Linkedin 1:1 Ad Personalization Test Results

ROASted Labs (because why not)

1. Overview

Linkedin has introduced a 1:1 ad personalization feature that inserts a viewer’s first name, company name, and some other stuff in ad copies. Given the volume of B2B attention that flows through the platform, and the frequency with which small interface changes reshape feed behavior, our question was: whether this capability meaningfully alters how prospects notice, process, and act on paid messages.

Between Sep 24 and Oct 21, we ran a controlled field test across eight companies (7 B2B SaaS, 1 GTM community) targeting marketing (4), sales (2), cybersecurity (1), and revops (1) leadership cohorts. The feature was eligible for engagement, traffic, and conversion objectives. Test window is four weeks; longer horizons may show different fatigue dynamics, but current evidence suggests decay, not compounding value.

To isolate the mechanism, we held audiences, exclusions, creatives, bids, and budgets constant. The only manipulated variable was the presence of first-name and company tokens in ad copies. Controls were the immediately prior period with otherwise identical setup. This structure allowed us to attribute observed differences to personalization rather than targeting, creative, or pacing artifacts.

So this report investigates three questions:

1. Does 1:1 personalization change in-platform metrics (e.g., engagement rate, CTR)?

2. If interaction changes, does it translate into off-platform behavior (session depth, website actions, and conversions)?

3. What is the time profile of any effect (onset, durability, decay), and how should it inform campaign use cases?

   
   

Realistically, I was almost certain that personalization was going to improve the in-platform metrics, so my core question was:

Does personalization improve downstream performance (website engagement, lead conversion, and pipeline contribution) beyond its expected impact on in-platform interaction metrics?

So this report shows results by objective, traces the week-to-week pattern, and explains the mechanism in a B2B buying context, then I give some practical guidance on when, where, and how to use personalization.

TLDR: Personalization increased platform-level attention (engagement rate and CTR) but did not improve landing-page engagement, conversion rates, or pipeline. The effect profile is consistent with a pattern-interrupt/novelty mechanism: attention is captured, intent is not created. Therefore, personalization “works” for attention, but it doesn’t convert attention into anything meaningful. 

2. Context and Objectives

• Accounts: 8 total; 7 B2B SaaS, 1 GTM community.

• Personas: 4 marketing leaders, 2 sales leaders, 1 cybersecurity, 1 revops.

• Objectives tested by company:

  • Engagement: 5 companies

  • Traffic: 5 companies

  • Conversion: 2 companies

  • One company ran engagement + traffic + conversion; one ran traffic + conversion; others ran engagement + traffic; one ran engagement only.

    
    

• Feature availability: Personalization was not available for brand awareness; tests ran on engagement, traffic, and conversion objectives.

• Hypothesis (operationalized): Early-mover personalization will lift CTR/engagement and, but the primary question was: Will 1:1 personalization improve downstream value (LP engagement, conversions, pipeline), beyond raising platform attention?

3. Experimental Design

Variable	Controlled	Notes
Creative	Yes	Same image/video assets across control and experimental conditions
Audience	Yes	Same audience definitions, same exclusions
Bidding/Budget	Yes	No allocation, pacing, or bidding strategy changes
KPI Measurement Windows	Yes	Same attribution windows
Independent Variable	Personalization in ad copies	No other copy structural changes

Control condition: Prior period performance using non-personalized copy.

Test condition: Identical campaigns with personalization enabled.

Campaign Types and Distribution

Objective	# Companies	Notes
Engagement	5	Increase in ER & CTR, not in clicks to landing page
Traffic (Cold)	5	Click-through and retargeting pool development
Traffic (Retargeting)	2	Same as above - website metrics remained the same
Conversion	2	Website metrics remained the same

4. Results

4.1 Engagement campaigns 

• Engagement rate: +15–25% lift .

• Clicks to landing page: ~+1-2% lift

Metric	Baseline	Personalized	Δ
Engagement Rate	Baseline variable	+15–25%	+18% avg
Link Clicks → Landing Page	Baseline variable	~+1-2%	negligible
Downstream Sessions/Depth	No change	no change	0 effect

Interpretation:

Personalization increases acknowledgment behaviors (likes, low-effort clicks) but does not translate into exploratory intent or movement toward evaluation.

4.2 Traffic campaigns  (Cold)

Early effect (Week 1):

• One outlier account saw CTR ~2×.

• The other four were +3–7% CTR (~+5% average)

• Without the outlier, this would have failed a Week-2 continuation threshold.

  
  

Delayed novelty (Day 20+):

• Broader ~+50% CTR lift appeared across most cold traffic campaigns.

• No change to offer, audience, or creative; the lift is best explained by novelty/pattern-recognition rather than value resonance.

  
  

In our case, novelty/pattern-recognition means the tokens acted as a salience cue (“that’s me!!!”), triggering an orienting response and curiosity clicks without changing motivation; the actual value resonance would mean the message/offer itself mapped to a real need and drove deeper evaluation. 

We held offer, audience, bids/budgets, and creative constant, only added first-name/company tokens, so the mechanism behind the lift is unlikely to be a better value-message fit. The metric shape confirms this: CTR/engagement rose (including a ~+50% wave around days 15–20, consistent with recognition building at frequency ~3–5), while time on page fell ~20%, pages/session fell ~50%, and conversion stayed flat. If this were value resonance, you’d expect attention and downstream depth/CVR to rise together; instead we saw higher noticing, shallower sessions, and no economic movement.

In short: these tokens increased salience, not relevance; they manufactured attention, not demand.

On-site quality of those clicks:

• Time on page ~ −20%

• Pages per session ~ −50%

• Bounce ~flat

Interpretation: 

These were curiosity clicks, not intent clicks. Personalization improved CTR and lowered CPC, and grew retargeting pools, but did not increase the buying intent (surprised?) 

Metric	Observation
CTR Week 1	One outlier at +100%, others at +3–7% (mean +5%)
CTR Week (~Day 20+)	Cohort-wide +~50% shift
Time on Page	− ~20% vs baseline
Pages/Session	− ~50% vs baseline
Bounce Rate	~unchanged

So the performance lift is attributable to novelty-driven curiosity, not improved message resonance or value-offer alignment. This is a pattern-interrupt effect, not a demand effect.

4.2 Conversion campaigns  

CTR: Up in Week 2 (~+50%), then stagnation/−20% by Week 4 from peak.

Landing-page conversion rate: No improvement.

Pipeline: No impact.

Metric	Result
CTR (First two weeks)	+15–25%
Conversion Rate	No change
CTR (After week three)	− ~20%
Pipeline Impact	None

Interpretation:

Initial attention does not translate into economic outcomes. The effect decays as novelty fades. There is no evidence of personalization influencing buyer progression stages.

The attentional lift is transient and non-economic. We observe an early CTR increase (including a cohort-wide ~+50% wave around later days), followed by stabilization and ~20% decline by week four, while conversion behavior remains unchanged. On-site quality indicators move in the wrong direction (time on page ~−20%, pages/session ~−50%, bounce ~flat), confirming that incremental clicks are curiosity-driven rather than higher-intent. In other words, attention does not translate into economic outcomes. 

5. Analysis

Observed pattern: Early/mid-period attention lift (ER and CTR), followed by quality shortfall on site and no conversion lift.

Likely mechanism: Salience/pattern-interrupt. Personalized tokens trigger recognition and curiosity, not problem recognition or solution-value alignment.

Causality boundary: Personalization alters surface attention; it does not alter motivation. It amplifies an ad’s visibility; it does not create demand.

So,

1. Personalization increases attention, not intent.

2. Curiosity clicks ≠ meaningful engagement.

3. No signal of increased qualification, depth of evaluation, or readiness to consider.

4. Effect decays as novelty diminishes, indicating a finite pattern interrupt mechanism.

5. Downstream economic value is unchanged.

The result is consistent with cognitive load and relevance theory: Personal cues increase salience but do not modify intrinsic motivation or problem recognition.

6. Strategic & Practical Implications

When it is useful

• Top-of-funnel cold reach where the goal is cheap exposure and retargeting pool expansion.

• Short sprints where CPC reduction and CTR screens are acceptable interim goals.

Use Case	Recommendation	Rationale
TOFU audience expansion	Sort of	Lower CPC, larger retargeting pool
Retargeting	Not really	No movement in intent, qualification, or pipeline progression
Conversion campaigns	Nah	Attention without meaning does not convert
Offer or messaging testing	Potentially	Only for testing reactions, not effectiveness

When it is not useful

• Retargeting or conversion campaigns where demo/lead creation is the target.

• Any scenario where pipeline or sales outcomes are the KPIs.

  
  

Operating guardrails

• Personalization should be treated as a tactical pattern-interrupt, not a strategy.

• Cap spend and monitor a novelty curve (Weeks 1–4). Sunset when CTR plateaus/declines and LP metrics underperform (especially for the retargeting stuff)

• Keep tokens secondary to proposition. Do not stack multiple tokens if it reduces clarity.

• Track CPC alongside LP quality; do not accept CPC improvements at the expense of engagement depth.

7. Conclusion 

Personalization “works” in the narrow sense: it increases attentional salience, lifts vanity metrics (engagement rate, CTR), lowers CPC at the margin, and can expand cold retargeting pools. 

But the effect profile is novelty/salience, not demand creation. Name-token cues alter surface attention without changing underlying motivation or problem–solution fit; they operate as a visual stimulus, not a value-based relevance driver. 

As a result, personalization amplifies visibility but does not convert attention into meaning or economic value. Use sparingly at the top of the funnel; do not expect it to move pipeline.