The Basics
This must be defined before starting any experiment.
A Hypothesis.
Testing this campaign element helps determine whether users respond better to different offers or call-to-actions.
Ad copy components.
What is the term for the practice of paying for advertisements to reach a specific audience on platforms like Facebook, Instagram, and LinkedIn?
Paid Social Media Advertising.
This marketing tactic involves showing ads on social media to users who have previously interacted with your PPC ads or visited your website via a PPC campaign.
Retargeting or Remarketing.
This statistical concept describes the phenomenon where marketing efforts in one channel indirectly contribute to conversions that are ultimately attributed to another channel often making it difficult to isolate the true impact of individual touchpoints.
What is marketing halo effect (or cross-channel influence)?
This metric calculates the total revenue generated from an advertising campaign relative to the cost of that campaign, often expressed as a ratio or percentage.
Return on Ad Spend (ROAS).
This kind of keyword prevents your ad from showing for certain search queries.
Negative keyword.
This ad format allows Facebook to dynamically test combinations of headlines, images, and calls-to-action.
Dynamic Creative Ads.
What strategic advantage can be gained by aligning the messaging and creative assets used across both your PPC and Paid Social campaigns?
Increased Brand Recognition and Brand Awareness.
Identifying the specific group of users who will be exposed to the different variations in your test is known as defining the what?
Target audience.
The two most common types of marketing tests/experiments used to measure the effectiveness of different strategies.
A/B and Pre/Post tests.
What is the acronym for the auction-based system where advertisers bid on keywords to have their ads appear in search engine results?
PPC
What type of audience targeting on paid social allows advertisers to reach users similar to their existing customer base or website visitors?
Lookalike Audiences.
What is a common metric used to evaluate the efficiency of both PPC and Paid Social campaigns in driving desired actions, such as jobs booked or app downloads?
Conversion Rate.
In statistical terms, what does "statistical significance" indicate about the results of an A/B test, helping you determine if the observed difference between variations is likely due to chance?
That the observed difference between the variations is unlikely to have occurred randomly and is likely a real effect.
The strategic metric that determines how much confidence you can have in your test results.
Statistical significance.
This metric, calculated by dividing the total cost of clicks by the number of clicks, indicates the average price an advertiser pays each time their ad is clicked.
CPC
What is the name for the type of social media ad that blends seamlessly with the organic content of a user's feed?
Native Advertising.
How can retargeting audiences built from website visitors (tracked via PPC campaigns) be utilized in paid social campaigns to re-engage potential customers?
Personalized ads on social media platforms.
Beyond simple A/B tests, what is a more complex experimentation method that allows you to simultaneously test multiple elements and their combinations to identify the optimal experience?
Multivariate Testing
This advanced attribution model uses algorithms and machine learning to assign credit to different marketing touchpoints based on their actual contribution to conversions.
Data-Driven Attribution.
What is the term for adjusting your bids in a PPC campaign based on factors like the user's device, location, or the time of day?
Bid Adjustments or Bid Modifiers.
This advanced paid social tactic involves uploading a list of your existing customers to a platform to target them or create lookalike audiences.
Customer Match.
This experiment splits audiences or regions to isolate the incremental effect of one channel on another.
Incrementality Test.
When analyzing the results of a digital marketing experiment, what potential threat to the validity of your findings might occur if external factors unrelated to your test variations influence user behavior during the testing period?
Confounding Variables