CPM = Cost Per Mille

Count the number of thousands of people watching a program or an ad

Nielsen Age breakdowns: 13-17, 18-24, 25-34, 35-54, 55+ and 18-49

HUT = Households Using TV / Total TV Households

Rating = Channel X Households / Total TV Households

Share = Channel X Households / Households Using TV

GRPs = Monday = X Rtg, Tuesday = X Rtg, Wednesday = X Rtg, Thursday = X Rtg, Friday = X Rtg, Saturday
= X Rtg, Sunday = X Rtg. GRPs = Total (SUM) GRPs

Reach = Channel X Households / Households Using TV

Frequency = GRPs / Reach

Media Buying Agencies = create, plan, handle campaign, exclusive relationship with client

Network = place, ad serving, measure, work with competing clients and publishers

Media buying and analyzing tools: [1] insight into digital marketing analytics tools, [2] use tools to
interpret campaign performance, [3] apply tools with the data sets provided and generated, [4] use
tools to analyze and measure marketing performance

Digital marketers: [1] build campaigns, [2] buy media or ad placement, [3] optimize media campaigns

5 methods of digital marketing: [1] display ads, [2] search engine marketing (SEM), [3] social media
marketing, [4] mobile advertising, [5] direct messaging and email marketing

Ad bidding = vickrey auction model

Quality score factors = [1] keywords, [2] ad copy, [3] landing page, [4] CTR of ad

Ad determination
Probability of ad campaign: CPC, conversion rate, sale value, profit margin, repeat behavior / loyalty /
lifetime value of customer (LTV)

CPC from search engine ad system

Conversion rates from search engine analytics or internal analytics

Repeat sale is cheaper to produce than a new sale

Low CPC = Target long tail keywords and combination of high CTR and high conversion rate

Break-even point analysis:

Metric Value
Avg. CPC (cost per click) $1.00
Conversion Rate 5.00%
Avg. Sale Value $60
Profit Margin 15%
Lifetime Value (LTV) $180

PPC Conversion Cost = $1.00 / .05 = $20

Profit Margin per Sale = $60 x .15 = $9

BEP CPC per Avg. Sale = $9* .05 = $0.45

(what we should be spending per click for one sale)

BEP CPC for LTV = $27 * .05 = $1.35

Advertising sales

Fixed cost / flat rate (per time interval)

CPM = Cost per mille (1,000 impressions)

CPV = Cost per view (actual view of an ad)

CPC = Cost per click

CPA/CPL = Cost per acquisition or per lead

CPS = Cost per sale (“commission”)

Example: TV ads sold on CPM basis. Search ads on CPC basis. Hybrid ads on CPV+CPC basis. If want to
spend $1050 to air a spot for 30,000 views. If CTR = 2%, then $35 CPM = $0.035, CPV = $1.75, CPC =
$0.025, CPV + $0.50 CPC

Ad Risk Principle

Advertising risk with creative: high vs. low CTR (click thru rate) ads

Operations management (OM) = management (operation, and improvement) of the processes that
transform material, labor, energy, and information into goods and services

Decisions in OM: [1] Strategic decisions (long term impact), [2] Tactical decisions (midterm impact), [3]
Operational decisions (short term impact)

Strategic decisions (long term impact) = Product / process selection, location, capacity

Tactical decisions (midterm impact) = Number of employees, number of hours worked, inventory levels,
order quantity and frequency

Operational decisions (short term impact) = Job scheduling, priorities

Goals of OM: improve efficiency, improve effectiveness, increase value (value = quality / price)
Queuing systems

Customer population = finite or infinite

Arrival rate = constant or variable

Line lengths = finite or infinite

Queue discipline = FCFS, reservation first, highest profit first

Service rate () = constant or variable

Probability of reservice = high or low

Assume the system is running at steady state

Does the time between successive arrivals follow some statistical distribution?
−t
f ( t )=e
How many arrivals might enter the system within a time period T?
n −T
(T ) e
PT ( n ) =
n!
Does the service time follow some statistical distribution?
− μt
f ( t )=μ e
M/M/1 Model Assumptions: Assumptions: infinite customer population, random arrivals (Poisson),
unlimited queue length, single line, FCFS, single channel, single phase

Examples: 1 lane toll bridge with manual pay, drive thru at McDonalds

M/M/1 Equations

Utilization ρ=λ/ μ

Average # of customers in the system Ls =λ/(μ−λ)

2
λ
Average # of customers in queue Lq= =Ls∗ρ
μ( μ−λ)
Average time a customer spends in the system W s=1/ μ− λ=Lq / λ

λ2 L
Average time a customer spends in the queue W q= = q
μ (μ− λ) λ
 n
Probability of n units in the system Pn=(1− ρ)(ρ)

Example: A bank is considering opening a drive through window for customer service. Management
estimates that customers will arrive at the rate of 15 per hour. The teller who will staff the window can
service customers at the rate of one every three minutes. Assuming Poisson arrival and exponential
service times, calculate performance metrics of this queue.

Utilization ρ=λ/ μ=15/20=75 %

Average # of customers in the system Ls =λ/(μ−λ)=15/(20−15)=3 customers

2
λ
Average # of customers in queue Lq= =Ls∗ρ=3∗75 %=2.25 customers
μ( μ−λ)
Average time a customer spends in the system W s=1/ μ− λ=Lq / λ=3 /15=0.20=12 min

λ2 L
Average time a customer spends in the queue W q= = q =2.25 /15=0.15=9 min
μ (μ− λ) λ
Because of limited space, the banker would like to ensure with a 95% confidence (or service level) that
no more than 3 cars will be in the system at any time. What confidence exists currently for no more than
3 cars? What rate would the teller need to operate at to meet a 95% service level? What would be the
utilization at this point?
n
Probability of n units in the system Pn=(1− ρ)(ρ) , r =15/20=0.75
0
0 cars P0= (1−0.75 ) ( 0.75 ) =0.25
1
1 car P1=( 1−0.75 ) ( 0.75 ) =0.188
2
2 cars P1=( 1−0.75 ) ( 0.75 ) =0. 141
3
3 cars P1=( 1−0.75 ) ( 0.75 ) =0.1 05

P=P 0+ P 1+ P2 + P3=0.25+0.188+0.141+ 0.105=0.684=68.4 %

95% service level P0 + P1 + P2 + P3=0.95

4
1−r =0.95 ,r =0.47
r =l/m, 0.47=15/m , m = 32 people / hour
Service rate of 32 people / hr equates to a 47% utilization rate

8 dimensions of product quality = Performance, functionality, durability, reliability, conformance to

specifications, serviceability, aesthetics, perceived quality

Dimensions of service quality = consistency, courtesy, convenience/availability, communication,

accuracy/reliability, timeliness/responsiveness, credibility/trustworthy, security

4 costs of quality
Common v assignable causes of variation

Common causes = Inherent in the process used, unavoidable with current process, can do nothing about
this

Assignable causes = Can be identified, can be corrected/fixed (ex: new operator error)

Control charts are used to identify assignable causes of variation

Upper and lower control limits are set based on common causes of variation for the process (we know
these will lead to a normal distribution)

Data plotting and monitoring is to watch for assignable causes of variation (these are causes of variation
we can do something about)

x_bar chart monitors the mean

Assuming 3 σ limits: UCLx = X́ + A 2∗R , LCLx = X́ −A 2∗R

Mean factor A2

R chart monitors the spread

Assuming 3 σ limits: UCLr = D 4∗R , LCLr = D 3∗R

Upper range D 4 , Lower range D 3

Process Capability Index (Cp)

Cp=(Upper specification−Lower Specification)/6 σ

Cp >= 1.0 indicates process is capable

Six Sigma equates to a Cp >= 2.0

This value only looks at spread, not how well a process is centered on its target value

Process Capability Index (Cpk)

{
Cpk=Minimum of [ upper specificatio n−
x́
3σ }
, {x́−lower specification/3 σ }]

Cpk gives the proportion of variation between the center of the process and the nearest specification
limit
Cpk = 1 means process meets specifications

Cpk < 1 Process does NOT meet specifications

Cpk > 1 Process is better than the specification requires

Example: Coke bottleneck diameter must be .600 +-.050. Above the tolerance and cap will not fit. Below
the tolerance and the cap will fall off. Assume the standard deviation is .012.

Cpk = Minimum of [{upper specification - x́ / 3 σ}, { x́ - lower specification / 3 σ}]

Cpk = Minimum of [{(0.0650 – 0.597) / (3 * 0.012)}, { (0.597 – 0.550) / (3 * 0.012)}]

Cpk = Minimum of [{ 1.306 }, { 1.4272 }]

Cpk = 1.306