MORTGAGE BUSINESS INTELLIGENCE

MORTGAGE BUSINESS INTELLIGENCE
AUTOMATING THE INDUSTRY’S HIDDEN MANUAL PROCESSES
COPYRIGHT © 2015 MOTIVITY SOLUTIONS, INC
INTRODUCTION
The use of technology to expand human capabilities goes back to the dawn of man. The first stone tools were
an attempt to enhance the abilities of prehistoric man, and this movement continued over millennia with the
development of simple machines such as the wheel, the lever, and the pulley. Windmills and waterwheels followed,
and these powered machines dramatically reduced the level of human effort required in production tasks while
increasing output and quality. The advent of computers pushed the automation movement into overdrive, and
today digital technology is an indispensable component of production automation in every industry around the
world.
Production is an inherently time sensitive undertaking, so it’s no surprise that production sectors have the most
prevalent and advanced users of automation on Earth. Businesses whose operations involve a production
component understand that it’s critical to automate as many manual processes as possible, as doing so can
compress production timelines, reduce costs, and improve quality. In fact, the producers in any given industry
that tend to fail are those with substandard automation. They simply can’t compete with larger or more capable
players, all of whom operate at a higher capacity with lower costs, while exhibiting the same or higher levels of
quality. In most production environments, manual operations (human tasks) are simply cost prohibitive.
In some cases automation movements can stall, as they have in the mortgage industry. Initially, the introduction
of computer technology in the 1970s brought transformative automation to mortgage lending. In a classic
example of automating manual processes, calculators and hand typed forms gave way to data entry screens and
laser printed documents. Today, anyone would be shocked to find a typewriter preparing loan documents. An
outdated, manual process is simply an unnecessary expense if it can easily and cost-effectively be automated.
Automating loan origination and document preparation, however, is merely half the battle. The mortgage
industry continues to struggle with an array of time consuming manual processes, largely because they are not
seen as manual. There are a number of hidden manual processes within traditional mortgage operations which,
if automated, can have the same galvanizing impact that lenders enjoyed when they retired their typewriters.
There are four areas where manual operations are presently the standard in the mortgage industry:
• Composition and distribution of spreadsheets
• Interpretation of spreadsheets
• Communications and ad hoc information needs
• Performance Management
The element of time as a commodity in the mortgage industry is well known. It would be difficult if not impossible
to find a mortgage professional who professes to have a mountain of available time each and every day, or any
day for that matter. Even weekends and holidays are often leveraged to catch up on activities not completed
during business hours. Mortgage lending is commonly described as stressful, and the fundamental source of
that stress is simply lack of time. Operations at every level feel hectic, as most professionals rarely feel there’s
enough time to devote to individual tasks or decisions. Amazingly, this has been the accepted norm for so long
that many industry veterans would actually react with suspicion to mortgage lending operations that lacked this
frenetic edge.
WHAT GETS MEASURED, GETS RESULTS.
COPYRIGHT © 2015 MOTIVITY SOLUTIONS, INC
These symptoms are emblematic of rampant inefficiencies. In the case of the mortgage industry, these inefficiencies
appear in the form of excessive manual operations, which in any other industry would have been automated
long ago. They have prevailed in mortgage lending because they’re generally not considered manual, as these
processes employ computers and software (most commonly Microsoft Excel). An understanding of the manual
nature of these processes emerges once lenders are shown how all other production industries have been
automating these manual operations for decades. The equivalent innovation in mortgage lending is termed
Mortgage Business Intelligence (MBI).
MANUAL PROCESS 1: SPREADSHEET COMPOSITION AND DISTRIBUTION
The act of extracting data from a loan origination system (LOS), dropping it into a spreadsheet and then
manipulating it for personal consumption or distribution is the generally accepted method for constructing
mortgage business analytics. In virtually every company, most employees have a spreadsheet program available
to them at their workstation so this activity is often decentralized, and an inordinate number of man hours are
spent composing and distributing spreadsheets organization-wide. Even though most people don’t think of it as
such, this is a manual process which, if automated, can tremendously broaden organizational bandwidth.
This is by far the easiest activity to fully automate, and mortgage companies that have done so have seen
hundreds of man-hours repurposed toward increased production and/or quality initiatives. While business
intelligence users in other industries may be reluctant to replicate the row and column format within their BI
platforms, spreadsheet replication during the first wave of automation has proven to be quite useful in the
mortgage industry. MBI providers understand the time-constrained nature of mortgage business operations, so
while replicating a spreadsheet may not represent best BI practices in other industries, and while it’s arguably
the first of many steps in implementing MBI, even this initial step consistently proves to be of enormous value
within the mortgage industry.
Spreadsheet replication and automation is quite often the first wave of MBI adoption for mortgage lenders.
Eliminating excessive man hours spent on manual analytics opens up time that can be spent getting acquainted
with the advanced analytics models and visualizers that MBI has to offer. MBI systems are often termed “embedded
analytics”, which connotes the fact that these platforms are directly integrated with production data to capture
any/all changes in the database in near-real-time. MBI platforms create a high performance copy of a lender’s
production data, and use this ‘query model’ to send actionable data to the users that need it, when they need it.
MANUAL PROCESS 2: SPREADSHEET INTERPRETATION
The interpretation of spreadsheet data is another process that is rarely considered manual, particularly given
the lack of discernible physical activity (e.g. manual entry/manipulation of data) associated with it. Interpreting
spreadsheets involves the simple act of reading. But business users are never trying to read and absorb an
entire spreadsheet. Instead, they’re scanning the rows and columns, looking for individual data points that will
direct them to friction points or other issues that require their attention. Even though no physical effort is being
expended, we can see this as a manual and automatable process for two reasons. First, it’s being performed by
a human resource, and second, it’s repeatable and predictable.
In the same way that an MBI platform can automatically map production or other data into an onboard row/
WHAT GETS MEASURED, GETS RESULTS.
COPYRIGHT © 2015 MOTIVITY SOLUTIONS, INC
column report, as well as automatically run and deliver this replicated spreadsheet at specified intervals to
any number of specified users, such a system can also be tasked to recognize the patterns and outliers that
business users spend their days searching for within spreadsheets. The BI function that is associated with this
level of automation is termed a “Key Performance Indicator”, or KPI.
Once a user community has automated the composition and distribution of spreadsheets, the next logical step
is to allow their MBI platform to automate the manual, row by row, column by column analysis they’ve been
spending their time on. Allowing an MBI system to take over this activity yields many levels of efficiency in the
form of a KPI (see Figure 1).
Seasoned BI users streamline
their analytics to the greatest
extent possible, striving to create
dashboards and/or other visualizers
displaying KPIs that track all their
critical areas of concern in a single,
easy-to-read view. Some may not
take action until they see red KPIs
appear, and the more proactive users
open up yellow KPIs by “drilling”
into them (accomplished by simply
clicking on the KPI) to reveal lower
level detail around the various tasks
in operations that the KPI is tracking.
There is no faster, more efficient
way to pinpoint operational issues
than having all manual analytics fully
automated and feeding near-realtime KPIs within a readily available
scorecard or dashboard, and no faster
way to uncover causes and effects
than by drilling down into a given
KPI to instantly reveal the related
operational detail.
FIGURE 1 - KEY PERFORMANCE INDICATOR (KPI)
TRENDING ARROW INDICATES
WHERE THIS MEASUREMENT
FALLS RELATIVE TO A PRIOR
SPECIFIED TIME PERIOD, IN THIS
CASE 2 MONTHS AGO
ANY MEASUREMENTS THAT A BUSINESS USER
HABITUALLY LOOKS FOR WITHIN SPREADSHEET
ROWS AND COLUMNS CAN BE AUTOMATICALLY
SURFACED AND PUT INTO CONSTANT VIEW WITHIN
A SCORECARD OR DASHBOARD VISUALIZER
COLOR CODING QUICKLY REVEALS WHERE THIS
MEASUREMENT FALLS RELATIVE TO A GOAL SET
FOR IT. YOU CAN SEE SEVERAL OTHER THINGS
AT A GLANCE AS WELL: IF THE METRIC IS AT OR
ABOVE THE ESTABLISHED GOAL (GREEN) WITHIN
A SPECIFIED TOLERANCE FOR A GOAL (YELLOW)
OR BELOW THE SPECIFIED GOAL (RED), EVEN
WITHOUT READING AND INTERNALIZING THE
ACTUAL MEASUREMENT.
MANUAL PROCESS 3: COMMUNICATIONS & AD HOC INFORMATION NEEDS
An outgrowth of manual analytics is excessive communication between business users. Spreadsheets may be
sent to individuals or groups of users, generating questions for the sender. Business users whose analytics
overlap with those constructed by other business users may be inclined to compare their results to help each
other find data irregularities or other inconsistencies. When users finally pinpoint meaningful data points that
require action, that action most often comes in the form of communication, typically one or more emails sent
individually (i.e. manually) to one or more loan participants. The most time consuming chore of this type is the
ad hoc report: called for by isolated, unique circumstances that require quick data discovery.
WHAT GETS MEASURED, GETS RESULTS.
COPYRIGHT © 2015 MOTIVITY SOLUTIONS, INC
Just as spreadsheet interpretation is manual, repeatable, and predictable, so are the subsequent communications
patterns that this activity necessitates. From originators or processors receiving emails (often including manually
produced spreadsheet attachments that must be opened and manually reviewed again by the recipient)
regarding loans that have specific issues or have slowed down relative to other loans in the pipeline, to closers
receiving messages regarding looming compliance deadlines, these and any number of other events can be
recognized as they happen by an MBI system, which can automatically send predefined email notifications to
internal or external loan participants as needed.
Ad hoc reporting requirements present a unique analytics challenge, particularly to the spreadsheet user, who
must invent a format for the data discovery task and then build the spreadsheet from the ground up. MBI systems
automate this activity to the greatest possible extent by providing users with a ‘data on demand’ grid. Taking
full advantage of the aforementioned ‘query model’, which is a high performance copy of a firm’s production
data, a ‘data on demand’ grid can profoundly streamline the ad hoc data discovery process. This intuitive, easy
to use tool provides unprecedented access to a firm’s entire data store, and business users can typically find the
information they need within one to six clicks of their mouse.
Users can also leverage MBI’s automation functions with an ‘on demand’ grid to automatically repeat a drill
path that a business user may have performed pursuant to a data discovery task. The drill path becomes a data
filter and the results of the discovery can be rendered in the form of a report, and if appropriate, the system can
generate and send this report automatically to a user or groups of users at specified intervals.
MANUAL PROCESS 4: MANAGEMENT OVERSIGHT AND GUIDANCE
One of the more fascinating attributes of BI in any industry is the intrinsic effect it has on production levels
once installed. MBI systems have the same effect on loan operations: their very presence induces reflexive peak
performance. There are two aspects of this ‘automated’ boost in human performance, which can be seen as
automating what would ordinarily take the form of management oversight and guidance: accountability and
context.
ACCOUNTABILITY
A heightened level of accountability comes along with the persistent situational awareness provided by
business intelligence in any production environment. While the ebb and flow of operational energy and
focus typically correspond to the rhythm at which periodic management reports are run, having live metrics
constantly at hand changes this dynamic permanently.
THE HAWTHORNE EFFECT
The Hawthorne Effect, also called the ‘observer’ effect, is often cited to explain persistent situational
awareness. The Hawthorne Effect takes its name from the factory complex where it was first observed.
Located in Cicero, Illinois, it was the largest factory complex in the U.S. back in the 1930s, when factories had
little to no automation, and factory floors were filled with people performing thousands of manual tasks. This
factory was the site of many industrial studies over the years, the most well-known of which sought to study
the effects of factory light levels on worker performance.
WHAT GETS MEASURED, GETS RESULTS.
COPYRIGHT © 2015 MOTIVITY SOLUTIONS, INC
When graveyard shifts were introduced, electric lighting became a new expenditure, previously not needed
during the day as skylights dominated factory ceilings to let in as much light as possible. Once they began
spending the money necessary to illuminate factories overnight, a question emerged as to just how much
light was needed. The thinking was that more light would lead to higher production levels as it might energize
the workforce, and conversely less light might induce lower performance levels.
As the study ensued, factory owners began experimenting with lower light levels. They started a measured
reduction of the light levels in a section of the factory and waited for the point they knew would come: a point
at which the workers would begin to slow down. They reasoned that if they could identify this threshold, they
could reduce energy costs without sacrificing performance. But as they reduced the level of light, the level
of performance increased with each reduction until they were seeing record levels of production in near total
darkness. This counterintuitive result quickly gave rise to the realization that the light had no effect on the
workers: they were merely responding to being observed.
This type of accountability can be quite powerful, but it’s only part of the equation when it comes to
establishing reflexive peak performance in a production environment. The other, much more powerful
component is context.
CONTEXT
Context, when applied correctly, can be even more powerful than the accountability that comes from
persistent situational awareness. Context can transform the way people think about their production work and
optimize performance, quality and output, all without direct managerial oversight. This type of automated
management and guidance is one of the most talked about aspects of mortgage business intelligence.
Anyone that has worked in loan operations understands the nature of task oriented workflow. Without
the automation offered by MBI, operations staffs view their work as somewhat monotonous in that they’re
presented with file after file, loan after loan, task after task. They understand each task and know that each
day they’ll be faced with more of these same tasks. The pace at which they perform them is dependent upon
things that are impossible to manage: how much sleep an employee may have had the previous night, what
they ate for breakfast, whether they’re distracted by personal matters that eclipse their ability to completely
focus on their work, and so on.
Once meaningful MBI metrics are consistently delivered to employees in operations, things begin to change.
Through the use of KPIs, employees can see for the first time the impact that their day to day, hour by
hour, and even minute by minute tasks are having on the established goal for their department, region, or
corporation. Through the simple use of red, yellow, or green KPI indicators to constantly display where a
given measurement lies relative to its established goal, employees instantly become goal oriented, forever
abandoning their task oriented mindset.
Consistently focusing on outcomes or beginning with the end in mind are well-known recipes for being
more effective and successful, and it’s this principal at work when context creates goal oriented workflow.
With an understanding and internalization of performance targets, operations employees quickly become
more dynamic thinkers. They carry more information in active memory as they reflexively prioritize their work.
Establishing a constant focus on the goal instead of the task is one of the fastest and most powerful ways to
WHAT GETS MEASURED, GETS RESULTS.
COPYRIGHT © 2015 MOTIVITY SOLUTIONS, INC
optimize any production operation, and it will bring about workflow optimization that is simply not possible
in addressing task oriented workflow ‘manually’, without MBI automation in place. Managers have been
attempting this for years, installing training, motivational programs and incentives all designed to motivate
and inspire people to maintain peak performance levels. But these approaches lack context because they
are not integrated with an employee’s day to day workflow. Embedded analytics have proven time and again
that they can automatically optimize production without direct management involvement.
CONCLUSION
Extensive automation is the rule in virtually every type of production endeavor. The time sensitive nature of
production necessitates the highest possible degree of automation. The average observer would be stunned to
find a typewriter being used in any business setting, but the vast array of manual analytics, communications, and
management processes present in mortgage lending do not provoke the same reaction. This is due to the fact
that they’ve not historically been recognized as manual, but the tide is turning.
More and more lenders are discovering the power of MBI, and the extent to which they’re wasting an alarming
number man hours by continuing to embrace manual processes. Automation through MBI holds the promise of
not just improving mortgage operations, but transforming them. And while some may feel that the automation
movement threatens their job or the jobs of their colleagues, it’s worthwhile to consider what has become known
as the Automation Paradox.
This paradox states that the more efficient an automated system is, the more critical the presence of human
operators. While they may be less involved, they are moved to their highest and best use, solving complex
problems that are beyond the capabilities of automation. Further, the mortgage industry is engaged with long
range business issues that will require just this type of complex problem solving. One looming issue that never
gets enough mindshare due to the time constraints produced by rampant manual processes is the overarching
problem of cost escalation. The cost of producing a mortgage loan has risen steadily since the post-war years,
and there is no end in sight.
Any other industry would not have survived as long without corrective action to reverse this cost trend, but the
mortgage industry has remained immune – until now. In the last few years, organizations such as Quicken Loans
and others have been attempting to reengineer how mortgages are produced, with the goal of eliminating the
human element altogether. Some envision removing functional divisions of labor, abandoning specialization
segments such as sales, processing, underwriting, and so forth, producing a digital/mechanical version of the
industry’s long standing business model.
Whether or not this would be in the best interest of the consumer is debatable, but lenders hoping to compete
in the coming years must have better control over their operations to endure. The good news is that with the
present state of the industry and the vast number of manual processes that have eluded automation, mortgage
lenders aiming to transform their businesses into formidable, competitive entities have a great deal to look
forward to. Once they recognize the extent to which they’re still using manual processes and begin to embrace
automation, they’ll have taken the first step toward having enough time and energy to address more complex,
long term, industry shaping issues.
WHAT GETS MEASURED, GETS RESULTS.
COPYRIGHT © 2015 MOTIVITY SOLUTIONS, INC