2. Today’s topics
“Frameworks are brilliant because rather than teaching a specific solution, they teach the reader how to think.
The best possible gift you can give them.” - Avinash Kaushik
● Problem: squirrels staying in their silos, not focusing (enough) in a
structured manner on insights, recommendations and business impact.
● Solution: frameworks and tools to answer each of the main questions:
1. Insights: why did metrics change?
2. Recommended Actions: what should we do to improve results?
3. Business Impact: what can we expect after implementing these recommendations?
4. Prioritization: based on impact and ‘time-to-useful’, what should we do first / spend most
of our time on and how should we align our metrics with our efforts and
leadership altitude?
● Wrap up
3. First, some definitions…
Reporting Squirrel Analysis Ninja
“Reporting Squirrels spend 75% or more of their time
in data production activities. The primary
manifestation of this is in creation of reports for their
direct leader, or team or division or bunch of people.”
“Analysis Ninjas spend 75% or more of their
time in analysis that delivers actionable
insights. The primary manifestation of this
is expressed in English (or native country
language).”
VS
12 Signs To Identify A Data Driven Culture
4. And just in case you like squirrels..
If your job description looks like this:
Compile and compute data according to statistical formulas for use in statistical studies. May
perform actuarial computations and compile charts and graphs for use by actuaries. Includes
actuarial clerks.
WillRobotsTakeMyJob.com
How Fast Will You Lose Your PPC Job to a Robot? A comparison of 5 Marketing Areas
5. Let’s increase your ‘robot survival’ chances threefold
If your job description looks like this:
Develop or apply mathematical or statistical theory and methods to collect, organize, interpret,
and summarize numerical data to provide usable information. May specialize in fields such as
bio-statistics, agricultural statistics, business statistics, or economic statistics. Includes
mathematical and survey statisticians.
6. Let’s increase your ‘robot survival’ chances even further
If your job description looks like this:
Formulate and apply mathematical modeling and other optimizing methods to develop and
interpret information that assists management with decision making, policy formulation, or
other managerial functions. May collect and analyze data and develop decision support
software, service, or products. May develop and supply optimal time, cost, or logistics networks
for program evaluation, review, or implementation.
7. So Avinash, how do we become an Analysis Nina?
• Puke data, unless it’s a customized data puke (CDP) for tactical and siloed decisions.
• State the obvious:
“As you can see graph one shows that visits in Sept. are down by 4%.”
They can see it, the graph is right there!
“CPA went up because the conversion rate went down.”
It’s either that or because of a higher CPC, but why did the conversion rate go down?
Must-read:
Digital Dashboards: Strategic & Tactical: Best Practices, Tips, Examples
8. So Avinash, how do we become an Analysis Nina?
• Select a handful of KPI’s for trended graphs and tables and compare performance with the
target and across segments.
• Use visuals to highlight key elements.
Write out the corresponding IABI – our focus of today:
• Insights: why did KPI’s change (causal factors)?
• Actions (recommendations / next steps): what do you recommend based on the performance
and insights?
• Business Impact: what is the expected impact on the business if the
recommendations are implemented (this is hardest part)?
11. 1: Insights – Why and how do metrics change?
It’s best to have a structured approach when you want to answer questions like:
Looking backward:
• Why did CPA go up?
• Why did revenue go down?
• Why is our competitor outranking us?
• Why aren’t our Shopping ads showing on X?
Looking forward:
• How can we increase our Quality Score?
• How can we get 10% more revenue?
• How can we increase our conversion rate?
This structured approach is called “Root Cause Analysis (RCA)” and you may need
5 Whys to get to a root cause.
12. 1: Insights – Root Cause Analysis Flowcharts
First time I fell in love with a PPC flow chart:
Rimm-Kaufman’s (now Merkle) dossier 3.2 in 2012
PPC Performance Troubleshooting Made Easy
13. 1: Insights – Root Cause Analysis Flowcharts
AdAlysis and Optmyzr have built-in root cause analysis:
14. 1: Insights – Root Cause Analysis Flowcharts
I’ve expanded on the Rimm-
Kaufmann, AdAlysis and Optmyzr
flow charts and created a new
version to capture and categorize
all PPC causal factors.
Download it for free (no email
needed) at:
trueclicks.com/goodies
Share if you like it!
16. 2: Recommended Actions – What should we do?
Now you’ve found the root causes of the changes you (want to) see, which actions will you
recommend?
If you really defined the root cause(s) thoroughly, the recommendations should flow
naturally.
These will sound like: “Metric x is down because of our inability to take advantage of trend
y and hence I recommend we do z.”
Or: “We missed our target for customer satisfaction because our desktop website performs
horribly on mobile platforms hence we should create a mobile friendly website.”
English. Your insights. The so what based on data you've summarized and snapshotted.
From: Digital Dashboards: Strategic & Tactical: Best Practices, Tips, Examples
17. 2: Recommended Actions – What should we do?
Google Ads has its recommendations built in:
18. 2: Recommended Actions - The TrueClicks™ Framework
After performing 100’s of audits, while writing the AdWords audit blog series and while building
TrueClicks, I realized all PPC analysis & corresponding recommendations can be categorized as following:
Issues to monitor daily (scripts or 3rd party tools):
1. Urgent Issues: critical things that are broken and
need your attention right now.
2. Errors: things that are missing, disapproved or
other mistakes that have been made (by other
people of course) Just fix it asap.
3. Account activity: lack of recent and meaningful
changes Stop being so lazy.
4. Anomalies: large and sudden deviations in KPI’s
Use root cause analysis flow charts to explain.
5. Budget Issues: (expected) deviations from target
budget Adjust as needed (and overthrow the
tyranny)
Issues and opportunities to audit or analyze:
1. Essentials: everything that should be
implemented or improved first, as it’s essential
for any successful PPC campaign.
2. Cost-saving opportunities: based on identified
waste.
3. Growth opportunities: everything that could
increase reach, traffic or conversions.
4. Account hygiene issues: to improve account
management & organization.
19. 2: Recommended Actions - The TrueClicks™ Framework
Monitoring
(daily)
Urgent issues
Account down
Broken / out of
stock landing
pages
Conversion
tracking down
Errors
Disapprovals
Conflicts
Accidental
mistakes
Anomalies
Sudden and large
changes in cost,
conversions, etc.
Budget
deviation
Compared to
monthly target
budget
Auditing
(weekly)
Essentials
Tracking
Landing pages
Quality score
Ad testing,
extensions,
granular ad
groups
Cost-saving
opportunities
Negative
opportunities
Underperforming
segments (e.g.
devices,
locations)
Underperforming
targeting options
Growth
opportunities
Impression &
click share
Newer extensions
Top of page
visibility
Audiences
Unused targeting
options
Account
hygiene issues
Duplicate
keywords & ads
Consistent
naming
convention
Non-serving
keywords
20. 2: Recommended Actions - The TrueClicks™ Framework
We can also subcategorize each recommendation based on the time it would take to
implement or fix them:
1. Quick wins: can usually be implemented within 1 hour per account.
2. Initiatives: will take longer than 1 hour.
So let me share a Google Sheet with you that combines 130+ Google Ads checks
using these action-oriented categories, to be found on
trueclicks.com/goodies and click on the Google Ads checks link
By filtering you can quickly zoom in on the categories you’re interested in.
23. 3: Business Impact
What will be the impact on the business if the CXO accepts your recommendation and
the business takes action? I almost never see this.
A small part of the problem is that Analysts often don't have the skills to compute
impact of the recommended actions.
A bigger part is that it is actually quite a bit of effort to compute impact.
But, what better way to create a sense of urgency than tell the CXO what the expected
outcome will be if they do based on your insights and recommended actions?
From: Digital Dashboards: Strategic & Tactical: Best Practices, Tips, Examples
24. 3: Business Impact
Based on the categories of the TrueClicks Framework™, let me share the following guidelines to
inspire action for each category. It’s always best if you can also compute the expected impact (in
terms of cost, revenue or even profit).
• Urgent issues, errors and essentials: focus on the terrible things that will happen (or are
happening) by not fixing the issue asap, such as: a negative user experience, reputation
damage, being in the dark without tracking, not showing ads at all, etc.
• Cost-saving opportunities: this is the easiest category, just use the cost of the identified waste
and use that to calculate x in a statement like “by adding these negatives we will save £ x per
year”
• Growth opportunities: if you can find them, use benchmarks from other similar accounts or
from Google (they often mention uplifts of beta users when announcing a new feature on the
Google Ads blog) like “by adding the price extension we expect an 18% increase in CTR,
resulting in 231 more clicks, £ 173 more cost and 7 extra conversions per month”. Make sure
the reader understands this isn’t a promise, but an expectation based on historical data /
benchmarks seen elsewhere.
• Account hygiene issues: focus on the time savings of a better organized and cleaned-up
account.
26. 4: Prioritization
A classic and proven way to prioritize actions is by plotting them in an impact-effort matrix
But if you can’t (quickly) quantify the impact and effort, other frameworks can help you out.
27. 4: Prioritization within The TrueClicks™ Framework
Monitoring
Urgent issues Errors Anomalies
Budget
deviation
Auditing
Essentials
Cost-saving
opportunities
Growth
opportunities
Account
hygiene issues
You can work on your recommendations from left to right,
starting with the quick wins (if any) within each category
28. 4: Prioritization with the Impact Matrix
Avinash strikes again with another must-must-read:
The Impact Matrix | A Digital Analytics Strategic Framework
(July 24, 2018)
• To paint a simple picture of the big, complicated world of analytics in a 2×2 matrix.
• Each cell contains a metric.
• The business impact is on the y-axis, illustrated from Super Tactical to Super Strategic
and is exponential.
• The time-to-useful is on the x-axis, illustrated from Real-Time to 6-Monthly.
29. 4: Prioritization with the Impact Matrix
“If you understand why each metric is where it is, the rest of this post will fill
you with euphoric joy rarely experienced without physical contact.”
30. 4: Prioritization with the Impact Matrix
46 metrics plotted on impact and time-to-useful, now what?
31. 4: Prioritization with the Impact Matrix
1: Assess Maturity: What metrics are most commonly used to make decisions
that drive actual actions every week/month/more?
35. Wrap Up
1. Always use the IABI format whenever you perform an analysis and when you write out
recommendations.
2. To help you write out solid IABI’s, use these tools and frameworks:
Insights:
- Root Cause Analysis flowcharts (including 3rd party tools)
- 5 Whys
Recommended Actions:
- Google’s built-in recommendations (if you agree)
- The TrueClicks Framework™
Business Impact:
- Quantify in terms of cost, revenue or profit if possible
- If not, describe the (possible) outcomes of (not) implementing the recommendations
3. Assess your companies’ analytics maturity using the impact matrix and move from
Solid Foundation Intermediate Advanced
4. Align your metrics with the corresponding leadership altitude and
analytical effort