We counted how often Anthropic's own 19 ads (those with a full description) state each requirement. This shows the filters the ads apply before anyone reads a resume. We have no data on why candidates are rejected after applying, and we do not guess.
Years of experience. 6 of 19 ads (32%) state a number of years. The median is 8 years; 83% ask for 5 or more and 17% for 10 or more.
Education
Sort:
Requirement ⇅
Ads that state it ⇅
Share of ads ⇅
Bachelor's degree
19
100%
PhD or doctorate
2
11%
'or equivalent experience' accepted
1
5%
Master's degree
1
5%
Working conditions
Sort:
Requirement ⇅
Ads that state it ⇅
Share of ads ⇅
Nights, weekends, holidays or on-call
2
11%
Skills and tools named most often
Sort:
Requirement ⇅
Ads that state it ⇅
Share of ads ⇅
python
12
63%
machine learning
9
47%
kubernetes
5
26%
docker
3
16%
excel
3
16%
c++
2
11%
aws
1
5%
gcp
1
5%
java
1
5%
fpga
1
5%
linux
1
5%
How to stand out
First, cover the basics the ads ask for most (see the requirements above: python (63%), machine learning (47%), kubernetes (26%), docker (16%)). Then look at what separates higher-paid ads:
Among the 19 ads that publish a pay range, we compared the median pay of ads that name each item with ads that do not. This is a correlation in the ads, not a promise: an item can pay more because it comes with senior roles.
Sort:
Item named in the ad ⇅
Ads naming it ⇅
Ads not naming it ⇅
Median pay midpoint with it ⇅
Difference vs ads without it ⇅
python
12
7
$600,000
+$0
kubernetes
5
14
$600,000
+$0
machine learning
9
10
$445,000
-$155,000
We describe only what the ads state. We do not know how recruiters rank candidates.
Same role elsewhere
Where else the same role family is open, with the pay each posts: