2. Confidential & Proprietary
Main topics for this session
1. Choosing the right metric to focus on
CPC? CPA? PNO? ROAS? ROI? Profit? CLV?
(Goal: Understand differences between various KPIs)
2. CLV: The Holy Grail of Customer Centricity
(Goal: Provide overview of customer strategies and relevant analytics)
3. Actionability
(Goal: Put chosen KPI into practice)
4. Marketing experiments
(Goal: Quick introduction to marketing experiments)
5. Confidential & Proprietary
How easily can you predict customer behavior?
History Present Future
Customer 1
Customer 2
Customer 3
Average number of transactions in a learning period
6. Confidential & Proprietary
Your 11 projects work under specific conditions
11/11 Focused on end customers (B2C)
11/11 Non-contractual settings
11/11 Non-membership status
11/11 Always-a-share (vs. lost-for-good)
11/11 Continuous buying
11/11 Variable-spending environment
11/11 Partial identification possibilities
7. Confidential & Proprietary
What are your performance marketing goals?
How many of you are oriented on
profit? Revenue? New customers?
What else?
10. Confidential & Proprietary
A
B
5 %
10 %
PNO
Which campaign is performing better?
Podíl nákladů na obratu
(Cost to Revenue Ratio,
Effective Revenue Share)
= Cost / Revenue
13. Confidential & Proprietary
CPA Conversions
Average
order value
Total
revenue
Media
costs PNO ROAS
Total
Profit
(simplified
without product
costs)
$50 10 100 € 1 000 € 50 € 5 % 20 : 1 950 €
10 € 30 100 € 3 000 € 300 € 10 % 10:1 $27,000
5 €A
B
20:1
2 700 €
Key idea:
Total profit balances the trade-off between unit costs and volume
5 %
… well, it depends
14. Confidential & Proprietary
Focus of performance marketing
CPA
ROAS
Profit
Revenue
Value Optimized
ROI and Total
Profit Optimized
Conversions
Cost OptimizedCPC
Clicks
Cost Optimized
PNO
(CtRR, ERS)
Conversions
Cost Optimized
Long term
Profit
Customer Equity
Customer Lifetime Value
Optimized
18. Confidential & Proprietary
Finding elasticity using Bid Simulator
Use Google’s Bid Simulator for estimation of elasticity on specific campaigns.
19. Confidential & Proprietary
What simplifications do you cover?
Seeing all touchpoints
Attribution of conversions and costs
Word of mouth and referrals
Cross-environment behavior
(cross-device, omnichannel)
Data in the right moment
Individual campaigns vs. portfolio approach
Future incremental purchases
Past behaviour of a customer (new customer?)
Variable spending
Volume of sales
External and indirect effects
Averages vs customer heterogeneity
21. Confidential & Proprietary
Can you choose your KPI?
CPA
ROAS
Profit
Revenue
Value Optimized
ROI and Total
Profit Optimized
Conversions
Cost OptimizedCPC
Clicks
Cost Optimized
PNO
(CtRR, ERS)
Conversions
Cost Optimized
Long term
Profit
Customer Equity
Customer Lifetime Value
Optimized (CLV)
22. Confidential & Proprietary
Recap of Part 1
CPA
Cost optimization
ROAS, ROI, PNO
Revenue optimization
Profit, CLV, CE
Profit Optimization
24. Confidential & Proprietary
“ Customer centricity is a strategy that
aligns a company’s development and
delivery of its products and services with the
current and future of a select group of
customers in order to maximize their long-
term financial value to the firm”
25. Confidential & Proprietary
“ Customer Lifetime Value (CLV) is the
present value of the future (net) cash flows
associated with a particular customer “
“ Customer Equity is the sum of customer
lifetime values across a firm’s entire
customer base“
http://noca.cz/clvbook
27. Confidential & Proprietary
Pareto analysis
How many customers is the company currently dependent on?
Are there customers who causes losses?
Visualizations via decisivedata.net/blog/pareto-charts-in-tableau/
29. Confidential & ProprietaryConfidential & Proprietary
Important
questions on
profitability
Who are our most profitable customers
now? And next year?
How do the most profitable customers
differ from one-time buyers?
Who has the most discounted
purchases?
Who makes the most returns?
How much can you invest in a customer
and still be profitable?
How long do you need to wait for ROI?
30. Confidential & Proprietary
Firmy.cz
Cost per click 9 Kč
Cost per action 135 Kč
First purchase value 1 842 Kč
More B2B office customers
Incremental purchases within the first year
Example of TisknuLevne.cz (2010)
Zboží.cz
Cost per click 1 Kč
Cost per action 64 Kč
First purchase value 1 749 Kč
More B2C home customers
Low ratio of returning customers
http://old.webtop100.cz/files/2010/martin-penicka.pdf
31. Confidential & Proprietary
Kevin Hillstrom
There is a direct correlation
between annual repurchase rates
and the length of time you are
willing to wait for payback.
http://blog.minethatdata.com/2015/10/lifetime-value.html
32. Confidential & Proprietary
How does CLV look like in reality?
For each customer you typically estimate lifetime profit
(discounted in following years).
I’ve found out that for better actionability it is useful to
estimate profit for some shorter term: ¼ to 3 years. This
should be selected depending on the nature of business
when your customer have high probability of
repurchasing.
Also, you typically calculate CLV each month/week/day in
order to see how your predictions evolve.
33. Confidential & Proprietary
How does CLV look like in reality?
Your monthly calculations for each
customer. Naturally, CLV models
change when customer purchases and
“fade out” the value when the customer
is inactive.
34. Confidential & Proprietary
Main areas where online marketers
can benefit from CLV
Theoretically:
➔ Customer Acquisition - Expansion - Support - Retention
➔ Direct Campaigns
➔ Customer Intelligence (CRM, managerial reporting)
Ideas of use cases like those mentioned on The Wise Marketer, on Econsultancy
and Custora are nice, but lack details.
35. Confidential & Proprietary
1) Customer Acquisition
How much can we afford to pay for a new customer? What is the true value per
acquisition? Should that differ from CPA / ROAS targeting?
What products drive higher CLV?
How fast can we estimate CLV for a fresh user/customer?
When can we compare CAC and Historical Profit + CLV?
36. Confidential & Proprietary
2) Customer Expansion
For what segments of customers should we increase/decrease marketing
activities (/costs)? When?
When can we push marketing efforts on fresh customers?
What is the impact of (up|x)-selling on CLV?
What less profitable customers should we remove from mailing?
What if CLV estimation rises?
37. Confidential & Proprietary
3) Customer Support
Who can (not) be given a discount?
Who should wait in a queue for a support?
Should we give a customer a gift or an exclusive deal?
38. Confidential & Proprietary
4) Customer Retention
How much can we afford to pay to retain a customer and still being profitable?
What to do when CLV estimation drops?
How to treat customers with low or negative CLV?
Should we give incentives when CLV rises?
39. Confidential & Proprietary
5) Direct Campaign Evaluation
Decide which customers by CLV to select for a campaign.
Can we get top 10% customers by CLV?
Use ratio of CLV as max costs per campaign.
Evaluate campaign by change in CLV.
40. Confidential & Proprietary
6) Customer Intelligence and managerial reporting
of your customer base
How does CLV/Customer Equity evolve? For various companies, markets,
customer types, segments of customers.
How well can we forecast sales?
What activities can you do to support the growth?
Where will high profits come from? What are profit drivers?
43. Confidential & Proprietary
Actionability concerns
Reporting and optimization for the chosen KPI
Using the data: support of bidding mechanisms
Having the right technology platform for all of it
Opportunity to increase investments for incremental conversions?
Managing KPIs across several campaigns = optimizing account-level KPI by
weighting individual campaign’s contribution.
44. Confidential & Proprietary
Advertising Platforms, with
Measurement enabled
Measurement Platforms, with
Ad Tracking enabled
Measures
Google Media
Measures
All Media
Online/app
measurement
Online/offline
marketing mix
Google’s Measurement Solutions
45. Confidential & Proprietary
Reporting support in AdWords
Native support for CPC, CPA, ROAS, ROI. Calculated column for PNO
Possibility to import profit
46. Confidential & Proprietary
Reporting support in Analytics
Revenue via e-commerce. Calculated column for PNO. Custom metrics for profit.
Cost Import for non-Google ad costs.
47. Confidential & Proprietary
Support of bidding mechanisms in AdWords
Reporting and optimizing on Conversions.
Conversion Optimizer, eCPC, Target CPA.
Targeting on ROAS usually requires more conversion data.
Experimenting with multiple bidding strategies (AdWords Drafts and Experiments)
48. Confidential & Proprietary
Recap of Part 3
Reporting
Choosing the right
measurement platform
Optimization
Ability to use data for
manual or automatic
campaign optimization
Data Integration
Using the right platform
50. Google Confidential and Proprietary
1996
1997
1998 1999
2000
14.8
15.0
15.2
15.4
15.6
15.8
16.0
200 250 300 350 400 450 500 550
TotalUSHighwayFatalityRate
Fresh Lemons Imported to USA from Mexico (Metric Tons)
Any clue on what really happened here?
51. Google Confidential and Proprietary
Experimentation Basics
Key Terms
All definitions from Wikipedia.
Experiment CorrelationControlled
Experiment
Causality Uplift
Orderly procedure
carried out with the
goal of verifying,
refuting or establishing
a hypothesis
Experiments that provide
insight into cause-and
effect by demonstrating
what outcome occurs
when a particular factor
is manipulated
Refers to any of a
broad class of
statistical relationships
involving dependence,
which is any statistical
relationship between
two random variables
or two sets of data
The relation between an
event (the cause) and a
second event (the effect),
where the second event
is understood as a
consequence of the first
Usually defined as the
difference in response
rate between a treated
group and a randomized
control group, allowing
marketers to isolate the
effect of a marketing
action and measure the
effectiveness or
otherwise of that action
52. Google Confidential and Proprietary
Experimentation Basics
Questions to get started
Do you know how experimentation can prove the value of your marketing efforts?
Are you relying on correlations, or can you base decisions on measured
causal impact?
Do you know the incremental value of each of your media investments?
Do you have a plan to evaluate the effectiveness of new marketing channels?
53. Google Confidential and Proprietary
Experimentation solutions can help you measure
the value of your marketing efforts
Time Series
Experiments
User-level
Experimentation
Campaign
Experiments
Marketing Geo
Experiments
thinkwithgoogle.com/articles/proving-marketing-impact.html
54. Confidential & Proprietary
Recap of this session
1. Choosing the right metric to focus on
CPC? CPA? PNO? ROAS? ROI? Profit? CLV?
2. CLV: The Holy Grail of Customer Centricity
(Managing CLV and Customer Equity)
3. Actionability
(Putting chosen KPI into practice with your campaigns)
4. Marketing experiments
(Controlled experiments to understand causality)