A note from Credit Karma's Founder and CEO

[Barry]

Hello All,

I received a request from Credit Karma's Founder and CEO, K. Lin (Forums screen name: CKFounder), to correct some misinformation that's been written in the FICO Forums about Credit Karma scores and to clarify some of the most commonly misunderstood aspects of Credit Karma scoring.  I appreciate that he's come to us with this information.

-Barry

---

Hello All. I have been tracking the discussion about our service for some time. Given some of the comments, I thought it would be a good idea to come into the lion's den and address some of comments. 

Are our scores valuable?

Contrary to many of the comments, all credit scores are highly correlated. One large bank study suggests that all credit scores, including ours, are 97.5% correlated. While the ranges may be different and there will always be difference between the models and bureau data, the end products are very similar and most importantly, consistent. 

Users can misinterpret our information and intent. 

Credit Karma strongly believes in access, education, and transparency. Many of our tools are designed to shed light on the credit scoring process. However sometimes users will read our data differently. For example, a recently myFico discussion refers to how Credit Karma recommends they have at least 21+ accounts. (http://ficoforums.myfico.com/t5/Credit-Cards/Credit-Karma-says-you-should-have-atleast-22-credit-acc...). If you look at our data, we are showing the correlations of credit scores based on a sample of over 400,000 credit files. More accounts and more breadth mean higher credit scores, on average. We don't suggest you rush out and open 21 accounts. This is no different than FICO's break out of: Recent Inquires, Breadth of Credit, On-time Payments. Etc. Our intent is to show the differences in a simple way. http://www.creditkarma.com/question/has-credit-karma-helped-anyone-improve-there-score. 

Are the score differences meaningful?

With almost any score comparison, there will be differences. Some users note that scores are different by as much as 50 points. This certainly is true. However there is quite a bit of statistical error in most comparisons. Consumers don't control for bureaus or time. Even FICO states that over 30% of consumers see a 50 point difference in scores between bureaus. Regardless of the absolute difference, the movement in one score will 97.5% of the time be a movement in another score. Our score is a guide just like all the other scores in the marketplace. 

The only score that matters. 

FICO is most certainly the leader in the mortgage space. However it is disingenuous to suggest that purchasing a FICO score is the only score that matters. FICO produces many scores for banks (e.g. FICO Classic vs. FICO 98), they have many more for auto and other industries. FICO only sells one score to consumers, so the score you buy may not actually be the score that the banks use. This is not a knock on FICO, just the realities of the marketplace.

I suspect I won't be able to change many of your opinions but I thought it would be good to share our point of view in a forum of savvy credit users. Credit Karma is looking to build a useful and truly free service. I welcome your thoughts and feedback on how to make our service more useful.

Best Regards,
K. Lin

Founder & CEO, Credit Karma 

[SocalMark]

But why soooo many different ways to do one thing...Give me my score

[haulingthescoreup]

Well, in the end, I want to have some idea what scores my lenders are looking at. And while it's very true that not all use FiCO's, or they use different versions of FICO's, or they use home-grown versions, I'm not aware of any lenders who use Credit Karma scores. There might well be, but I can't think of any now.

Perhaps CK scores track FICO scores better than do other non-FICO scores. Back when my scores were changing a lot, CK wasn't around. But I did have TrueCredit, and their non-FICO scores were pretty humorous in their failure to match the movements of my FICO's.

Credit monitoring services are great where they let you see all three reports and track the changes. But I'm pretty skeptical about the value of their scores, and after going on 4 years of this, I haven't seen anything to make me think differently.

Glad you waded into the fray, though. 

edit, lol, to fix the wrong smiley

[plasticman]

CKFounder,

Thank you for taking time to post here. Hopefully it will not be your last as most information disseminated  here credit related is of some value.

With respect to your request for suggestions regarding your web site, one suggestion immediately comes to mind.

My suggestion is for Credit Karma to indicate the average age of all open and closed accounts. It appears Credit Karma's report only advises the user of the average age of all open accounts.

This indicates to me that your scoring model  may weight the average age of open accounts, as opposed to open accounts and closed accounts, much differently than FICO...dunno. When I pull my FICO scores it indicates the average age of open and closed accounts and that is the number which is of importance to me in terms of average age of accounts reporting.

In any event, nowhere does Credit Karma inform the user of the average age of open and closed accounts. Please correct me if I am mistaken.

Your feedback will be appreciated.

[Anonymous]

I understand your point. I kind of feel like your beating a dead tree, but your word counts. I understand the difference between the scores. I use your site and its useful and free. I use it all the time for the dgf and I. The score might not be a fico score but atleast it keeps a record of all your past scores. I like to see that go up. I pay fico money and they don't even give me a graph with historical scores. Atleast not for doing it once in a while.

I must say however... I can't believe everyone is rolling over like a dog to get pet here... Where I got my education we had to CITE our references. I am almost astonished you didn't. I'm not just going to take your word for it.

Please cite "One large bank study" for me please.

[plasticman]

---

@Anonymous wrote:

 I must say however... I can't believe everyone is rolling over like a dog to get pet here... Where I got my education we had to CITE our references. I am almost astonished you didn't. I'm not just going to take your word for it.

Please cite "One large bank study" for me please.

---

I do not understand your assertion regarding this thread. Is attempting to engage in a civil discourse acting like a pet dog?

Please enlighten me.

[Anonymous]

Your request is quite straightfoward. When building our credit report card, we struggled with keeping things concise while providing context and detail. I'll forward your request to the product team. I suspect we could get it into a future release if we see a correlation with the score. On some level, I think the correlative value may be lower since the data may be co-linear with many of the other attributes. 

[Anonymous]

It is an internal BofA study. Obviously I can't publish it but I'm sure anyone in the know would call me out if I was not being truthful.

[plasticman]

---

@Anonymous wrote:

Your request is quite straightfoward. When building our credit report card, we struggled with keeping things concise while providing context and detail. I'll forward your request to the product team. I suspect we could get it into a future release if we see a correlation with the score. On some level, I think the correlative value may be lower since the data may be co-linear with many of the other attributes. 

---

Please advise of the feedback you receive from your product team and post the response on this thread.

Thank you.

[Anonymous]

My experience with CreditKarma is less than enchanting.

CreditKarma provides TransRisk FAKO scores.

I’ve been tracking TransRisk scores for both DH and myself for ten months.

DH’s TransRisk score was 100 points off his FICO score when we originally pulled.

Since that time, his TransRisk score has gained 70 points; his FICO score has lost ten points during the same time period.

I’ve watched TransRisk scores jump up, and have pulled FICO scores to look similar changes, but they didn’t exist.  In fact some TransRisk score increases came on the same day as FICO decreases.

For me, my CreditKarma’s TransRisk score (for TU) is 698.  30 points off my TU FICO.  And CreditKarma tells me that at 698, I’m scoring higher than 61% of consumers nationwide.  On the other hand, a 698 FICO would be below the midpoint.

What I’m really finding interesting is on CreditKarma’s website under the “We found some savings for you in the following categories:”

Under Credit Cards, my reported balance is reported correctly as $1.

I’m then told I can lower my payment by $282.

Under Banking, it indicates I have a $10,000 balance – however my bank balances are not listed on credit reports and are not available to CreditKarma.

At least the comment about “Details Needed” under insurance is forthcoming.