Content Moderation Policy

1. The principles we work from

Content moderation on a review platform is a balancing act between protecting honest critics from being silenced and protecting honest businesses from being lied about. Our principles below are what we use to keep that balance.

• Speech first, then exceptions. We start from the position that a user's words should stay up. We remove them only where a clear rule applies.
• Specificity beats sentiment. A specific factual claim that is shown to be false is more removable than a general feeling expressed strongly.
• Two-sided process. If a review is challenged, the reviewer gets to respond before we act. If a flag is rejected, the flagger can appeal.
• Plain language. Every decision we send includes a one-sentence reason in plain English, plus the rule it applies.
• Equal treatment. The same conduct gets the same outcome, whether the business is on a free plan or paying us our highest tier.

2. What we moderate

This policy covers all user content posted on Ratecrest, including reviews, ratings, photos attached to reviews, business replies, profile information, and any other public content a user puts on the platform.

Private messages between users and Ratecrest support are not moderated for content, but they are still bound by the Terms of Use.

3. How content reaches us

We do not pre-screen reviews. They go live as soon as you post them. We become aware of problem content in four ways.

3.1 Flags from users and businesses

The flag link on every review and business reply is the main way. Anyone can flag content, with or without an account.

3.2 Automated signals

Our systems watch for patterns that often correlate with abuse: bursts of accounts created shortly before posting, identical text across many businesses, posting from networks linked to past abuse, suspicious editing patterns, and similar. These signals do not result in automatic removals. They trigger a human review.

3.3 Lawful demands

Courts, regulators, and law enforcement agencies may serve us with orders. We respond to those that are properly issued by a Nigerian authority with jurisdiction. We push back where the demand is over-broad or improperly authorised.

3.4 Our own judgement

Our staff find things during the normal course of work. Where they spot a clear breach, we act on it.

4. The categories of action we take

We have a graduated set of actions, so that the response fits what was actually wrong.

• No action. The flag does not show a breach.
• Correction request. The reviewer is asked to update a specific factual claim. The review stays up while we wait.
• Redaction. Specific words are blacked out (the home address of a staff member, for example). The rest stays up.
• Removal. The review or content is taken down. The reviewer is told and the business is told.
• Temporary unpublish. The content is taken offline while a serious dispute is investigated. It returns if the investigation supports keeping it up.
• Account warning. Adds a note on the user's account. Repeated warnings can lead to suspension.
• Account suspension. The user cannot post or reply for a set period.
• Account closure. The user is removed from the platform.
• Profile penalty. For a business, can include loss of paid features, removal from category rankings, or display of a public moderation notice on the business profile.

5. How we tier the response

The tier we apply depends on the type of breach and any history we have on the account.

5.1 Serious and immediate

Some categories are removed on first contact, with no warning, and may be reported to authorities. They include child sexual abuse material, credible threats of violence, doxxing of private individuals, content that promotes terrorism, and content covered by an enforceable Nigerian court order.

5.2 Serious

Other categories are removed on first occurrence and result in an account warning. They include false statements of fact that cause harm, hate speech, sexual harassment of staff, and serious harassment of named individuals.

5.3 Moderate

These categories are usually corrected, redacted, or removed on first occurrence, and warned on repeat. They include minor factual errors, off-topic content, promotional content for unrelated businesses, and posting another person's contact details.

5.4 Soft

For low-grade issues like strong language inside an otherwise on-topic review, we may leave the content up with a note to the reviewer about platform expectations. We reserve the right to act on repeat patterns.

6. Defamation and the Cybercrimes Act

Defamation under Nigerian common law applies online. Certain conduct also falls within offences under the Cybercrimes (Prohibition, Prevention, etc.) Act, 2015, including the publication of false information that the publisher knows to be false and that causes harm. We take these laws seriously.

6.1 How a business raises a defamation concern

Use the flag link and select the option for false or defamatory content. Include the specific statement you say is false, why it is false, and any evidence you can share (your records, communications, photos). General complaints that "this is defamatory" without identifying the specific statement and explaining its falsity are not enough.

6.2 What we do with it

We assess whether the statement is one of fact or opinion. If it is opinion, we tell you so and the review usually stays up. If it is fact, we ask the reviewer to respond and, where useful, to support the claim. Based on the response, we may correct, redact, remove, or leave the content up. For complaints with credible evidence and serious potential harm, we may temporarily unpublish while we work it through.

6.3 What we are not

We are not a court. We do not adjudicate defamation in a binding way. A removal decision by Ratecrest is not a finding that defamation occurred; it is a finding that platform rules were breached. A decision to leave content up is not a finding that no defamation occurred; the business can still pursue its remedies in court.

7. Copyright complaints

Copyright in Nigeria is governed by the Nigerian Copyright Act, 2022. If a review uses your copyrighted work without permission, you can ask us to take it down.

A valid notice should include: identification of the copyrighted work, identification of the content on Ratecrest you say infringes it, a statement that you are the owner or authorised to act, your contact information, and a signature (electronic is fine). Send it through the contact form with "Copyright takedown" in the subject.

We remove content quickly where the notice is in order. The poster is told of the takedown and the reason. They can file a counter-notice if they believe the takedown was wrong, and we may restore the content if the counter-notice is in order and the original complainant does not pursue the matter further. The full procedure sits in the Copyright Policy.

Repeat infringers lose the ability to post on Ratecrest.

8. Targeted harassment and pile-ons

The single biggest moderation risk on a review platform is a coordinated attack on a business or an individual. We watch for it and act on it.

Signs we watch for include sudden bursts of negative reviews from accounts created shortly before posting, identical or near-identical wording across multiple reviews, coordination on external platforms, and posting from networks linked to past abuse. When we see a pattern that fits a pile-on, we may pause new reviews on the affected business, remove the suspicious ones, and clearly explain to legitimate reviewers what we did and why.

The same applies in reverse: a business owner or their associates orchestrating fake positive reviews to bury a negative pattern is also a pile-on. We act on those the same way.

9. Personal data in reviews

Reviews sometimes name people who are not the business owner: a specific staff member, another customer, a director, a service provider. Names alone are usually fine. Contact details and identity numbers are not.

We remove on request: home addresses of private individuals, personal phone numbers and email addresses of staff or customers, identity numbers (BVN, NIN, passport numbers), banking details, and photographs of staff or customers taken without their consent. The wider rules sit in the Privacy Policy.

10. Appeals

Almost every moderation decision can be appealed once. The exceptions are decisions made under a binding Nigerian court order, which we cannot lift on appeal.

An appeal is reviewed by a different member of our team than the one who made the first decision. They look at the content, the original complaint, the original reasoning, and anything new you give them. The appeal outcome is communicated in plain language with a reason. The appeal decision is final on the Ratecrest platform.

For serious matters we aim to resolve appeals within 7 days. For most other matters, 14 days.

11. Repeat offenders

The platform tolerates occasional mistakes. It does not tolerate repeated, deliberate breach. Accounts that accumulate multiple warnings or removals over a rolling 12-month period escalate to suspension, then to closure.

For businesses, repeated breaches around reviews (paying for reviews, instructing staff to post, gating asks, retaliating against reviewers) result in loss of paid features, removal from category rankings, and in serious cases removal of the business profile.

12. How our team is organised

Our front-line moderation team triages flags and applies routine actions under the rules above. A senior reviewer handles appeals and complex cases. A small policy group is responsible for these documents and for the harder calls where the principles in Section 1 are in tension.

Where a case involves legal complexity, including credible defamation claims, court orders, or potential offences, it is handled by senior reviewers with input from our legal advisors.

Staff training emphasises Nigerian law, plain-language reasoning, and the documented standards in this policy and the Terms of Use.

13. Transparency

We publish a transparency summary once a year. It includes aggregate numbers on flags received, content removed, accounts actioned, court orders received and complied with, and copyright notices processed. We do not publish the names of individual reviewers, businesses, or complainants, except where the law requires us to.

For specific decisions that affect you, we always explain what we did and the rule we applied. If you think the explanation does not match the rule, that is grounds for appeal.

14. What we do not do

• We do not remove reviews because a business owner is unhappy with them.
• We do not remove reviews in exchange for payment, advertising spend, or any other consideration.
• We do not give paying customers softer treatment than free users.
• We do not silently suppress reviews. Where a review is removed or unpublished, the reviewer is told.
• We do not respond to informal requests from law enforcement. They must be served through proper legal channels.
• We do not act on threats of litigation alone. We act on the underlying conduct.

15. Changes to this policy

We update this policy as the platform evolves and as we learn from cases. Small clarifications happen without fuss. Material changes are highlighted on the site when you next visit. The "Last updated" date at the top of this page always reflects the current version.

16. How to contact us

The contact form is the right way to reach us about anything in this policy. Write "Content moderation" in the subject so the right team handles it. For urgent matters involving immediate harm, write "Urgent: content moderation" and we will prioritise.