1
Marketing Mix Modeling in
Financial Services
POV Prepared by Ninah
2
Whilst the use of Marketing Analytics to support
Marketing Mix Investments is only around 19.8%
for all industries (Source CMOSurvey.ORG, 2015)
we know that amongst the main national retail
banks it is 100%.
Perhaps this isn’t so surprising. Given the level
of accountability demanded of banks following
at first Sarbanes-Oxley and more recently Dodd
Frank for accountancy standards and capital
allocation support, looking to analytics to justify
investments is a natural. And in any case, surely
the ROI argument could not find a better vertical to
resonate in; not only is it the language of finance,
it is also the lifeblood for risk scoring and credit
appraisal.
So with many banks (and insurance, brokerage
and tax preparation firms) leveraging marketing
mix the more interesting question is not the degree
of penetration for Marketing Mix Modeling (MMM)
but rather what is it that distinguishes a finance
marketer in the what, how and why of its use.
Working this backwards we’ve seen the single
greatest force propelling all banks to use MMM
is a greater focus on managing customer value.
This in turn has led to linking marketing with the
drivers of customer value and that in turn has sped
greater centralization of the marketing function –
centralizing both geographically and across Lines
of Business (LOBs). With this impetus MMM has
been leveraged to enable a more effective center
by providing it both with the optics into the business
as well as some levers to pull in managing it.
So how can we understand the different levels of
progression across finance companies? Broadly
speaking we see a continuum of user-ship and
enablement levels:
•	 On the production side from a Do it For Me
to a Do it Yourself or fully outsourced, to fully
internalized,
•	 On the decision making side from a simple
annual channel budget with static Economic
Profit or Shareholder Value Added to a live
investment optimization across message and
channel to drive dynamic Economic Profit and
Shareholder Value Added.
Where any given client sits on these continuums
typically depends on the level of adoption and use
that has been established:
•	 Some may be using their analytics vendors to
do both relatively simple modeling as well as
very advanced, yet others have full internalized
both channel focused MMM and the advanced
modeling. Most typically of course there is a
balance, with the more basic work being done
internally with the analytics partner used for
scale and for advanced developments.
•	 More complex analytics is typically performed
where earlier “quick wins” have been
established and credited to the modeling.
Once finance departments look to the models
to manage brand and promotional investments
to drive greater Customer Lifetime value then
the marketing department is seen as a key
force in centralizing all banking activities
So what drives the level of adoption? In our opinion
there are two fundamental drivers:
•	 Aligning analysis with the actual decision
makers needs – build to fit.
•	 Deliver results with a narrative that is based on
the companies own style and language vs. the
analytics firms
In addition today another more passive but also
more fundamental force is playing into adoption
and that is the requirements emanating from
model governance and the compliance needs of
the Federal Reserve. This is the technical flipside
to the organizational appetite above. Both these
adoption drivers require a marketing analytics
function that can be both highly technical and
dexterous whilst also being highly tuned and
consultative in their approach.
MMM: POV Prepared by Ninah
3
Once a virtuous adoption cycle is established
companies are then able to evolve the use of
marketing analytics and MMM to more advanced
use cases. More specifically in financial services
we’ve seen a wide range of applications unique to
the industry:
•	 Multiple Outcome measures: The value
equation in financial services tends to
be more complex with multiple business
outcome metrics impacting value and ROI
e.g. acquisition, usage, balances and churn.
This dynamic creates a more complex set
of dependent variables to fully understand
marketing ROI.
•	 Segmented consumer dynamics: Financial
services companies have information not
just on buying behavior but who is buying
and the tiers of products creating a need for
a segmented understanding of marketing
impact on business outcome across customer
segments and product tiers/types.
•	 Understanding customer value: Product value
used in marketing effectiveness evaluation
is dependent on customer behavior and
usage, therefore good customer lifetime
valuation models at product and/or customer
level e.g. Net Present Value SVA models are
key to marketing effectiveness in financial
services. Engagements will often include
some component of building, modifying or
deconstructing these models to understand
marketing ROI.
•	 Integrating and balancing direct and mass
marketing: Financial services marketers
rely on a complex interplay between direct
channels e.g. mail, email, search, social
and mass channels e.g. display, TV, print.
Understanding the contribution of all channels
and the interaction effects between channels,
particularly direct and mass channels is key in
financial services.
•	 Complex distribution environment: Financial
services are often distributed and serviced
through multiple outlets e.g. web and
branch with different marketing impact by
distribution channel, and regional differences
in distribution strength. This requires models to
be built at a geographically segmented level
across multiple channels.
•	 Balancing branding, acquisition and
deepening: Financial services marketers often
have to balance a complex set of messaging
objectives ranging from brand building to
deepening and acquisition. This creates
a more complex taxonomy for organizing
media and marketing data for marketing mix
modeling.
•	 Broader set of macro drivers: Financial services
and brokerage in particular is impacted by
a broader set of macro metrics that a typical
MMM model e.g. market index, Volatility Index
(VIX), yield curve, options expiration dates,
housing market, data release, fed actions,
unemployment etc. Controlling correctly
for these macro drivers is key to getting an
accurate marketing estimate.
MMM: POV Prepared by Ninah
4
Marketing Mix Modeling in
Financial Services
Case Studies
5
Case Study: Building a Strong Case for “Brand Effects” in Insurance Advertising
Building a Strong Case for
“Brand Effects” in Insurance Advertising
Situation
The highly competitive U.S. auto insurance
marketplace poses a real challenge for marketers,
where the path to growth involves building and
maintaining brand equity, and demands for a more
efficient marketing spend than the competition.
With annual budgets for the top insurance firms
often exceeding $500MM, building the case for
marketing requires multi-year measurement and
understanding marketing’s role in building a brand.
Challenge
The client, a mid-size auto insurance firm, which
specializes in direct-to-consumer personal car
insurance, acknowledged that advertising does not
pay off in the short-term given the high acquisition
costsinadditiontocompetitivepricing.Theywanted
to measure the long-term effects of advertising on
brand equity and base sales that had been built
over time. The idea was to optimize their media mix,
and to ensure that any increased investment would
be directed towards the most efficient messaging.
It wasn’t enough to understand how each message
performed individually; instead it was necessary to
understand how they worked together in the short
and long-term to drive baseline quotes completed.
The client had a firm grasp of its short-term impact
of advertising on quotes completed, which was
measured by the marketing mix model developed
internally by their analytics team. However, they did
not have a clear understanding on how to quantify
the influence of brand equity and deduce the
long-term effect of advertising on baseline quotes
completed. Therefore, it was crucial that the results
provided insights on both the short and long-term,
as well as the overall impact on quotes completed
when both were taken into consideration.
Goal
The client looked to Ninah Consulting to help
optimize their media mix balancing short-term
sales with long-term brand equity. They set out to
understand how brand equity affected business
outcome, how the long-term effect of advertising
differs from the shorter term and what the optimal
allocation is across messages to maximize returns.
Essentially, they were building the business case
for continued investment behind their brand.
Working with client stakeholders, we identified the
following key business questions:
•	 Which brand equity metrics are the most
important to us?
•	 What is the relationship between brand equity
and baseline quotes completed?
•	 What is the long-term effect of advertising on
brand equity, which ultimately leads to baseline
quotes completed?
•	 What is the short and long-term combined
effect of advertising on quotes completed?
•	 What is the optimal message mix (Brand vs.
De-positioning vs. Savings)?
Solution
Ninah Consulting designed an analytics framework
to identify which of the 15 Brand Equity measures
had the strongest relationship with baseline
quotes. Once the right brand equity measure
was determined, an econometric model was
built to measure the impact of brand equity on
baseline quotes completed while controlling for
other baseline factors such as seasonality and
macroeconomic factors. A dynamic linear model
was also developed previously to compare
its “moving base” against the client provided
baseline, which was derived from their short-term
marketing mix model, to build confidence in using
their baseline as our target metric; the two were
very similar and highly correlated with one another.
Afterwards, another econometric model was built
to measure the long-term effect of advertising on
brand equity, accounting for any lagged effects,
6
carry-over effects, and diminishing returns. Having
both models in place, the relationship between
advertising and baseline quotes completed by
way of brand equity could be quantified. As a
result, Ninah was able to optimize the message
mix to balance short term sales growth with long
term brand equity building.
Results
Ninah Consulting created a simulation tool based
on the marketing mix models. The tool enabled the
client to scenario plan different media allocations
by message type; adjusting for the differences in
response during the short and long-term, while
capturing the overall effect on quotes completed.
In addition, Ninah Consulting also provided a
number of valuable insights for the client:
•	 Unaided Consideration is the most important
brand equity metric that best captures changes
in baseline sales.
•	 Brand equity had a contribution of over 50% to
baseline quotes completed.
•	 Television & Video drove 70% of brand equity
in the long term, which translated to 37% of
baseline quotes completed and 19% of total
quotes completed.
•	 Advertising has a long half-life or carry-over
effect of 13 weeks and decays relatively slowly
against brand equity.
•	 There is more room to grow brand equity,
and in turn quotes completed, with Brand
messaging than De-positioning or Savings;
Brand messaging had a relatively linear
relationship in driving brand equity, whereas
De-positioning and Savings have moderate to
strong degrees of diminishing returns.
Case Study: Building a Strong Case for “Brand Effects” in Insurance Advertising
7
Case Study: Media Mix Optimization Delivers 17% Increase in ROI
for Insurance Provider
Media Mix Optimization Delivers 17% Increase in
ROI for Insurance Provider
Situation
The Irish insurance market is highly competitive,
battling increasing degrees of consumer switching.
Throughout the industry, media activity has risen
in recent years, with year-over-year double-
digit growth in advertising. That spend includes
traditional media, though insurance providers were
also directing more euros to digital.
Challenge
Facing increased pressure in a rapidly changing
media and consumer engagement landscape, the
client, insurance provider, wanted to ensure they
were effectively allocating their media budget.
They enlisted Ninah Consulting to help them
better understand the overall media influence on
customer acquisition and revenue.
Goal
The primary objective was to understand the true
contribution of media and the subsequent optimal
media mix to drive acquisition and revenue. To that
end, Ninah interviewed key stakeholders across
the client organization and identified the following
key questions:
•	 How does performance vary by media channel
– TV, online, radio etc.?
•	 What is the optimum level of TV weights
through the year?
•	 What is the effectiveness of different creative?
Campaigns?
•	 To what extent does competitor media activity
impact performance?
•	 What is the influence of other key factors such
as underlying seasonality and the economy?
Solution
Ninah established econometric models across key
channels (traditional and digital, including mobile
– we didn’t do mobile! There is no data on mobile
in Ireland, so this is not a claim we can make, I’m
afraid!) in order to understand key performance
drivers, the influence of each media channel, and
the cost efficiencies of each channel in driving
acquisition. Utilizing sales conversion rates allied
to revenues, Ninah was also able to derive ROI by
media channel, which was then used as a basis for
budget optimization.
Results
The models revealed that the digital channels
were the most efficient and increasingly important
for driving customer acquisition and revenue. TV
was still an important part of the mix, but other
traditional/offline channels proved least efficient.
As a result of the data and analysis provided by
Ninah, the client optimized TV seasonally and re-
allocated budget from less efficient offline channels
to more efficient online channels such as Search.
Ninah Consulting’s actionable results armed
the client with a more effective media strategy
with which to compete in the rapidly changing
customer engagement landscape and, notably, a
17% improvement in ROI.
8
For more information contact:
Sebastian Shapiro
Managing Partner, Ninah
sebastian.shapiro@ninah.com
1-212-820-3302
375 Hudson Street
New York New York 10014
United States

Marketing Optimization in Financial Services

  • 1.
    1 Marketing Mix Modelingin Financial Services POV Prepared by Ninah
  • 2.
    2 Whilst the useof Marketing Analytics to support Marketing Mix Investments is only around 19.8% for all industries (Source CMOSurvey.ORG, 2015) we know that amongst the main national retail banks it is 100%. Perhaps this isn’t so surprising. Given the level of accountability demanded of banks following at first Sarbanes-Oxley and more recently Dodd Frank for accountancy standards and capital allocation support, looking to analytics to justify investments is a natural. And in any case, surely the ROI argument could not find a better vertical to resonate in; not only is it the language of finance, it is also the lifeblood for risk scoring and credit appraisal. So with many banks (and insurance, brokerage and tax preparation firms) leveraging marketing mix the more interesting question is not the degree of penetration for Marketing Mix Modeling (MMM) but rather what is it that distinguishes a finance marketer in the what, how and why of its use. Working this backwards we’ve seen the single greatest force propelling all banks to use MMM is a greater focus on managing customer value. This in turn has led to linking marketing with the drivers of customer value and that in turn has sped greater centralization of the marketing function – centralizing both geographically and across Lines of Business (LOBs). With this impetus MMM has been leveraged to enable a more effective center by providing it both with the optics into the business as well as some levers to pull in managing it. So how can we understand the different levels of progression across finance companies? Broadly speaking we see a continuum of user-ship and enablement levels: • On the production side from a Do it For Me to a Do it Yourself or fully outsourced, to fully internalized, • On the decision making side from a simple annual channel budget with static Economic Profit or Shareholder Value Added to a live investment optimization across message and channel to drive dynamic Economic Profit and Shareholder Value Added. Where any given client sits on these continuums typically depends on the level of adoption and use that has been established: • Some may be using their analytics vendors to do both relatively simple modeling as well as very advanced, yet others have full internalized both channel focused MMM and the advanced modeling. Most typically of course there is a balance, with the more basic work being done internally with the analytics partner used for scale and for advanced developments. • More complex analytics is typically performed where earlier “quick wins” have been established and credited to the modeling. Once finance departments look to the models to manage brand and promotional investments to drive greater Customer Lifetime value then the marketing department is seen as a key force in centralizing all banking activities So what drives the level of adoption? In our opinion there are two fundamental drivers: • Aligning analysis with the actual decision makers needs – build to fit. • Deliver results with a narrative that is based on the companies own style and language vs. the analytics firms In addition today another more passive but also more fundamental force is playing into adoption and that is the requirements emanating from model governance and the compliance needs of the Federal Reserve. This is the technical flipside to the organizational appetite above. Both these adoption drivers require a marketing analytics function that can be both highly technical and dexterous whilst also being highly tuned and consultative in their approach. MMM: POV Prepared by Ninah
  • 3.
    3 Once a virtuousadoption cycle is established companies are then able to evolve the use of marketing analytics and MMM to more advanced use cases. More specifically in financial services we’ve seen a wide range of applications unique to the industry: • Multiple Outcome measures: The value equation in financial services tends to be more complex with multiple business outcome metrics impacting value and ROI e.g. acquisition, usage, balances and churn. This dynamic creates a more complex set of dependent variables to fully understand marketing ROI. • Segmented consumer dynamics: Financial services companies have information not just on buying behavior but who is buying and the tiers of products creating a need for a segmented understanding of marketing impact on business outcome across customer segments and product tiers/types. • Understanding customer value: Product value used in marketing effectiveness evaluation is dependent on customer behavior and usage, therefore good customer lifetime valuation models at product and/or customer level e.g. Net Present Value SVA models are key to marketing effectiveness in financial services. Engagements will often include some component of building, modifying or deconstructing these models to understand marketing ROI. • Integrating and balancing direct and mass marketing: Financial services marketers rely on a complex interplay between direct channels e.g. mail, email, search, social and mass channels e.g. display, TV, print. Understanding the contribution of all channels and the interaction effects between channels, particularly direct and mass channels is key in financial services. • Complex distribution environment: Financial services are often distributed and serviced through multiple outlets e.g. web and branch with different marketing impact by distribution channel, and regional differences in distribution strength. This requires models to be built at a geographically segmented level across multiple channels. • Balancing branding, acquisition and deepening: Financial services marketers often have to balance a complex set of messaging objectives ranging from brand building to deepening and acquisition. This creates a more complex taxonomy for organizing media and marketing data for marketing mix modeling. • Broader set of macro drivers: Financial services and brokerage in particular is impacted by a broader set of macro metrics that a typical MMM model e.g. market index, Volatility Index (VIX), yield curve, options expiration dates, housing market, data release, fed actions, unemployment etc. Controlling correctly for these macro drivers is key to getting an accurate marketing estimate. MMM: POV Prepared by Ninah
  • 4.
    4 Marketing Mix Modelingin Financial Services Case Studies
  • 5.
    5 Case Study: Buildinga Strong Case for “Brand Effects” in Insurance Advertising Building a Strong Case for “Brand Effects” in Insurance Advertising Situation The highly competitive U.S. auto insurance marketplace poses a real challenge for marketers, where the path to growth involves building and maintaining brand equity, and demands for a more efficient marketing spend than the competition. With annual budgets for the top insurance firms often exceeding $500MM, building the case for marketing requires multi-year measurement and understanding marketing’s role in building a brand. Challenge The client, a mid-size auto insurance firm, which specializes in direct-to-consumer personal car insurance, acknowledged that advertising does not pay off in the short-term given the high acquisition costsinadditiontocompetitivepricing.Theywanted to measure the long-term effects of advertising on brand equity and base sales that had been built over time. The idea was to optimize their media mix, and to ensure that any increased investment would be directed towards the most efficient messaging. It wasn’t enough to understand how each message performed individually; instead it was necessary to understand how they worked together in the short and long-term to drive baseline quotes completed. The client had a firm grasp of its short-term impact of advertising on quotes completed, which was measured by the marketing mix model developed internally by their analytics team. However, they did not have a clear understanding on how to quantify the influence of brand equity and deduce the long-term effect of advertising on baseline quotes completed. Therefore, it was crucial that the results provided insights on both the short and long-term, as well as the overall impact on quotes completed when both were taken into consideration. Goal The client looked to Ninah Consulting to help optimize their media mix balancing short-term sales with long-term brand equity. They set out to understand how brand equity affected business outcome, how the long-term effect of advertising differs from the shorter term and what the optimal allocation is across messages to maximize returns. Essentially, they were building the business case for continued investment behind their brand. Working with client stakeholders, we identified the following key business questions: • Which brand equity metrics are the most important to us? • What is the relationship between brand equity and baseline quotes completed? • What is the long-term effect of advertising on brand equity, which ultimately leads to baseline quotes completed? • What is the short and long-term combined effect of advertising on quotes completed? • What is the optimal message mix (Brand vs. De-positioning vs. Savings)? Solution Ninah Consulting designed an analytics framework to identify which of the 15 Brand Equity measures had the strongest relationship with baseline quotes. Once the right brand equity measure was determined, an econometric model was built to measure the impact of brand equity on baseline quotes completed while controlling for other baseline factors such as seasonality and macroeconomic factors. A dynamic linear model was also developed previously to compare its “moving base” against the client provided baseline, which was derived from their short-term marketing mix model, to build confidence in using their baseline as our target metric; the two were very similar and highly correlated with one another. Afterwards, another econometric model was built to measure the long-term effect of advertising on brand equity, accounting for any lagged effects,
  • 6.
    6 carry-over effects, anddiminishing returns. Having both models in place, the relationship between advertising and baseline quotes completed by way of brand equity could be quantified. As a result, Ninah was able to optimize the message mix to balance short term sales growth with long term brand equity building. Results Ninah Consulting created a simulation tool based on the marketing mix models. The tool enabled the client to scenario plan different media allocations by message type; adjusting for the differences in response during the short and long-term, while capturing the overall effect on quotes completed. In addition, Ninah Consulting also provided a number of valuable insights for the client: • Unaided Consideration is the most important brand equity metric that best captures changes in baseline sales. • Brand equity had a contribution of over 50% to baseline quotes completed. • Television & Video drove 70% of brand equity in the long term, which translated to 37% of baseline quotes completed and 19% of total quotes completed. • Advertising has a long half-life or carry-over effect of 13 weeks and decays relatively slowly against brand equity. • There is more room to grow brand equity, and in turn quotes completed, with Brand messaging than De-positioning or Savings; Brand messaging had a relatively linear relationship in driving brand equity, whereas De-positioning and Savings have moderate to strong degrees of diminishing returns. Case Study: Building a Strong Case for “Brand Effects” in Insurance Advertising
  • 7.
    7 Case Study: MediaMix Optimization Delivers 17% Increase in ROI for Insurance Provider Media Mix Optimization Delivers 17% Increase in ROI for Insurance Provider Situation The Irish insurance market is highly competitive, battling increasing degrees of consumer switching. Throughout the industry, media activity has risen in recent years, with year-over-year double- digit growth in advertising. That spend includes traditional media, though insurance providers were also directing more euros to digital. Challenge Facing increased pressure in a rapidly changing media and consumer engagement landscape, the client, insurance provider, wanted to ensure they were effectively allocating their media budget. They enlisted Ninah Consulting to help them better understand the overall media influence on customer acquisition and revenue. Goal The primary objective was to understand the true contribution of media and the subsequent optimal media mix to drive acquisition and revenue. To that end, Ninah interviewed key stakeholders across the client organization and identified the following key questions: • How does performance vary by media channel – TV, online, radio etc.? • What is the optimum level of TV weights through the year? • What is the effectiveness of different creative? Campaigns? • To what extent does competitor media activity impact performance? • What is the influence of other key factors such as underlying seasonality and the economy? Solution Ninah established econometric models across key channels (traditional and digital, including mobile – we didn’t do mobile! There is no data on mobile in Ireland, so this is not a claim we can make, I’m afraid!) in order to understand key performance drivers, the influence of each media channel, and the cost efficiencies of each channel in driving acquisition. Utilizing sales conversion rates allied to revenues, Ninah was also able to derive ROI by media channel, which was then used as a basis for budget optimization. Results The models revealed that the digital channels were the most efficient and increasingly important for driving customer acquisition and revenue. TV was still an important part of the mix, but other traditional/offline channels proved least efficient. As a result of the data and analysis provided by Ninah, the client optimized TV seasonally and re- allocated budget from less efficient offline channels to more efficient online channels such as Search. Ninah Consulting’s actionable results armed the client with a more effective media strategy with which to compete in the rapidly changing customer engagement landscape and, notably, a 17% improvement in ROI.
  • 8.
    8 For more informationcontact: Sebastian Shapiro Managing Partner, Ninah sebastian.shapiro@ninah.com 1-212-820-3302 375 Hudson Street New York New York 10014 United States