Inman

Let them eat dough

The trouble with Internet mortgage leads is that people prefer to eat bread not dough. The average Web lead is not cooked properly; it’s just a name, not even a lead and certainly not a prospect. Here’s what I mean:

The average Web lead converts (to funding) at less than 3 percent. This low yield is not compatible with the way most lenders operate. These leads are too cold. A typical inbound call converts at 10 percent to 20 percent for mortgage brokers. If one in 10 offline leads is a winner, why buy Web leads that take 30 calls before closing one deal?

In the mid-90s, my company, INPHO, thought we were selling bread, but we were selling warm dough. We faxed and e-mailed leads to top lenders–GMAC, Bank of America, Countrywide and Chase–that resulted in closed loans at a $600 acquisition cost. So far so good. That was about half of their cost. But the conversion rate was just 3 percent. That was not compatible with the infrastructure and expectations of the consumer-direct operations of these top lenders. My leads converted at 3 percent, but lender practices were geared to 20 percent conversion. That’s a big gap! If they modified their business practices to my model, all was well. But anyone who deals with large institutions knows that little guys win when they adapt to big guys, not the other way around.

So, fast forward to 2004. The same problem occurs with Internet leads: low yield, big gap. Most lenders are burned out on Internet leads. Internet leads are like dough, a product released prematurely.

Stand back for a moment as I illustrate how most Internet lead-generation schemes work, to demonstrate that they are not ready for prime time in 2004.

A Web site visitor takes 90 seconds to populate a lead capture form. There’s no money down and it’s a blink of an eye to hit “submit.” This suggests low commitment to me. This person has a one in 500 chance of executing a loan application in 90 days. Further, the Web site lead generator tries to give that lead to multiple competing lenders. So, now the chance for success drops to one in 2,000. Unless the lender’s handling of the lead is automated or based on labor costs in Bangkok, this lead is expensive. Such low conversion rates burn out the loan officers or call-center reps who handle them. Somebody needs to bridge the gap between lead generators and lead users. Warming up these leads is a job for sophisticated analytics technology, not people.

There are two solutions to the problem of cold, doughy mortgage leads off the Web: a bad one and a good one.

The bad solution is for the lead buyer to adapt its business practices. Remember what I said about big firms adapting to small ones?

A small proportion of the nation’s lenders and real estate brokerages have set up specialized units to handle these leads. They can convert low-yield leads into high-conversion prospects. For real estate brokerages it is a RELO-like operation with a low-cost structure, but take away the expensive high-touch customer service and hand-holding of the traditional RELO department. Substitute automated e-mail response technology for the high-cost neighborhood savvy RELO counselor, or use a low-cost team that simply stays in touch with the leads until they warm up. When this “RELO-lite” unit has isolated cold leads from the list, only then should the resulting warm leads pass to the traditional agent.

What yield or conversion level should leads reach before they are turned over to your existing sales channels? The answer is simple: Internet leads need to exhibit the same conversion metrics as traditional leads in your channel today. Your current conversion rate is the standard that Web leads should meet. If not, you should pass on those leads. Some lenders and Realtors have succeeded as they set up specialized units. Don’t try to handle Web leads in your regular channels. You will burn out your people.

I mentioned that trying to bridge the gap between Internet leads and traditional channels was a bad idea. Here’s the good solution: lead vendors, adapt to thy customer. And repeat after me:

ATTRACT
REGISTER
INCUBATE
DETECT through analytics
CONVERT

If your Web site or your lead vendor’s Web site does not deliver these five steps, the leads are not ready for prime time. Go back to the drawing board. Here I focus on incubation and detection. Call me if you want to talk about the other steps.

Most site visitors are not ready to transact when they hit a Web site. They need to be warmed, turned like eggs in the nest and incubated until they are ready. How do you stay engaged with them? How do you detect when they move into the buy zone and get hot? That’s incubation and detection!

None of this is new thinking. With StopRunoff.com, my goal is to take the best of agent behavior and embed it in the software that drives a Web site. A good loan officer or Realtor knows how to incubate leads. Interestingly however, neither can really afford to incubate leads for more than 90 days. They need warm prospects that are right on the edge of becoming customers. 

In lending and real estate, most site visitors are not ready to transact at that time. They want to get some information, but remain anonymous. You need to register them, and convert them from anonymous users–freeloaders on your valuable content–to known marketable registrants. MLS sites register prospects at under 2 percent; home-price sites register users at rates in the teens. If you do register them, then what? Incubation. You need to take that warm dough and bake it slowly until it is ready to come out of the oven. 

Incubation means staying engaged with registrants until they are ready to transact, and keeping them from leaking to a competing lender or broker. This is especially critical for online or consumer-direct (call-center) players, because there lurks an agent or loan officer ready to poach your prospect merely by their local presence. So incubation is preventing leakage, as well as reminding prospects that you are there and ready to serve. 

Detection refers to knowing when the registrant moves from being a tire kicker who’s curious about listings, rates, etc., to being a hot prospect in the buy zone. In lending, the first bank to contact the borrower captures the application 65 percent of the time, so timing is critical. As soon as a prospect moves into the buy zone, you want to initiate a solicitation call. If you call too early, you run the risk of irritating the prospect. And you waste your sales resources and risk Web-lead burnout. Timing is critical because the borrower’s window for selecting a lender is tiny. Borrowers are receptive to lender solicitations for just 21 days! While looking for a home, borrowers become interested in rates and programs, defining their own needs and sifting through lender programs. And 21 days later they make a selection. Game over.

Detection of the buy zone is critical. Here’s how we do it with StopRunoff.com:

The ever-vigilant sentry in StopRunoff’s incubation and detection black box looks at three inputs:

  • Registration information disclosed by the consumer

  • Click-stream data observed–as the user surfs our home-price service

  • Response e-mail feedback–where did users click?

And these three inputs yield three outputs:

  • Urgency: are you 30-90 or more days out from needing to close a loan?

  • Product interest: what loan do you need? Purchase? Refinance? Equity loan/line?

  • Segmentation: information-hungry, first-time home buyer? Or rate-sensitive move-up buyer?

Suppose a registrant discloses that he/she lives in an urban condo, say in Boston’s Back Bay. But your software detects a pattern of searches in a single-family suburb. Bingo. You have found a purchase borrower plus a sell-side opportunity for the condo, and a buy-side opportunity in the suburbs. Urgency is revealed in search patterns such as time of day (10 p.m. has more weight than 10 a.m.) or an increased frequency of searches (a dozen searches in a week after months of only occasional searches). And segmentation is revealed in registration data (“Are you a first-time buyer?”) as well as e-mail response–do they start by clicking on rates (experienced buyer) or on links that define various loan programs (first-timer)?

So by mining all available data, you can make reasoned judgments about your online prospect’s urgency, loan requirements and segment. These insights let you know when the lead is baked (urgency) and the product and segmentation insights guide the call center’s outbound script.

Steve Kropper is SVP at Equinox, which provides mortgage business process outsourcing, as well as the next generation of customer acquisition and retention services called SmartAcquisition and StopRunoff. The programs focus on strategies for purchase loans and top-30 lenders. Kropper lives in Lexington, Mass., and can be reached at skropper@equinoxco.com; (617) 306 9312.

***

What’s your opinion? Send your Letter to the Editor to newsroom@inman.com.