Duplicate records are a well-understood CRM data quality challenge, and Zoho CRM provides several tools to identify and merge them at the record level. But duplicate images present a different type of problem — the same product photograph attached to multiple records, the same property image appearing under several listings, or the same ID document uploaded across duplicate contact entries. These visual duplicates can go undetected by standard deduplication tools, which compare field values rather than image content.

Zia Duplicate Image Detection addresses this gap by analysing the visual content of images uploaded to CRM records and identifying which are similar or identical to images already stored elsewhere in the module. When a potential duplicate image is detected, Zia surfaces the matches — giving your team the information needed to investigate and resolve the duplication without automating a merge that might remove legitimate data.

Where This Feature Adds Value

• Product Catalogues — Identify product records where the same image has been uploaded under different names or SKUs, a common issue when catalogue data is imported from multiple sources or managed by different teams.
• Property and Asset Records — Flag property photographs or asset images that appear across multiple records, helping to consolidate or investigate entries that may represent the same physical item.
• Contact and Lead Deduplication — Where profile photographs or ID documents are associated with CRM records, duplicate image detection can surface contact pairs that standard field-matching may have missed.
• Document Quality Control — Detect when the same scanned document — a signed form, a compliance certificate, or a supporting attachment — has been uploaded across multiple records, helping to maintain clean audit trails.

How Detection Works

When an image is uploaded to a configured field, Zia compares it against the existing image database for that module using visual similarity analysis rather than file-name or file-size matching. This means near-duplicate images — such as the same photograph at different resolutions, cropped differently, or with minor colour adjustments — can still be detected as potential matches.

When Zia identifies a visual match, it surfaces up to 10 duplicate samples for that image, displaying them alongside the current record so that your team can review each potential duplicate and decide on the appropriate action. Zia does not automatically merge or delete any records — the detection is advisory, and all resolution decisions remain with the user.

Configuring Duplicate Image Detection

1. Navigate to Setup > Zia > Vision > Duplicate Image Detection. This is the central configuration area for the feature within your CRM setup.
2. Select the target module and image field. Choose the module where duplicate detection should be active and the specific image field to monitor. Rules are applied at the field level.
3. Enable the rule and save. Once activated, Zia will analyse images uploaded to the configured field and surface matches within the module’s existing image library.
4. Review detections from the record view. When a potential duplicate is detected, the similar images are surfaced directly on the record, allowing the reviewing user to take action — such as investigating the matched records or flagging them for manual deduplication.

Known Limitations

• Maximum 10 duplicate samples per image. Zia will surface up to 10 visually similar images for any given upload. If more than 10 matches exist in the module, only the 10 most visually similar results are shown.
• No automatic merge. Duplicate Image Detection is a detection and advisory tool only. It does not automatically merge, delete, or modify any records — all resolution decisions are made manually by your team.
• Blurred or low-quality images cannot be reliably detected. Images that are heavily blurred, very low resolution, or otherwise of poor quality may not be identifiable by the visual similarity analysis, resulting in missed matches or inconclusive results.
• Minimum user requirement. This feature requires a minimum of 20 CRM users in the account — it is not available to smaller organisations below this threshold.

Availability