This project aimed for Business Analytics Specialization on Wharton School, University of Pennsylvania. Case discussed is GYF Ads who has threat from Ad Blocking that cannibalized organic user base. Problem Statement, Strategies, and Measurement are embodied in this presentation.

1. Business Analytics Capstone
Framework for Strategy
Fitrie Ratnasari
30th of July 2020

2. Problem Statement

3. Problem Statement–
Describe the Problem Adblockers present to GYF
1. How many end-users from our base are vanished due to ad-blocker and what segment of
users are they?
Decreasing end-user would make incomplete data analytics so that we could not capture perfectly the end-user
background, preferences, behaviour, from whole end-user porpulation as this is crucial to the prospected B2B/B2C to
reach their invaluable customer.
2. What are the main drivers of users who uses Ad-blockers? (as a root of causal problem)
Is it because irrelevant ads from their preferences, or too much ads they’re obtained in 1 page, or the design were
confusing or the image/video size proportion are too large or the page deliverable is so slow?
3. What kind of behaviour our customer before they’re starting to use Ad-blockers? How
was our ads appearance before they’re waving goodbye to our service?
Identify their Recency, Frequency and Monetary (RFM) to know their CLV and behaviour prior departure.
4. What kind of data are fleet from our end user and its implication to our customer
analytics and marketing strategy?
Are they also activate to block our GYF-Analytics? If yes, then we’re unable to get the cookies, time spent, links clicked
etc to conduct further analytics which is very crucial to our service and revenue.
5. How large the monetary aftermath would cannibalized due to this threat?
Declining end-user would also degraded our CTR possibility that would have bad impact to conversion and revenue
subsequently.

4. Problem Statement–
Application Exercise 1 – Research Methods and Tools (Optional)
• Research we’ll be using is Descriptive Research, why?
Considering we are aware of the issues of ad-blocker existence, what we need is to reveal the
implication of the issue by unleash the numbers of how big it would impact to our customer base and
to our financial bottom line and we need to identify the root of causal problem or reason of end-users
uses ad-blockers.
As the next step we’ll be using Causal Research, especially A/B testing to conduct new native
advertisement appearance to make end-user more comfortable, which ones would be preference to the end-
users to take up the conversion rate. By then we could subsequently decide what strategy we need to
conduct to anticipate and mitigate if the result go rough.
• Research sources are from :
1. General model of scanner data in the back end, which could track whether end user is using ads-blocker
or not, if using it, we also need to record their time stamp (by using its JavaScript) to investigate their
behaviour prior using ads-blocker. By this data we can do modelling through Machine Learning, of which
from our user base are alleged to churn by using Ad-blocker.
2. Mobile survey (to end user population), to investigate the main drivers of end user who uses ad-blocker
as the causal problem.
3. Mobile Data Analytics to know whether segment of our end user are using adblockers, and how far it
would impact our end-user data analytics.
4. Social Media to capture our ads engagement & ads persona.

5. Strategy

6. Strategy
Strategy of ad-blockers threat are basically lies on in creating sustainable advertisement ecosystem, with 2
main strategies:
1. Generating much greater advertising experience for user to see the ads without pushing them, as key
line following:
• Tighten Quality of Assurance (QA) of advertising content, that ads content’s deliverable should be delightful,
informative, meaningful, invisible inspirational, and indispensable part of product/service experience, with taking
improvement of page load time into consideration. Company will build 1 dedicated team to handle the creative ads
content for B2B/B2C customer of GYF ads can consult the content value.
• Executing A/B Testing on different ads experience to each segment of user-base, to know end-user preference,
and deploying the result which new native ads appearance suits best for each segment.
• Create GYF built-in adblocking that could make end-user to choose their own preference to prohibit the access of
ads or not. If they’re choosing to activate the GYF built-in adblocker, GYF-analytics still can fetch the end-user
cookies, link clicked, and other data to do further analytics.
• For end-user who already used other 3rd party ad-blockers, we’re send the prompt that asking whether they will turn
off the old one and activate GYF built-in adblocker considering their data security is crucial to unknown 3rd
party.
2. The ambidextrous tactics to boost the strategy itself. The key line strategies are:
• Create the solid culture of ads team member, by building execution culture that lays out key decision, action,
support the model and execution plan with effective incentives and controls, which take agreement and commitment
are sought and rewarded into consideration. This culture would lead successful strategy execution and to avoid
member turnover challenge.
• Create collaborative ecosystem all across search engine companies and publishers by holding consortium to ‘drive’
the industry that potentially become an industry barrier, as for the case to standardize the ads content to not be so
invasive and too bloated.

7. Strategy
Application Exercise 2 – Hiring a Team Leader (Optional)
3. Hiring a Team Leader
• Carrie Candidate is chosen to be Senior Associate Director for Digital Advertising Strategy,
considering the result of analytics research (from Multivariate Regression Method) that shown
work samples, cognitive ability tests, structured interview, are the most correlated to
performance.
• Looking at the basis of hiring decision of
metrics assessment result Carrie Candidate
portrays higher result than Peggy Prospect
in terms of work samples, cognitive ability
tests, and structured interview.
Ryan & Tippins 2004,
“Attracting and selecting: What psychological research tells us”

8. Effects and Measurement

9. Effects
Describe the anticipated effects of your strategy
• Please describe the anticipated effects of your strategy. Make sure you address the effects on customers, revenue, and the internal organization.
Here are the implication of our strategy of creating sustainable advertisement ecosystem:
1) The first proposed strategy is to generating much greater advertising experience (by tighten Quality of
Assurance of advertising content’s deliverable, deploying the result which new native ads appearance as
result of A/ B testing) would addressed end-user loyalty by staying in our page without taking any action to
install add-blocking software. End-user would pleasantly stay on page because they are perceiving valuable
and informative insight from the ads content. Subsequently the brand awareness and customer
engagement can be reached.
For such a circumtances we’re maintaining our user base, that become key players to the ads impression,
click, conversion (if any of them are piqued by the product/service), and subsequently would bring
beneficiary impact to profit our customer and become our revenue. Internally, as mentioned 1
dedicated team of creative content advertisement is needed to ensure the quality, this might need the
accrual cost and hind the financial statement, but it would make significant impact to long term corporate
value.
2) With the second strategy, ambidextrous tactics by create the solid innovative & collaborative culture in
advertising team would also make successful strategy execution and great way to avoid member turnover
challenge. As for trade off the considerable incentive for those who performs well are encountered,
but it will make a condusive ecosystem to achieve considerable growth, which improve the intangible
asset & competitiveness factor of company and become indirect driver of long term growth.

10. Measurement
Please outline your plan for measuring these effects using data. Make sure you use techniques you learned about in the courses
Measurement of Ads Performance:
ü Ads experience can be measured by number of impressions, and video play length 25,50,75,100%
(quantiles).
ü Ads performance can be measured by :
1. Return on Investment (ROI) metrics (Revenue - Cost of goods sold) / Cost of goods sold
2. Brand awareness which can be seen from number of impression
3. Customer engagement can encounter by Click-Through Rate (CTR) and conversion rate
These measurements are common used in ads industry (such as Google Ads, Yahoo, etc)
Measurement for Internal Performance:
ü Innovative Culture : % growth of innovation produced
ü Collaborative Culture : % drop of employee turnover & evaluation result of Collaboration Network
ü Growth in Intangible asset
Measurement for Company Competitive Advantage across industry:
ü Return on Equity
ü Profitability which measures Return on Assets and Return on Sales
ü Efficiency, encountered Assets Turnover as key ratio.
ü Balanced score card

11. Effects
Application Exercise 3 – Designing a Deterministic Optimization Model
• Provide an explanation of the calculations you performed to build an optimization model using decision variables, constraints, and an objective; this model could use
the scenario in Application Exercise 3, or one of you own devising.
Optimization Model on Training Program
Variable definition used:
Algebraic Statement
Objective : Maximize 0.2*H/I + 0.7*H/E + 0.6*S/I + 0.4*S/E
Decision Variable : H/I, H/E, S/I, S/E
Constraints:
Spending on Hard Skills/Internal = H/I
Spending on Hard Skills/External = H/E
Spending on Soft Skills/Internal = S/I
Spending on Soft Skills/External = S/E
1. Budget spending
H/I + H/E + S/I + S/E <= US$ 65,000
2. Hard Skills requirement
0.2*H/I + 0.7*H/E >= US$ 20,000
3. Soft Skills requirement
0.6*S/I + 0.4*S/E >= US$ 12,000
4. Internal & External Proportion
(0.2*H/I + 0.6*S/I) / (0.4*S/E + 0.7*H/E) >= 60%
From the analytics of software tools (solver) used that shown
in appendix page, in order to maximize the productivity return
(obtained US$ 42,824 after doing calculation), the spending
amount of training on Hard Skill/Internal and Soft
Skills/External are 0, means does not need to conduct any
training on these two area. In contrast, company should
consider the budget expense US$ 26,765 for training on Soft
Skills/Internal and US$ 38,235 for Hard Skills/External
training.
However from my perspective, considering 1) Any modelling
has limitation that could not scope the correlation between
predictor and objective entirely, and 2) Training program are
related to human resources of which we know that people are
the product of dynamic change and growth, thus training Hard
Skills/Internal and Soft Skils/External might also imperative
for employees to do the job greatly. This is the case when
blending the experts and analytics are considerably
needed to achieve much greater outcome towards
company’s objective.

12. Effects - Appendix
Application Exercise 3 – Designing a Deterministic Optimization Model

13. Measurement
Application Exercise 4 – Identifying Key Drivers
Causal Business Model Performance Measurement Framework
Goal : Achieve net income growth from advertisement service against ad-blocking software threat.
Greater experience of
ads content
End-user loyalty (stays without
taking action to install add-
blocking software)
Impression
Click
Conversion
Profit for
company
More Profit Per Impression
(PPI) would lead more GYF’s
revenue in the long run
Measurement :
How great the experience from ads content (as a key driver) can be seen from
Brand Awareness (measured by number of impression) and Customer
Engagement (measured by Click-Through-Rate), and subsequently to ROI.
Causal Business Model (Hypothesis): when user perceiving greater experience
of ads content (as a key driver), would increase end user loyalty (they won’t install
addblocking software), that would impact to number of impressions, CTR,
Conversion which drives revenue of GYF’s Clients. The more Profit per
Impression achieve, leads more GYF’s revenue in the long run.
Hypothesis Validation:
Validation method this hypothesis is to conducting a research and modelling
through Multivariate Regression Analysis, to know whether experience of ads
content taking the most valuable predictor for revenue increase (against ad-
block threat) amongst other key driver, i.e : solid innovation & collaborative
culture of ads team member. If it does, then the p-value of predictor experience
of ads content would be the highest.
Supporting data needed:
- Mobile survey from end-user population regarding ads performance measurement
- Scanner data in the back end to know growth of impression, click, conversion and
turnover of ad-blocking users
- Numbers of innovation produced and performs
- Evaluation result of collaboration networks
- Financial measurement metrics
Causal Business Model (Hypothesis)

14. Measurement
Application Exercise 4 – Identifying Key Drivers
Linking non financial metrics to financial objective
However there is also probability when the key driver mentioned would not perform. If this happened, we
should find out why. At least there’s 3 (three) possibility reason : 1) the strategy aligned is wrong, 2) the
measurement are garbage (we can eliminate this possibility since we use common measurement in industry for
decade), and 3) the people who are doing the job are not dependable.
This is when non financial metrics take considerable role to financial objective. We should evaluate and dothe
right measurement on non financial metrics (in this case, the innovative and collaborative culture, management
capability, employee relation, quality and brand value) to know the causal reason of the issues. If the case is
solved, which we see the highest predictable factor of key drivers makes increasing net income successfully, this
means both non financial metrics and financial outcome are highly correlated as it supposed to be, that is why
most of corporate nowadays use Balanced Score Card to measure how well one company is doing, especially
their competitive advantage across industry.

15. Conclusion (1)
ü The existence of ad-blocking software becoming threat for the company as advertisement takes 70% of
revenue stream of the company. The first thing we need to do further descriptive research aimed to find the
numbers of end-user are vanished due to this threat, the causal reason of end-users are determined to avoid
ads and using ad-blocking software, in what any extent impact to our GYF analytics and financial bottom
line. With the supporting data needed such as general model of scanner data in the back end, mobile survey,
mobile analytics, social media would greatly help to emphasize the real phenomenon.
ü Regarding this threat, management proposing strategies aligned that heavily lies on creating sustainable
advertisement ecosystem by generating much greater advertising experience (by tighten Quality of
Assurance of advertising content’s deliverable, deploying the result which new native ads appearance as
result of A/ B testing) and ambidextrous tactics by creating the solid innovative and collaborative culture of
ads team member internally & paralelly building collaborative ecosystem all across search engine companies
and publishers by holding consortium to ‘drive’ the industry that potentially become an industry barrier. For
trade off, additional accrual cost for dedicated team of ads creative content and incentive for those who
performs well are encountered, but it will make a condusive ecosystem to achieve considerable growth.
ü To measure whether our strategies especially key factors we defined are having high correlation to increasing
income growth despite add-blocking threat, we need to conduct evaluation for each strategies aligned by
using measurement we have set, such as number of impressions, and video play length 25,50,75,100%
(quantiles), Return on Investment, Click-Through Rate (CTR) and conversion rate. As for internal evaluation
we need to conduct Innovative Culture, Collaborative Culture, and Growth in Intangible asset. It is
imperative also to recognise our company competitive advantage using measurement of financial ratios
(R.O.I, R.O.A, R.O.S., Asset Turnover) and balanced score card.

16. Conclusion (2)
ü As final step we need to linking the non financial strategies to financial outcome by conducting research and
modelling through Multivariate Regression Analysis, to know whether experience of ads content (as key
driver) taking the most valuable predictor for revenue increase (against ad-block threat) amongst other key
driver, i.e : solid innovation & collaborative culture of ads team member. When the outcome went wrong,
which is not implies the correlation to increasing net income, we need to investigate by evaluating further the
reason on strategies aligned, measurement, and people who are involved in the project.
ü Once we have solved the real reason behind (might change the causal business model), we’ll see the highest
predictable factor of key drivers makes increasing net income successfully, this means both non financial
metrics and financial outcome are highly correlated as it supposed to be. This means non-financial takes
important role to financial outcome,.