cancel
Showing results for 
Search instead for 
Did you mean: 

Fun with Credit Karma

tag
bs6054
Valued Contributor

Re: Fun with Credit Karma


@navigatethis12 wrote:

They determine this by what other people's profiles looked like on approval. They do not know the underwriting standards for any of the cards listed.


How is that tracked?  If you apply through their site, then they can get your current profile and see if that card shows up in a month or so, but do they do that?

If so, unless the sample slze is very small, it sounds like they could be pretty accurate," 7 out of 10 profiles that looked like yours (in ways we measure) got approved."

Message 11 of 27
tcbofade
Super Contributor

Re: Fun with Credit Karma


@bs6054 wrote:

@navigatethis12 wrote:

They determine this by what other people's profiles looked like on approval. They do not know the underwriting standards for any of the cards listed.


How is that tracked?  If you apply through their site, then they can get your current profile and see if that card shows up in a month or so, but do they do that?

If so, unless the sample slze is very small, it sounds like they could be pretty accurate," 7 out of 10 profiles that looked like yours (in ways we measure) got approved."


There in lies the problem.  What are they measuring?  How many people with blonde hair applied?  How many persons that are married to goats?

 

Whatever they ARE measuring, it isn't accurate.  Smiley Tongue

02/01/24 Fico 8: EX 757, EQ 803, TU 783.
Fico 9: EX 760 12/16/23, EQ 790 02/04/24, TU No idea.

Zero percent financing is where the devil lives...
Message 12 of 27
navigatethis12
Valued Contributor

Re: Fun with Credit Karma


@bs6054 wrote:


How is that tracked?  If you apply through their site, then they can get your current profile and see if that card shows up in a month or so, but do they do that?

If so, unless the sample slze is very small, it sounds like they could be pretty accurate," 7 out of 10 profiles that looked like yours (in ways we measure) got approved."


Well the links are most likely referrals so they probably are told if someone is approved or not. I do not see when writing a review how you can input what score one had when applying. Some people leave it in the review so they may go by that, or they track it like any other referral website.

Message 13 of 27
Anonymous
Not applicable

Re: Fun with Credit Karma

It was eerily accurate for me. 

 

I had one card at the time that was 3 months old $300 Cap Plat. 

 

I had NO other credit history at all. 

 

It said my odds of approval for Amex Delta were very good and green bad. I applied for both.....Got the Gold, didnt get the green

 

IT said my odds for Discover it and citi forwward were very good and I got both. 

Message 14 of 27
bs6054
Valued Contributor

Re: Fun with Credit Karma

Well, presumably they are measuring the information in the TU report.    Now the lenders get info that might not be supplied in the CK profile (income, years at work, years at address, these can be input into CK but don't know who does).  Another difference of course would be if a lender pulls from EX or EQ.  So one test, are CK predictions more accurate on TU pullers!

 

But really, in the TU case (+ the optional data) they really could have very similar data to what the issuers see, and so could collect meaningful data.

Message 15 of 27
Duncanrr
Valued Contributor

Re: Fun with Credit Karma


@Anonymous wrote:

It was eerily accurate for me. 

 

I had one card at the time that was 3 months old $300 Cap Plat. 

 

I had NO other credit history at all. 

 

It said my odds of approval for Amex Delta were very good and green bad. I applied for both.....Got the Gold, didnt get the green

 

IT said my odds for Discover it and citi forwward were very good and I got both. 


A broken watch is "eerily accurate" twice a day!



Starting Score: EQ 551 TU 548 CK 607on 6/8/12, EX 542(AMEX pull 3/4/12)
Current Score: EQ 808 TU --- EX --- CK 804(FAKO-EX 821, EQ 823, TU 803)
Goal Score: 750

Take the FICO Fitness Challenge
Message 16 of 27
Repo-ed
Senior Contributor

Re: Fun with Credit Karma

Even this site here, suggested some sub-prime cards for me.

5/2012: 560 credit scores across the board
12/2014: 750+
3/2017: 780+
11/2019: 833
2/2023: Experian via Chase United Explorer CC pull - 891
Message 17 of 27
bs6054
Valued Contributor

Re: Fun with Credit Karma


@Repo-ed wrote:

Even this site here, suggested some sub-prime cards for me.


But was it pulling your credit report when it did that?  To me, that's the potential advantage of CK, as compared to whogavemecredit for example!

Message 18 of 27
Anonymous
Not applicable

Re: Fun with Credit Karma


@Duncanrr wrote:

@Anonymous wrote:

It was eerily accurate for me. 

 

I had one card at the time that was 3 months old $300 Cap Plat. 

 

I had NO other credit history at all. 

 

It said my odds of approval for Amex Delta were very good and green bad. I applied for both.....Got the Gold, didnt get the green

 

IT said my odds for Discover it and citi forwward were very good and I got both. 


A broken watch is "eerily accurate" twice a day!


Sorry my last line was cut off. I also was told I had very poor odds for the Continental Mastercard and citi simplicity and was rejected for both of those (which are lesser cards than the three I was approved for - far lesser in the case of the continental card)

Message 19 of 27
Revelate
Moderator Emeritus

Re: Fun with Credit Karma


@bs6054 wrote:

@navigatethis12 wrote:

They determine this by what other people's profiles looked like on approval. They do not know the underwriting standards for any of the cards listed.


How is that tracked?  If you apply through their site, then they can get your current profile and see if that card shows up in a month or so, but do they do that?

If so, unless the sample slze is very small, it sounds like they could be pretty accurate," 7 out of 10 profiles that looked like yours (in ways we measure) got approved."


This wouldn't be hard to do and I suspect the sample size is quite large actually.

 

They might also be able to mine some data out of simple changes to the report; we know limit and APR ranges here on many credit products, it's not unreasonable that they know as well and could simply look at the lines that are being offered and make some qualified judgements based on the reported information.  This wouldn't be as accurate but if they're looking at ballpark numbers and given that BCE/P for example are well assumed to be the same underwriting criteria.

 

For that matter, I think even I put some ancillary information about my tradelines regarding AF's, APR's, and card type in some cases just for giggles to see if it affected the the recommended cards in it; oooh sneaky now that I think of it: their data is actually probably pretty good if I'm anywhere close to representative in that.




        
Message 20 of 27
Advertiser Disclosure: The offers that appear on this site are from third party advertisers from whom FICO receives compensation.