# Validating Primer Audiences

#### Step 1 — Validate Reach & Audience Integrity

**Goal:** Prove the audience is matching your ICP and can actually deliver reach across channels.

**Check** [**match rates & usable sizes**](https://support.sayprimer.com/primer/building-audiences/managing-audiences-in-primer/matched-audience-size) **(per destination).**

In each platform (Meta, LinkedIn, Google Demand Gen/YouTube, Reddit), confirm that the audience is eligible and large enough to serve.\
In Primer, use **History Log** on each audience to confirm synced/matched counts.

<figure><img src="https://888674264-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FGKE0Prjh2dfbNy05EsHY%2Fuploads%2F7cgFh7LcAYPB3LAkfsic%2FHistory%20Log.png?alt=media&#x26;token=1fa1f0bf-fe1b-48bb-b36b-91e32b895f7f" alt=""><figcaption></figcaption></figure>

[**Preview the audience**](https://support.sayprimer.com/primer/building-audiences/how-to-build-an-audience/audience-preview) **in Primer.**

Sanity-check that company, seniority, department, and geo line up with your ICP (and that obvious non-fit companies are minimal).

<figure><img src="https://888674264-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FGKE0Prjh2dfbNy05EsHY%2Fuploads%2FPq82loWREcIRzhsneDTe%2Fpreview.png?alt=media&#x26;token=66376692-d7ad-4909-bc57-019814fb296f" alt=""><figcaption></figcaption></figure>

**Validate against LinkedIn (neutral composition check).**

Sync the same audience to LinkedIn and open **Campaign Manager → View Insights** to review composition.

<figure><img src="https://888674264-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FGKE0Prjh2dfbNy05EsHY%2Fuploads%2FNltYio8w1sm4BXXLdG4H%2Fimage.png?alt=media&#x26;token=810fe028-27d2-4e89-a12d-5f36fbdec7e9" alt=""><figcaption></figcaption></figure>

**Important:** LinkedIn often buckets general C-suite titles (e.g., *Vice President, Board Member, Founder, Executive* without a function) into **Business Development**. If your audience includes these titles, expect to see an inflated “Business Development” function in Insights—this is a categorization artifact, not a flaw in your list.

#### **Confirm ICP engagement via reverse-IP**

Use 6sense/ Demandbase/ Warmly filtered by relevant **UTMs** to see which accounts are engaging with ads/content driven by your Primer audiences. You’re looking for the *right* companies showing up— early green light.

<figure><img src="https://888674264-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FGKE0Prjh2dfbNy05EsHY%2Fuploads%2FtBB0Ac5ceHMmzzgppeu6%2Fimage.png?alt=media&#x26;token=d968d4db-fae3-4e44-941c-bbd322885062" alt=""><figcaption></figcaption></figure>

**Pass/Fail heuristic:**

* Pass = healthy match/eligibility across 2+ channels, LI insights align with ICP (allowing for the BD bucket caveat), and reverse-IP shows the right logos.
* If fail, refine your audience (negative filters, industry/tech filters, title normalization) and re-sync.

***

#### Step 2 — Build Awareness & Retargeting Pools (Cross-Channel)

**Goal:** Create efficient, scalable retargeting layers before pushing for conversion.

**Run education/awareness to the inclusion audience:**

* **YouTube/Demand Gen:** View-through friendly formats to warm up accounts; add *% video viewed* segments to retargeting.
* **Meta:** Video views, ThruPlays, and Landing Page Views as signals (don’t rely solely on website visits).
* **LinkedIn:** Thought-leadership units and Document Ads to your ICP; add “engaged with ad” segments.
* **Reddit:** Interest + custom list inclusion to reach technical personas where relevant; build site/video engagement pools.

**KPI for this phase:** growth of qualified retargeting segments. Build retargeting segments from video viewers, ad engagers, LP viewers as well as from website visitors.&#x20;

<figure><img src="https://888674264-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FGKE0Prjh2dfbNy05EsHY%2Fuploads%2F8YKirFSWjxut4ojtxaKa%2Fimage.png?alt=media&#x26;token=d85b5c07-db81-45aa-b63d-e88529589b53" alt=""><figcaption></figcaption></figure>

***

#### Step 3 — Drive Qualified Leads (Parallel Paths)

**Goal:** Convert warmed audiences while respecting channel realities.

**Instant/Native Lead Forms (Meta/LI/YouTube Lead Form extensions):**\
Expect more personal emails by default on Meta and Youtube. Counter with a follow-up workflow that requests work email (or enriches and routes intelligently).

**On-site forms (preferred for work email capture):**\
Run in parallel. Ensure **mobile-first** page speed and UX (most social ad traffic is mobile; many B2B buyers submit on desktop later).

**Mind the mobile→desktop gap:**\
Touch-based attribution will under-credit channels like Meta/YouTube when the *impression → research → later desktop form* path is common. Watch for correlated lifts in **direct/organic desktop** traffic and brand search when last-click looks light.

{% embed url="<https://youtu.be/V4SmQSojeUY>" %}

***

#### Step 4 — Evolve Measurement to True Incrementality

**Goal:** Isolate the causal impact of advertising from everything else (SEO, email, outbound, events, SDR).

**Install the** [**Primer Pixel** ](https://app.gitbook.com/u/09zch0JCBORlmfSvSBQuqy75i2j2)**everywhere it should be.**\
This is non-negotiable. Proper pixel coverage lets you unify event streams, dedupe, and create consistent retargeting/holdout logic across channels.

<figure><img src="https://888674264-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FGKE0Prjh2dfbNy05EsHY%2Fuploads%2FV9K7y2vhfpSOsAPfLJH7%2Fimage.png?alt=media&#x26;token=f9f0a7e1-9356-4d67-9f15-0b5ed5f148e2" alt="" width="563"><figcaption></figcaption></figure>

**Use Primer's built-in** [**Holdout Groups**](https://support.sayprimer.com/primer/learn-more/conversion-lift-analysis-beta) **to measure lift.**\
Primer’s holdouts let you keep a statistically similar slice of your audience *unexposed* to ads. Compare exposed vs. holdout on downstream outcomes (site visits, high-intent pages, demo starts, MQLs, pipeline).

**Why it matters:** Holdouts isolate advertising’s impact **independent of** brand, content, outbound, PR, or seasonal effects. This is how you prove the ads moved the needle—not just touched it.<br>

<figure><img src="https://888674264-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FGKE0Prjh2dfbNy05EsHY%2Fuploads%2F104vxK5p24XmM6bUyaSX%2Fimage.png?alt=media&#x26;token=2be01de7-6f8d-43f3-8cb2-d8d3223d59c5" alt="" width="331"><figcaption></figcaption></figure>

**Cadence & decisions:**\
Review weekly for QA and pacing; run formal lift reads at day 30 and day 60. **Double down** on audiences/channels with positive, material lift; **pause** neutral/negative arms; **expand** winners with adjacent lookalikes/segments.

***

#### Practical Channel Notes

* **Meta:** Mobile-heavy exposure; prioritize video+LP view retargeting; expect personal emails on Instant Forms; ensure use of mobile optimized website forms to get last click conversions.
  * [Meta Lead Gen Use Case](https://support.sayprimer.com/primer/meta-lead-gen-custom-audiences)
  * [Meta Lookalikes Use Case](https://app.gitbook.com/u/09zch0JCBORlmfSvSBQuqy75i2j2)
* **LinkedIn:** Best for composition validation and high-fit reach; CPMs higher—use for thought leadership and CRM retargeting or if see bad titles slipping in via LinkedIn's native audiences.
  * [LinkedIn Lead Quality Use Case](https://support.sayprimer.com/primer/linkedin-audience-efficiency)
  * [ABM Aircover Use Case](https://app.gitbook.com/u/09zch0JCBORlmfSvSBQuqy75i2j2)
  * [CRM Retargeting Use Case](https://app.gitbook.com/u/09zch0JCBORlmfSvSBQuqy75i2j2)
* **YouTube/Demand Gen:** Cost-efficient reach and powerful video engagement pools; stitch to site via UTMs and pixel; strong for air-cover + assisted conversions.
  * [Demand Gen Use Case](https://support.sayprimer.com/primer/apply-audiences-to-demand-gen-campaign)
* **Reddit:** Great for technical sub-communities; test against relevant subreddits/interest clusters; when ready use Primer audiences to expand reach and target your ICP across any subreddit.&#x20;

***

#### QA & Experiment Hygiene (Do This Every Time)

* **Consistent UTMs** per arm for reverse-IP and analytics.
* **Creative consistency:** Keep creative consistent across any A/B testing to avoid confounds.
* **In Meta Experiments set up A/B tests on ad sets** (avoids resetting learning; steer clear of Advantage+ audience expansion)
* **In Google Experiments use cookie/user-level splits always** (avoid search-query splits for incrementality reads).

***

#### What “Good” Looks Like by Phase

* **Validation:** Eligible audience across platforms; LI insights match ICP (with BD caveat); reverse-IP shows right logos.
* **Awareness:** Growing retargeting pools; lift in branded/direct traffic from ICP geos/verticals.
* **Conversion:** Lead/MQL CPL competitive with (or beating) LinkedIn benchmarks on warmed cohorts; rising work-email capture rate.
* **Incrementality:** Stat-sig positive lift on site visits, high-intent actions, and pipeline from exposed vs. holdout.

***

#### Troubleshooting Fast

* **Low match / small eligible size:** Normalize titles; add firmographic anchors (industry/size/tech); remove edge geos; re-sync.
* **Right reach, wrong composition in LI Insights:** Audit titles—strip generic C-suite where irrelevant; add department/seniority filters; re-sync.
* **Thin retargeting pools:** Shift spend to video/view-through formats; increase frequency caps; broaden within ICP (adjacent industries/seniority).
* **Leads but low quality:** Tighten inclusion (company filters, tech filters); add Anti-ICP exclusions; gate assets with qualifying fields.
* **No measurable lift:** Verify pixel coverage, UTMs, and holdout integrity; extend duration; increase spend to hit MDE (minimum detectable effect).

***

#### TL;DR

1. Prove audience integrity everywhere (Primer previews → LI Insights → reverse-IP).
2. Warm them up cross-channel and build retargeting pools (not just website visits).
3. Convert with native + onsite in parallel, respecting mobile→desktop behavior.
4. **Install the pixel and run holdouts** to isolate causal impact from the rest of your GTM.
5. Reallocate to winners based on **incremental lift**, not touch-based vanity.