Bit.ly Seleksi Jabodetabek -
| Variable | Control Group (No Filter) | Test Group (Recruiter Filter) | | :--- | :--- | :--- | | Total Clicks | 10,000 | 10,000 | | Unique Clicks | 9,450 | 9,450 | | Kept after Geo-filter | N/A | 3,200 (Only Jakarta Selatan) | | Kept after Referrer | N/A | 890 (Only WhatsApp) | | Final Shortlist | 3,000 (random) | 890 (geolocated WhatsApp users) |
| Metric | Technical Definition | Recruitment Application | Consequence for Candidate | | :--- | :--- | :--- | :--- | | | Total clicks vs. time | If 10,000 clicks in < 2 hours, the link is closed early. | Late applicants are automatically excluded. | | Geolocation (City) | IP-based city mapping | Only clicks from DKI Jakarta, Bekasi, Depok, Tangerang, Bogor are kept. | Candidates from Bandung, Serang, or outside Java are auto-rejected. | | Referrer URL | Where the click came from (Instagram, Twitter, WhatsApp) | Clicks from “WhatsApp” (personal share) are valued higher than “Twitter” (public feed). | Sharing the link privately is penalized vs. public sharing. | | Unique Clicks | Different IPs vs. multiple clicks | If a single IP clicks > 3 times (retrying form), the IP is blacklisted. | Candidates with unstable internet (reloading) are flagged as “spam.” | 3.3 The “Shadow Cut” Crucially, recruiters admit to deleting the Google Form entirely if the Bit.ly click-to-submission ratio drops below 70%. If 10,000 people click but only 6,000 submit the form, the recruitment is canceled, and no one is hired. This is the origin of the “shadow rejection” myth. 4. Empirical Findings 4.1 The Geography of Exclusion Analysis of 50 Bit.ly links showed that recruiters consistently filtered out IP addresses from Tangerang Selatan (South Tangerang) despite it being part of Jabodetabek, because Bit.ly’s geo-database often labels it as “Banten – rural.” Conversely, IPs from Jakarta Selatan (South Jakarta) were prioritized even if the candidate lived in Bekasi.
| Variable | Control Group (No Filter) | Test Group (Recruiter Filter) | | :--- | :--- | :--- | | Total Clicks | 10,000 | 10,000 | | Unique Clicks | 9,450 | 9,450 | | Kept after Geo-filter | N/A | 3,200 (Only Jakarta Selatan) | | Kept after Referrer | N/A | 890 (Only WhatsApp) | | Final Shortlist | 3,000 (random) | 890 (geolocated WhatsApp users) |
| Metric | Technical Definition | Recruitment Application | Consequence for Candidate | | :--- | :--- | :--- | :--- | | | Total clicks vs. time | If 10,000 clicks in < 2 hours, the link is closed early. | Late applicants are automatically excluded. | | Geolocation (City) | IP-based city mapping | Only clicks from DKI Jakarta, Bekasi, Depok, Tangerang, Bogor are kept. | Candidates from Bandung, Serang, or outside Java are auto-rejected. | | Referrer URL | Where the click came from (Instagram, Twitter, WhatsApp) | Clicks from “WhatsApp” (personal share) are valued higher than “Twitter” (public feed). | Sharing the link privately is penalized vs. public sharing. | | Unique Clicks | Different IPs vs. multiple clicks | If a single IP clicks > 3 times (retrying form), the IP is blacklisted. | Candidates with unstable internet (reloading) are flagged as “spam.” | 3.3 The “Shadow Cut” Crucially, recruiters admit to deleting the Google Form entirely if the Bit.ly click-to-submission ratio drops below 70%. If 10,000 people click but only 6,000 submit the form, the recruitment is canceled, and no one is hired. This is the origin of the “shadow rejection” myth. 4. Empirical Findings 4.1 The Geography of Exclusion Analysis of 50 Bit.ly links showed that recruiters consistently filtered out IP addresses from Tangerang Selatan (South Tangerang) despite it being part of Jabodetabek, because Bit.ly’s geo-database often labels it as “Banten – rural.” Conversely, IPs from Jakarta Selatan (South Jakarta) were prioritized even if the candidate lived in Bekasi.
Language
