Your tracker shows your history. Shared insights show everyone's. Every opt-in user contributes anonymized outcome data (no resume text, no names, no companies). Once the pool has enough data for a role category, you see patterns. "For senior frontend roles, resumes missing TypeScript have a 67% lower interview invite rate." Real numbers, real corroboration.
3 steps. Each one runs in seconds. Scroll in order.
A single toggle on your profile. Off by default. You can flip it on or off at any time and nothing new gets contributed while it is off.
When you mark a job as denied, offer, accepted, or interview-invite, the system writes an anonymized row: job title (normalized), experience level, which keyword categories were present, what outcome happened. Never resume text, never the company name, never anything identifying.
Once the pool has ten or more records for your role category, you see patterns specific to it. Below that threshold we show general CBT-informed tips instead of noise from tiny samples.
One real example, same input reshaped two ways.
You are a senior frontend engineer. You have logged three rejections this quarter. Keyword matches ranged 70% to 85%. You opted in to shared insights.
For Senior Frontend Engineer roles in your level band: - Resumes missing TypeScript had a 67% lower interview invite rate (n = 142). - Roles listing GraphQL and Apollo together had a 2.1x interview rate when both appeared in the resume (n = 38). - Testing skills (Jest OR Playwright OR Cypress) correlated with a 1.4x invite rate (n = 218). Your own rejections: 2 of 3 were missing TypeScript. Worth a resume pass.
The mechanics nobody should have to guess at.
The shared insights table never stores a user ID, resume text, company name, or free-text notes. The schema is keyword categories, role normalization, experience level, and outcome. That is it.
No insight shows below n >= 10 for the slice. Small-sample noise is worse than no insight at all. Below the threshold we fall back to general advice.
Older data weights down. The hiring market changes. An insight pulled mostly from 2024 should not carry the same weight as one pulled mostly from the last 90 days. Recent outcomes carry more signal.
You toggle it on. You can toggle it off and from that moment nothing new contributes. Past rows that already went to the shared pool stay (they are anonymous) but your future outcomes stop flowing.
Honest limits read as trust signals. Hiding them does the opposite.
Normalized job title, experience level, which of a fixed set of keyword categories were present in your resume, and the outcome (interview invite, offer, denied). No resume text, no company name, no notes, no user identifier.
Yes. The toggle is reversible. From the moment you turn it off, nothing new is contributed. Previously contributed rows stay in the pool because they are not tied to you, but nothing new flows.
Once the pool has at least 10 rows for your role category and experience band. For common titles this happens quickly; for niche titles it may take longer.
Yes, free accounts see the top insight for their role. Pro accounts see the full breakdown and historical trends.
Industry reports tell you the average time-to-offer for a whole field. Shared insights tell you which specific skills correlate with callbacks for your exact title and level, using outcomes from people using the product today.
Mark a job as denied. Get an AI read on the likely gap, what the resume missed, and what to do differently next time.
Matched and missing keywords named exactly. No opaque score, no black-box percent.
Every saved job gets its own resume, rewritten to mirror the posting. ATS-safe, zero fabrication.
Create a free account in under a minute. First job tracked, first tailored resume, and first keyword breakdown all happen inside the onboarding flow.