Does the amount of content in the Description of my Listing really matter?
The answer is an unequivocal “YES!” and I have done some number crunching to prove this point. I have queried all the expired or completed listings from 2/6/06 to 11/1/07 and counted the number of characters in the descriptions. The left bar represents all the listings that have completed with full funding. The right bar represents all the expired listings that did not reach full funding. The resulting statistics was surprising but understandable.
Here is the raw data:
| Type of Listing | Count | Average Description Size |
|---|---|---|
| Completed Listings with 100% Funding | 31696 | 1676.3 |
| Expired Listings with <100% Funding | 205255 | 1229.1 |
The listings with 100% funding had, on average, 36.4% more content than listings that expired without full funding. This makes a lot of sense: lenders are more comfortable bidding on listings that they have a lot of information about. Lending money to a complete stranger is tough enough, but lending money to a complete stranger with nothing to say is even tougher. There are a lot of factors that you do not have control over in the listing, but what you do have full control of is what you want to say in the description. Take advantage of this opportunity and write in detail what you plan on doing with the loan and prove that you are able to pay the money back. If there is anything negative in the listing such as a delinquency or public record, explain why this has happened to you and how you plan or planned to alleviate it.
In response to NewHorizon’s suggestion in the Prosper forums, I broke down the data by credit grade to see if there was anything misleading.
Here is the raw data:
| Type of Listing | Credit Grade | Count | Average Description Size | % Difference between Funded and Non-Funded |
|---|---|---|---|---|
| Funded | AA | 3198 | 1269 | 23.9% |
| Non-Funded | AA | 2628 | 1024 | |
| Funded | A | 3104 | 1426 | 22.7% |
| Non-Funded | A | 3230 | 1162 | |
| Funded | B | 4008 | 1501 | 18.3% |
| Non-Funded | B | 5758 | 1269 | |
| Funded | C | 5692 | 1611 | 24.3% |
| Non-Funded | C | 12673 | 1296 | |
| Funded | D | 5797 | 1727 | 29.8 |
| Non-Funded | D | 23318 | 1330 | |
| Funded | E | 4581 | 1812 | 44.0% |
| Non-Funded | E | 38735 | 1258 | |
| Funded | HR | 5168 | 2106 | 75.2% |
| Non-Funded | HR | 118288 | 1202 | |
| Funded | NC | 282 | 1681 | 72.4% |
| Non-Funded | NC | 1680 | 975 |
The results are very interesting. Except for the NC group, there is a steady increase in content for funded listings as the credit grades get lower. The overall average of content between funded and non-fundings listings, 36.4%, was indeed askew due to the large number of HR listings where the funded listings had 75.2% more content than non-funded listings. Also, the dates used for this dataset is slightly different than the original: 2/2/06 – 11/3/07.
If you have any requests regarding statistics that you would like to see, please let me know. If it’s feasible, I will try my best to make it happen.
You may also be interested in reading the following posts:
- The Top 1000 5-letter Words in the Description of Funded Listings vs. Non-Funded Listings
- Which day of the week are lenders most active in bidding?
- What time of the day are lenders most active?
- How long should I have my listing for?
- Loan calculator with statistics for all listings from the last past year
- Which day of the week should I end my listing?
- How do I create a successful listing on Prosper?
Glossary for status of listings:
- Completed – The listing ran to completion and funded
- Expired – The listing failed to fund in time
Very cool blog… Keep it up!!!
Hey, my very first comment and it was positive! =)
Thanks. I will try to keep it up.
Hi Knix – great post. I think you mean “100% funding” (more than). In other words, the left bar is borrowers who got funded, and the right bar is borrowers who did not get funded.
Hi Andrew – Yes, you are right. The left bar represents all the listings that were completed with full funding. The right bar represents all the listing that were expired and did not reach full funding. It was tough for me to think of the terms to use but I think “Funded” vs. “Not Funded” should be more concise. Thanks for the suggestion.
Hi, great blog love ur work mate!
Just wondering if you conducted any statistical tests of significance on the findings?
I noticed that there are way more data pts for expired listings compared with completed listings. Perhaps an ‘analysis of variance’ is needed to determine whether the difference in the avg no. of chars is significant?
i’m no expert myself, just something i remember from the stats course i took many years ago.
Thx.
Hi theone,
I automatically assumed that the numbers were considered significant due to the number of listings involved. Although 31696 is small compared to 205255, it is a considerably large pool. Of course, it would still be nice to have some stats on the significance and I will definitely look into providing some. Thanks for your suggestion.
You have been doing some very interesting studies.
I plan to start doing some studies as well since I am a huge fan of numbers. In looking at the difference between Lending Club and Prosper, one of the things that I did not like about Lending Club was the very short descriptions available. Your data backs up what I was feeling while browsing potential borrowers.