Estimated Action Rate (EAR)

(System Behavior Framework — Comprehensive Edition)

1. What is Estimated Action Rate (EAR)?

Estimated Action Rate (EAR) is Meta’s internal prediction of how likely a specific user is to complete your optimization event—such as:

• Purchase
  
  

• Add to Cart
  
  

• Lead submission
  
  

• Landing page view
  
  

• App install
  
  

• Video watch
  
  

• Click

EAR is one of the three core components of Meta’s auction ranking formula:

Total Value = Bid × Estimated Action Rate × Ad Quality

It directly influences which ads win impressions, how much they cost, and who they are delivered to.

EAR is not visible to advertisers.
It is a fully internal, AI-driven probability model.

2. How does it work inside the ad platform?

EAR uses machine learning models trained on billions of user interactions across Meta’s ecosystem.

It combines three categories of signals:

A. User-Level Behavioral Signals

These predict what actions a person tends to take:

• purchase history
  
  

• recent browsing patterns
  
  

• app activity
  
  

• engagement patterns (scroll stops, likes, shares)
  
  

• time spent on landing pages
  
  

• video watch behavior
  
  

• session frequency
  
  

• device type & reliability
  
  

• connection speed

Each user receives an invisible probability score like:

“This person has a 0.042 likelihood of making a purchase in the next 24 hours.”

B. Ad-Level Predictive Signals

Meta evaluates how well your specific ad is expected to drive the desired action.

Signals include:

• creative clarity
  
  

• hook strength
  
  

• predicted scroll-stopping rate
  
  

• expected click behavior
  
  

• landing page consistency
  
  

• previous engagement patterns from similar creatives
  
  

• domain history (e.g., slow pages = lower predicted action rate)

C. Contextual & Auction-Time Signals

These include:

• time of day
  
  

• current session intent
  
  

• user’s recent actions
  
  

• competition level
  
  

• placement type (Reels, Feed, Stories)
  
  

• device/browser behavior
  
  

• real-time performance variance

EAR is recalculated for each impression opportunity.

3. Why does EAR affect performance?

EAR is one of the most important hidden levers inside Meta’s entire system.
It impacts:

A. Auction Wins

Higher EAR = higher Total Value = more affordable auction wins.

Even a lower bid can win if EAR is strong.

B. Cost Efficiency

High EAR → lower CPM, CPC, and CPA
Low EAR → higher CPM, CPC, and CPA

EAR is one of the strongest predictors of whether your ads become scalable winners or expensive failures.

C. Delivery Optimization

EAR determines which users the algorithm prioritizes for your ad.

The system tries to show your ads to users with high predicted conversion probability.

D. Learning Phase Speed

Ads that show strong user-response signals early develop a higher EAR → learning phase ends faster.

E. Creative Lifespan

As creative fatigue sets in, engagement and conversions fall → EAR drops → CPM rises → performance degrades.

EAR is extremely sensitive to creative fatigue.

F. Funnel Movement

EAR rises or falls depending on:

• audience freshness
  
  

• retargeting quality
  
  

• offer-to-audience fit
  
  

• landing page performance

EAR can dramatically shift between cold and warm audiences.

4. When does this become important to marketers?

a) When launching new creatives

EAR predicts which new creative gets early traction.

b) When scaling budgets

EAR determines auction competitiveness—high EAR = easier scaling.

c) When diagnosing CPM spikes

Low EAR → high CPM
High EAR → lower CPM

d) When choosing optimization events

EAR is closely tied to your Optimization Event (Purchase vs Add to Cart vs Link Click).

e) When creative fatigue hits

EAR naturally drops as repeat exposure increases.

f) When landing page performance is weak

Slow or dysfunctional landing pages drive EAR down—even with strong ads.

g) During major seasonal competition (Q4)

EAR must outperform competing advertisers for similar users.

5. Common pitfalls or misunderstandings

1. Thinking EAR is static

EAR updates in real time as user behavior changes.

2. Assuming EAR is only about creative

EAR is influenced by creative + landing page + audience + session context.

3. Trying to fix EAR with budget changes

EAR is behavior-prediction-driven, not budget-driven.

4. Underestimating landing page impact

Slow pages lower conversion likelihood → EAR drops → costs spike.

5. Over-editing ad sets

Frequent edits disrupt learning → EAR becomes unstable.

6. Assuming high EAR = high ROAS

EAR predicts likelihood of action, not profitability.

6. What should you understand next connected to this system?

Following the locked rule (must choose only from your keyword list), the next most relevant concepts directly connected to EAR are:

Ad Relevance Score

(because both relevance and EAR combine to determine auction value)

Optimization Event

(what action EAR is actually predicting)

Ad Delivery Optimization

(EAR feeds into delivery logic at every step)

Creative Fatigue

(EAR collapses as user engagement drops)

ROAS

(EAR influences the cost side of the ROAS equation)