Her finner du en samlet, strukturert oversikt over hvordan GoRabbit fungerer – for kunder, tekniske brukere og governance/styring.
Controls targeting and categorization.
The Segments module manages all categorization of groups, industries, themes, and locations. It determines which Facebook groups a campaign should publish to based on segment ID, rules, regions, and relevance. Segments acts as a smart targeting and filtering system.
Finds the right groups for each campaign.
Segments is responsible for matching campaign themes, industries, and geographies with relevant groups in fb_group_master.json. Without Segments, the proprietary internal mechanism handled automatically by the system. Segments provides the intelligence needed to decide which groups fit a campaign.
fb_group_master + rules + customer proprietary internal mechanism handled automatically by the system.
Segments uses three main sources: (1) fb_group_master.json, (2) segment_rules.json, and (3) customer-specific overrides (gruppe_mapping.json). Together, they form the basis for targeted publishing lists.
A list of relevant groups.
Segments generates segmented lists of groups, typically in normalized_group_list.json. This proprietary internal mechanism handled automatically by the system.
Criteria that match campaigns to groups.
Segment rules are criteria that tell the system which groups fit which campaigns. Examples: ‘Real estate → housing groups’, ‘Travel → holiday groups’, ‘Norway → region=NO’, ‘PoolMan → pool maintenance groups’. Rules are defined in segment_rules.json.
Relevance-based prioritization.
Segments have priority levels based on relevance and risk. Example: (1) High relevance (direct match), (2) Medium (related topic), (3) Low (general interest). proprietary internal mechanism handled automatically by the system.
Yes — full geographic support.
Yes. Groups can be linked to regions (e.g., NO, SE, Alicante, Oslo, Nordics, EU). Campaigns may target countries, counties, or cities. This makes GoRabbit ideal for proprietary internal mechanism handled automatically by the system.
Industry determines targeting groups.
Industry segments group Facebook groups by topic: real estate, travel, hobbies, IT, food, health, etc. Campaigns are assigned an industry ID, and Segments automatically finds matching groups.
Detailed interest categories.
Thematic segments are fine-grained categories such as 'families with children', 'renovation', 'garden', 'hobby photography', 'dog owners', 'motorhomes', etc. They allow highly precise targeting and are used by the proprietary internal mechanism handled automatically by the system.
proprietary internal mechanism handled automatically by the system.
RabbitEngine reads the segment ID, segment_rules.json, and fb_group_master.json. It then performs a scoring match for each group. Only groups with scores above a defined threshold are included in normalized_group_list.json.
Yes — hierarchical targeting.
Yes. Segments may have: (1) top level (industry), (2) subthemes, (3) location, (4) niche categories. proprietary internal mechanism handled automatically by the system.
Yes — multiple segments per campaign.
Yes. A campaign may be assigned multiple segment IDs. Segments merges all matching groups, removes duplicates, and returns one consolidated list.
Yes — custom segments supported.
Yes. Customers may have custom segments defined in client-specific segment files. These override global segments when matched.
Score + metadata matching.
Each group has metadata: category, region, type, member count, risk, language, and industry. The segment is compared against these fields and a scoring is calculated. The proprietary internal mechanism handled automatically by the system.
Rules for countries and regions.
Geographic rules define which regions the campaign allows and which regions the groups belong to. Example: The ‘Norway’ segment automatically excludes groups marked ‘SE’, ‘FI’, or ‘DK’.
Filters out low-quality groups.
Segments filters out groups with low activity, high risk, heavy spam, or inappropriate content. Groups requiring admin approval or strict posting rules are also excluded. The result is a clean, relevant, and safe list.
Customers can override global segments.
Customers can define their own segment preferences in gruppe_mapping.json or a dedicated segments proprietary internal mechanism handled automatically by the system. These override global segment rules and allow manual inclusion or exclusion of specific groups. This is useful for customers with special needs or private groups.
Yes — multilayer segmentation.
Yes. Segments use multiple layers such as industry → theme → region → niche → rule sets. RabbitEngine uses these layers to assign each group a combined relevance score. This allows highly precise targeting even in large segments.
Relevance score 0–100.
Segment score is a value from 0–100 determining how well a group matches a campaign. The score is calculated by analyzing: (1) category match, (2) region, (3) activity, (4) member count, (5) language, (6) risk, (7) segment rules. Only groups above a threshold are included.
Yes — groups may be excluded.
Yes. If a group has high risk, wrong region, low activity, or does not match the industry, it is automatically excluded. Customers may also manually exclude groups via the proprietary internal mechanism handled automatically by the system.
Yes — removes inactive groups.
Yes. Inactive groups or groups without recent posts are filtered out. This prevents campaigns from being published into dead or irrelevant groups.
Yes — spam groups blocked.
Yes. Segments analyzes risk flags, rules, activity, and language patterns. If a group appears spam-like, the proprietary internal mechanism handled automatically by the system. It is then removed from all segments.
Yes — size matters, but relevance first.
Yes. Groups with extremely low member counts are excluded. Large groups may receive a higher score, but Segments prioritizes relevance over size.
Assesses risk before publishing.
Risk is assessed by: (1) admin strictness, (2) how often posts are rejected, (3) language patterns, (4) rules, (5) historical blocks. High-risk groups receive low score or are excluded entirely.
Yes — language affects segmentation.
Yes. Norwegian campaigns prioritize groups in NO or those tagged as 'NO-language'. English campaigns have more global flexibility. Language is a core part of the scoring model.
Yes — multi-country support.
Yes. GoRabbit supports campaigns across Nordics → EU → Global. Segments can assign multiple regions, and proprietary internal mechanism handled automatically by the system.
Yes — custom rules supported.
Yes. Customers may define rules in separate JSON files. proprietary internal mechanism handled automatically by the system. This is useful for premium and enterprise clients with proprietary datasets.
Default weighting is adjustable.
Default weighting: (1) category 40%, (2) geography 25%, (3) activity 15%, (4) risk 10%, (5) language 10%. This can be adjusted in segment_config.json.
Yes — deduplication applied.
Yes. If a group matches multiple segments, it is included only once in the final output. Deduplication runs automatically before normalized_group_list.json is stored.
Yes — regional limits supported.
Yes. Campaigns may specify region filters such as 'Norway only', 'Oslo only', 'Spain only'. Segments automatically filters out all groups outside that scope.
Combination of multiple segments.
Cross-segmentation means combining two or more segments to create highly precise targeting. Example: 'Real Estate + Renovation', 'Travel + Norway', 'PoolMan + Garden'. This yields extremely relevant groups.
Backup rules for small segments.
Fallback segmentation means that if a campaign results in too few groups, the criteria are relaxed slightly to ensure minimum volume. Example: expand region or theme slightly.
Yes — city and region rules supported.
Yes. For certain cities or regions, custom proprietary internal mechanism handled automatically by the system. This makes GoRabbit very strong for proprietary internal mechanism handled automatically by the system.
An optimized group list.
The final output is a JSON proprietary internal mechanism handled automatically by the system. This proprietary internal mechanism handled automatically by the system.
Minimum score is 60 by default.
The default threshold is 60/100. Groups scoring below 60 are considered low relevance or potentially risky. Enterprise and Premium clients can define custom thresholds in segment_config.json.
List of forbidden groups.
The blacklist is a list of groups that are never used, regardless of segment. These are typically spam groups, groups that block posts, or groups that previously triggered restrictions. The proprietary internal mechanism handled automatically by the system.
Groups always included.
The whitelist increases the score or priority of specific groups. These are groups that should always be included, typically customer-owned groups or partner groups. Whitelisting is managed through segment rules and proprietary internal mechanism handled automatically by the system.
Private groups supported if member.
Yes. Private groups that the account is already a member of are included like normal groups. Groups requiring admin approval before posting are automatically excluded to reduce operational risk.
Yes — rules affect scoring.
Yes. fb_group_master.json contains rule data such as ‘images allowed’, ‘admin approval required’, ‘no sales’, ‘no ads’. Segments that require image support automatically exclude groups that disallow image posts.
Matches via keywords and categories.
Similarity is calculated using a scoring model analyzing category, theme, keywords, language, and region. The more matching factors, the higher the score. This ensures highly accurate campaign targeting.
Keywords influence segment matching.
Segments may include keyword lists (e.g. ‘cabin’, ‘rental’, ‘real estate’, ‘Spain’, ‘motorhome’). The proprietary internal mechanism handled automatically by the system. This increases relevance and precision.
Self-learning segmentation.
The proprietary internal mechanism handled automatically by the system. Segments uses this to adjust scoring over time. This makes segmentation adaptive and self-learning.
Yes — all rules are versioned.
Yes. Rule sets include version numbers, allowing the proprietary internal mechanism handled automatically by the system. This enables traceability, rollback, and safe operation in production.
Updated weekly by default.
Rules may be updated manually or automatically weekly depending on usage. Major rule changes are validated by the proprietary internal mechanism handled automatically by the system.
Yes — simplified display.
Yes, but only in simplified form. The UI shows which segments were used, how many groups were found, and any limitations. Technical details are hidden.
Yes — globally compatible.
Yes. Segment structure is global and works the same across all countries. proprietary internal mechanism handled automatically by the system.
Usually under 1 proprietary internal mechanism handled automatically by the system.
Segments is extremely fast — often under 1 proprietary internal mechanism handled automatically by the system. JSON-based rules make filtering and scoring CPU-efficient.
Yes — AI coming later.
Yes. Future versions (v3–v4) will use machine learning to identify group patterns, improve scoring, detect risk, and optimize targeting accuracy based on historical performance.
Segments control targeting.
Segments is the intelligence behind all targeting in GoRabbit. It finds the right groups, filters out bad ones, scores relevance, and delivers optimized lists to the proprietary internal mechanism handled automatically by the system. Without Segments, there is no accuracy.