Are the Mortgage Web Sites a Bust?
December 4, 2000, Updated October 10, 2008
This article was something of a forecast, so instead of revising it, I decided to keep it intact and add parenthetical remarks in brackets, commenting on how well the forecasts have turned out.
"I keep reading about mortgage web sites that have gone out of business or are struggling to stay afloat. Early predictions that they would take about 25% of the market within 5 years seem to be far off the mark. How do you explain this, and what’s your forecast?"
In one sense, the mortgage web sites have been a major success. Surveys indicate that more than half of all borrowers explore on-line before they get a mortgage, and the number keeps rising. The internet has become a major information source for mortgage borrowers.
[I would guess the proportion using the web in 2008 is closer to 90%].
The problem for the sites has been that few consumers transact on line. They take advantage of the net to educate themselves, then go to a lender or a mortgage broker to get their loan off-line. Mortgage web sites that depend on transactions to generate revenue are having a tough time.
Probably the most important reason why so few borrowers transact on line is that so few of them have the confidence to go through the process alone. Most want help from an expert who will be accountable to them. The Help Desks and Chat Rooms that some sites offer are not an adequate substitute for a loan counselor with an interest in getting the deal done.
[This continued to be true in 2008, but somewhat less so].
A second reason is that none of the mortgage sites have found an adequate way to deal with market nichification: the division of the market into literally millions of niches, based on characteristics of the borrower, the property and the transaction. Niches carry different interest rates, points, and down payment and other underwriting requirements. To find the niche in which a particular borrower falls, a great deal of information must be collected from that borrower.
Borrowers resist the tedious inputting of information required to place them in the proper niche. People on the net usually have short attention spans, they don’t understand why all the information is needed, and many suspect that it isn’t.
To some degree, therefore, every mortgage site caters to the impatience of users by limiting the information they request. But even a truncated list is too long for many users. And because of the omitted information, some users who do go through the process are placed in the wrong niche by the site’s automated system. The prices and underwriting requirements have to be corrected later, by human beings. This defeats a major purpose of transacting on-line.
[This also continued to be true in 2008]
But remember that the existing mortgage sites are only the first generation. A second generation of sites will emerge in the next few years that will integrate loan counselors into the on-line process. High-tech and high-touch will be combined.
The generation two sites will have three new high-tech features. First, the automated system will place every user in the correct market niche. The technology exists to do this now, but it has not been deployed because of user resistance to long questionnaires. That resistance will disappear, for reasons indicated below.
[In 2008, we were still waiting].
Second, the site will offer on-screen approval. Users would be shown the loan terms at which they have been approved, subject only to confirmation of the accuracy of the information the users have entered.
[One new web site does offer this feature (see Mortgage Grader: A Better Type of Web Site?), but it has some serious shortcomings as well.]
Third, the technology will help borrowers in making the four critical decisions that must be made on every loan: the type of loan, the down payment, the term, and the combination of interest rate and points. Existing sites require users to make their own decisions on these matters, providing little or no help.
[I struck out on this prediction, none of the newer sites have any of this].
The process of collecting the information needed to place the user in the correct market niche, and assistance in making the four critical decisions will be tied together. The flow of information will run in two directions, with the borrower’s answer to one question affecting the next one. Filling out the questionnaire will be a "give and get" process for the user, rather than just a "give". A counselor will be available to move the process forward.
In the multi-lender version, borrowers would select from mortgage brokers who have been approved by the site and screened for proficiency in using the site’s technology. The site would list approved brokers by location and URL. The brokers, however, would be accountable to and paid by the borrower, who would negotiate their fees directly with the broker in advance of service. Brokers could communicate with borrowers by telephone, email, in person or through any combination of these.
Single-lender sites would work essentially the same way, with employees serving as the counselors rather than mortgage brokers.
[None of this had happened by 2008]
This article was something of a forecast, so instead of revising it, I decided to keep it intact and add parenthetical remarks in brackets, commenting on how well the forecasts have turned out.
"I keep reading about mortgage web sites that have gone out of business or are struggling to stay afloat. Early predictions that they would take about 25% of the market within 5 years seem to be far off the mark. How do you explain this, and what’s your forecast?"
In one sense, the mortgage web sites have been a major success. Surveys indicate that more than half of all borrowers explore on-line before they get a mortgage, and the number keeps rising. The internet has become a major information source for mortgage borrowers.
[I would guess the proportion using the web in 2008 is closer to 90%].
The problem for the sites has been that few consumers transact on line. They take advantage of the net to educate themselves, then go to a lender or a mortgage broker to get their loan off-line. Mortgage web sites that depend on transactions to generate revenue are having a tough time.
Probably the most important reason why so few borrowers transact on line is that so few of them have the confidence to go through the process alone. Most want help from an expert who will be accountable to them. The Help Desks and Chat Rooms that some sites offer are not an adequate substitute for a loan counselor with an interest in getting the deal done.
[This continued to be true in 2008, but somewhat less so].
A second reason is that none of the mortgage sites have found an adequate way to deal with market nichification: the division of the market into literally millions of niches, based on characteristics of the borrower, the property and the transaction. Niches carry different interest rates, points, and down payment and other underwriting requirements. To find the niche in which a particular borrower falls, a great deal of information must be collected from that borrower.
Borrowers resist the tedious inputting of information required to place them in the proper niche. People on the net usually have short attention spans, they don’t understand why all the information is needed, and many suspect that it isn’t.
To some degree, therefore, every mortgage site caters to the impatience of users by limiting the information they request. But even a truncated list is too long for many users. And because of the omitted information, some users who do go through the process are placed in the wrong niche by the site’s automated system. The prices and underwriting requirements have to be corrected later, by human beings. This defeats a major purpose of transacting on-line.
[This also continued to be true in 2008]
But remember that the existing mortgage sites are only the first generation. A second generation of sites will emerge in the next few years that will integrate loan counselors into the on-line process. High-tech and high-touch will be combined.
The generation two sites will have three new high-tech features. First, the automated system will place every user in the correct market niche. The technology exists to do this now, but it has not been deployed because of user resistance to long questionnaires. That resistance will disappear, for reasons indicated below.
[In 2008, we were still waiting].
Second, the site will offer on-screen approval. Users would be shown the loan terms at which they have been approved, subject only to confirmation of the accuracy of the information the users have entered.
[One new web site does offer this feature (see Mortgage Grader: A Better Type of Web Site?), but it has some serious shortcomings as well.]
Third, the technology will help borrowers in making the four critical decisions that must be made on every loan: the type of loan, the down payment, the term, and the combination of interest rate and points. Existing sites require users to make their own decisions on these matters, providing little or no help.
[I struck out on this prediction, none of the newer sites have any of this].
The process of collecting the information needed to place the user in the correct market niche, and assistance in making the four critical decisions will be tied together. The flow of information will run in two directions, with the borrower’s answer to one question affecting the next one. Filling out the questionnaire will be a "give and get" process for the user, rather than just a "give". A counselor will be available to move the process forward.
In the multi-lender version, borrowers would select from mortgage brokers who have been approved by the site and screened for proficiency in using the site’s technology. The site would list approved brokers by location and URL. The brokers, however, would be accountable to and paid by the borrower, who would negotiate their fees directly with the broker in advance of service. Brokers could communicate with borrowers by telephone, email, in person or through any combination of these.
Single-lender sites would work essentially the same way, with employees serving as the counselors rather than mortgage brokers.
[None of this had happened by 2008]