We built a health score for your job search. Here's what it actually measures.

Job trackers show you a board. They don’t tell you if your search is working. We built a health score that does, and here’s the thinking behind it.

---

The problem with kanban boards

Every job tracker gives you a kanban board. Columns for Applied, Phone Screen, Interview, Offer. You drag cards between columns and feel organized. The board looks clean. You’re on top of things.

Except the board doesn’t actually tell you anything about whether your search is going well.

You could have 40 cards in the Applied column, all of them two weeks old with no response, and the board would look exactly the same as 40 cards that were added yesterday. You could be converting interviews to offers at a 50% rate or a 0% rate, and the board wouldn’t surface either number unless you counted manually. The board shows you where things are. It doesn’t show you where things are going.

We kept running into this while building JobHawk. The tracker part worked fine. But when we used it ourselves, we’d open the board, see our applications neatly organized, and close the tab without learning anything we didn’t already know. The board confirmed our memory. It didn’t challenge it.

That’s when we started thinking about what a job search tool should actually measure.

What “healthy” looks like in a job search

Before we could build a score, we had to define what we were scoring. And “healthy job search” is harder to pin down than it sounds.

It’s not just volume. Sending 30 applications a week sounds productive, but if none of them advance past the initial screen, it’s busy work. It’s not just conversion rates either, because early in a search you might have great conversations from a small number of applications, and that’s fine.

We landed on three signals that, taken together, give you a reliable picture of how your search is going.

The first is pipeline momentum. Are applications moving forward through stages, or are they piling up and going stale? A search with 10 active applications where three have interviews scheduled is healthier than a search with 50 applications where none have moved in two weeks. We measure this by tracking how long applications sit in each stage and whether the overall flow rate is positive (more things advancing than stalling) or negative.

The second is stage conversion. What percentage of your applications get responses? What percentage of phone screens become interviews? What percentage of interviews become offers? These numbers tell you where your funnel is breaking. If your application-to-response rate is below 5%, the problem is probably your resume or your targeting. If your interview-to-offer rate is low, the problem is prep. The health score doesn’t just flag that something is off; it points to the stage where it’s happening.

The third signal is activity consistency. Job searches fall apart when people go through cycles of intense activity followed by silence. You send 20 applications on Monday, then nothing for a week because you’re waiting to hear back. The waiting creates anxiety, which creates avoidance, which makes the next burst of activity feel harder. Consistent, moderate activity (a few applications a day, regular follow-ups, steady prep for upcoming interviews) produces better outcomes than sporadic bursts. The score tracks this rhythm and flags when you’re trending toward the boom-and-bust pattern.

How the score works

The health score is a single number that combines all three signals. It updates daily based on the state of your pipeline.

We deliberately kept it simple. One number, not a dashboard of 12 metrics. The goal is to open JobHawk, see your score, and know within five seconds whether your search needs attention or is on track. If it needs attention, the score links to the specific signal that’s dragging it down, so you know what to do about it.

The scoring weights aren’t equal. Pipeline momentum matters more than activity consistency, because a stalling pipeline is a more urgent problem than an irregular schedule. Stage conversion matters more in the middle and late stages of a search, when you have enough data for the percentages to mean something. Early on, the score leans more heavily on momentum and consistency, because you don’t have conversion data yet.

We went back and forth on whether to show the underlying numbers or just the composite score. We ended up doing both: the score is the headline, and the three component signals are one tap below it. Most days you’ll glance at the number and move on. When the number drops, you’ll dig into the components and see exactly what changed.

The design decisions we argued about

A few choices were harder than they sound.

Should the score go from 0-100 or use a simpler scale? We tried a five-star rating, a letter grade (A through F), and a 0-100 number. The 0-100 scale won because it’s granular enough to show change over time. Going from a B to a B doesn’t feel like anything. Going from 72 to 68 does. Small changes in the number keep you aware of trends before they become problems.

Should a declining score trigger notifications? We decided no. The score is there when you open the app. Sending you a push notification that says “your job search health is declining” felt like the opposite of what we want the product to be. JobHawk is supposed to reduce anxiety, not add to it. If you’re checking in regularly, you’ll see the trend. If you’re not checking in, a notification about a declining score is more likely to make you avoid the app than to bring you back.

Should stale applications count against you? Yes, but with a grace period. An application with no response after five business days starts counting as stale. After 15 business days, it’s effectively dead weight in your score. We picked those numbers based on recruiting industry data on response timelines: most companies that are going to respond will do so within two weeks. Keeping a three-week-old application in “Applied” and hoping for a response is understandable, but the score shouldn’t pretend it’s a live prospect.

What we learned from beta users

The most common reaction from beta users was that the score told them something they already suspected but hadn’t wanted to look at directly. One user described it as “the scale in your bathroom.” You kind of know whether things are going well or not, but having a number makes it concrete.

The second most common reaction was that the stage conversion breakdown changed how people spent their time. Users who saw a low application-to-response rate started spending more time on their resume and less time sending applications. Users who saw a low interview-to-offer rate started using the interview prep features more seriously. The score didn’t tell people what to do, but it told them where to look, and that was enough.

We also learned that the activity consistency signal needed tuning. Our initial version penalized any day with zero applications, which made the score drop over weekends and during weeks when users had interviews (and weren’t applying to new roles because they were focused on prep). We adjusted it to track a rolling seven-day average and to treat interview prep activity as equivalent to application activity. Your search isn’t stalling if you spent Tuesday prepping for Wednesday’s interview instead of sending applications.

What we’re building next

The current health score is retrospective: it tells you how your search has been going. The next version will add a predictive element. Based on your current pipeline state and historical conversion rates, it will estimate how many weeks until your next likely offer, and what you could do to shorten that timeline.

That’s a harder problem, and we’re not going to ship it until it’s accurate enough to be useful. A prediction that’s wrong is worse than no prediction at all, because it sets expectations that don’t materialize. We’re testing it internally now, and we’ll write about the results when we have them.

If you’re in the middle of a job search and want to try the health score, the free tier includes the composite score and the three component signals. The paid tiers add historical trend data and the detailed conversion analytics.

Try JobHawk free →