At our consultancy agency, we connect both our Microsoft 365 and LinkedIn accounts to Rolodex. Due to recent LinkedIn API restrictions, it is no longer possible to retrieve email addresses or phone numbers via the LinkedIn integration. As a result, we are seeing a significant increase in duplicate contacts: one contact synced from Microsoft with an email address but no photo or LinkedIn profile, and another contact with a photo and LinkedIn profile but without an email address or phone number.

Manually merging over a thousand duplicate contacts is simply unmanageable. However, using a blanket "Merge All" approach is also risky, as different individuals can share the exact same name.

We would strongly benefit from some intelligent auto-merging capabilities within Rolodex.

Suggested Auto-Merge Logic:

If two contacts have exactly the same name, and:

* Contact A has an email address with domain [mycompany.com] but no LinkedIn profile

* Contact B has a LinkedIn profile but no email address, and one of the companies listed in that profile matches (or almost matches) domain [mycompany]

Then, in 99.99% of cases, we can confidently assume it's the same person, and Rolodex should automatically merge those contacts.

Additional Matching Signals:

Another way to increase the probability of a correct match is by analyzing the first interaction timestamps. If the first email interaction and the first LinkedIn interaction with each respective contact happened around the same period (e.g., within a few days or weeks of each other), it is highly likely they refer to the same individual. This correlation could serve as a valuable signal for suggesting or auto-confirming matches.

Enhancing the Manual Merge Experience:

For all other cases, the "Merge & Fix" screen is helpful, but it would be far more efficient if Rolodex displayed a match probability percentage based on string similarity and field comparisons. Currently, we must open and analyze each contacts manually. If Rolodex could display something like "Match Probability: 85%", users could more confidently choose to merge without manual review, saving significant time.

This kind of smart, context-aware assistance would drastically improve the efficiency and user experience for teams managing large, complex contact databases.