	Awareness / Engagement
Custom list (warm)	20-30%	Engagement / Traffic
Retargeting (hot)	20-30%	Conversion / Lead Gen

Cross-Platform Audience Sync

Upload the same lists to both Facebook and LinkedIn simultaneously. Prospects seeing your brand on both platforms creates a surround-sound effect that amplifies recognition. The incremental cost is minimal (you are already paying to reach them on one platform), but the compound awareness effect is significant.

Monitoring and Hygiene

• Check audience sizes weekly -- if an audience suddenly shrinks, investigate (platform policy changes, list quality issues)
• Monitor frequency caps -- once frequency exceeds 8-10 on a custom audience campaign, creative fatigue will tank performance
• Remove bounced and invalid emails before uploading to prevent low match rates
• Track audience overlap between ad sets using Facebook's Audience Overlap tool to prevent self-competition

---

8. Advanced Techniques

Sequential Retargeting Funnels

Build multi-stage retargeting funnels using custom audiences:

1. Stage 1: Serve educational content to your uploaded list (engagement objective)
2. Stage 2: Retarget engagers from Stage 1 with case study content
3. Stage 3: Retarget engagers from Stage 2 with a direct offer or demo request
4. Stage 4: Retarget Stage 3 non-converters with a different angle or social proof

Each stage filters for progressively higher intent, and the audience self-selects through engagement.

Using Custom Audiences for Market Research

Before launching a new product or entering a new market, upload a target list and run engagement campaigns with concept-testing content. Measure engagement rates, comments sentiment, and click-through patterns to validate market interest before investing in full campaign buildout.

Combining Custom Audiences with Offline Signals

For B2B with long sales cycles, integrate offline conversion data:

• Upload closed-won customer lists as high-value seed audiences
• Upload sales-qualified leads as mid-funnel seed audiences
• Upload marketing-qualified leads as top-funnel seed audiences
• Create separate lookalikes from each to test which stage produces the best top-of-funnel performance

Closed-won lookalikes consistently outperform MQL lookalikes because the source signal is stronger and more predictive of actual buying behavior.

---

Quick Reference: Platform Comparison

Feature	Facebook/Meta	LinkedIn
Minimum audience size	~100 (flexible)	300 (strict)
Match rate (email)	60-75%	30-50%
Audience build time	30 min - 2 hours	24-48 hours
Lookalike sizing	1-10% adjustable	Fixed ~15x source
Engagement retargeting	Yes (granular)	Yes (limited)
Cost efficiency	Lower CPMs	Higher CPMs, better B2B targeting
Video view audiences	Yes (25/50/75/95%)	Yes (25/50/75/97%)
Company targeting	Via uploads only	Native company targeting