Potential Fake Review Playbook: Flag Processing Guidelines

This playbook outlines guidelines for identifying and processing potential fake reviews on our platform. While increased attention stems from recent FTC regulations, the primary goal is to maintain the authenticity of user feedback, prevent review manipulation, and uphold trust across our ecosystem.

Types of Flags and How to Handle Them

Reviewed a Product in a Different Interest Group

This flag is triggered when a reviewer has submitted reviews for products in entirely different interest groups, making it unlikely that they have genuine experience with both.

• To assess legitimacy, check the reviewer’s LinkedIn profile:
  
  

  • Have they had a significant career change that could explain their exposure to both product categories?
    
    

  • Have they acquired new skills or completed courses that make it reasonable for them to have used both products?
    
    

→ Do not publish the review if the reviewer’s background does not support the likelihood of using both products.

Phone Number Matches Another User

This flag appears when two or more user profiles share the same primary phone number.

• This may indicate an accidental duplicate account rather than an attempt to manipulate the system.
  
  

• If the names are nearly identical—differing only by a single letter, number, or special character (such as an underscore)—it is likely that the person accidentally created a duplicate account rather than attempting to bypass the system or impersonate someone else.
  
  

→  In this case, merge the accounts and publish the review if no other flags are present.

• If the two profiles belong to different individuals—evident by distinct names, pictures, email addresses, or other identifying details—yet share the same phone number, it strongly indicates fraudulent activity.
  
  

→ Flag both profiles and do not publish the review.

Received an Incentive for a Previous Review

This flag is raised when a reviewer has left a review for a product that qualifies for a gift card incentive and has previously received a similar incentive for another review.

• This measure prevents users from creating accounts solely to leave fake reviews in exchange for rewards.
  
  

• To verify the legitimacy of the reviewer:
  
  

  • Check their review history to see if they have repeatedly left incentivized reviews.
    
    

  • Review their LinkedIn profile to confirm that their corporate email corresponds with their job experience.
    
    

→  The review can be published if the reviewer’s profile appears credible and no other concerns arise.

 Credible profile: 

• Works for a company the user works for looks real (up-to-date website, the clickthrough options lead to a new page and don’t just reload the page).

• The product they are reviewing makes sense for their job experience (ex: marketing person leaving a review for an IT platform).

• Check to see if the use case aligns with their job experience/industry. Skim through the review to see if there is other information you can use to verify the user. Meaning, if the person works for a small sized company but says it’s saved them millions of dollars a year, that’s suspicious depending on what the company does.

• User is not in a tech hub city. Uganda and Kenya have a high spam rate.
  
  

However, if the profile seems suspicious—for example, lacking activity, missing job experience, or showing inconsistencies with the products they have reviewed

→  Do not publish the review.

Works for a Competitor (Not an FTC Flag, but Still Relevant)

Although this is not an official FTC flag, it prevents a review from being automatically published.

• This flag was implemented to address concerns that employees of competing companies may leave biased or negative reviews.
  
  

• The level of concern depends on the company’s size:
  
  

  • If the reviewer works for a small to mid-sized competitor

→ Flag the review and escalate it to a superior for further evaluation.

• If the reviewer is employed by a large company (such as Microsoft, SAP, or IBM), their company likely covers a broad range of categories, making it reasonable for them to have experience with the product. 

→ In these cases, the review can be published.