If you're considering making the switch to identity based analytics so you can track not only what is happing on your site/app but also who is doing it, then you need to make sure that you set up your data collection correctly.
We'll show you some tips for building an analytics schema and data dictionary so that you can manage your implementation with less trouble. We will discuss:
Surfacing key business questions to inform your metrics
Creating an analytics schema to describe all your events and properties
Adding a data dictionary so that the entire organization will be on the same page
Best practices for adding properties to the user identity to track events for easy report generation
3. Thue is the Kissmetrics Webinar Wizard and
Marketing Ops Manager. Before joining forces with
Kissmetrics, he was a Lyft driver in SF, which is also
how he ended up as a Kissmetrics marketer.
Whenever Thue is not trying to automate everything
around him, you can find him hiking in the Sierras.
THUE MADSEN
Marketing Operations Manager, Kissmetrics
@ThueLMadsen
Before founding SaaS Management, Ryan created a
number of data-driven testing methods that helped
him build multi-million user companies.
RYAN KOONCE
CEO, SaaS Management Group
@RyanKoonce
www.SaaSMgmt.com
5. 1 Section One - Introduction
Step 1: Identify your expert
Step 2: List Key Business Questions
Step 3: List Key Metrics and Definitions
Step 4: Decide How to Collect Data
Step 5: Choose your Tools
Step 6: Create your Analytics Schema
Step 7: Manage your Integration
Step 8: Audit Your Data
Step 9: Build your Baseline Reports
2 Section Two - 9 Steps for Success
3 Section Three - Q & A
TABLE OF CONTENTS
7. A LITTLE SEMANTICS
• Analytics is the process of testing a hypothesis based
on the data available
• Reporting is the ability to extract information from your
analytics system
• Business Intelligence is a fancy way to say analytics
• Data Visualization is drawing pretty pictures of
analytics
• Big Data + Data Science + Predictive is magic
• Data science generally doesn’t have ideas. It just tells you stuff (beer
+diapers).
8. 9 Steps for Success
BE THE HERO OF YOUR ORGANIZATION
9. STEP 1: IDENTIFY YOUR EXPERT
• You are a VP and doing your own analytics &
reporting
• You are a marketing manager who has been
tasked with “figuring it out”
• You are an engineer who has been asked to
“check the database”
UH-OH Scenarios
10. STEP 1: IDENTIFY YOUR EXPERT
Dedicate Someone to Analytics
and Reporting!
11. Key Business Questions come before Key Metrics
We’re going to focus on Marketing (Customer Acquisition) and Product (Engagement)
STEP 2: MAKE A LIST OF KEY BUSINESS QUESTIONS
Customer Acquisition
• What were our site visits per channel?
• How much money did we spend per visit by channel?
• For users that converted to paid subscribers, what activities were they most
likely to preform in their trial period?
12. Key Business Questions come before Key Metrics
STEP 2: MAKE A LIST OF KEY BUSINESS QUESTIONS
Engagement / Retention
• What is our revenue per site visit by engagement campaign (retargeting,
email)?
• How many photos did users download in their free trial by cohort?
• How many photos did active subscribers download compared to members
who canceled?
13. Key Business Questions come before Key Metrics
STEP 2: MAKE A LIST OF KEY BUSINESS QUESTIONS
Avoid non-actionable questions
• What is my bounce rate?
• How many total registered users do I have?
14. KEY QUESTIONS INFORM KEY METRICS
Which digital marketing channel is the most
profitable?
• Profit = Revenue Generated by User - Cost of Acquisition for User
• Marketing Channel = Campaign, Source, Term, Medium (think UTM
Parameters)
What kind of user engagement results in more
free trial conversions?
• Free Trial User Engagement = How many times did a user do event 1 AND
event 2 during their trial period that converted to paid subscribers?
15. A few examples:
STEP 3: LIST YOUR KEY METRICS AND DEFINITIONS
• Site Visits - Anytime someone visits the site within a 30
minute window.
• Revenue - Total revenue collected from customers
• Revenue Subtotal - Gross Revenue less discounts and
tax
• Products Sold - The total number of products sold
• Landing Page View - A Site Visit to a specific landing
page
• Login - Any time someone signs into the site
16. It’s time to look past the Javascript Snippet.
STEP 4: DECIDE HOW YOU WILL COLLECT YOUR DATA
Identify and Track
• Measure not only what happened, but who did it.
• Segment, Kissmetrics, AttributionApp.com
• SQL or Data Modeling experience generally not required
Data Warehouse
• Store your data in Amazon Redshift, PosgresSQL, etc.
• Looker, Mode, Periscope (Segment.com or TreasureData as translation
layers)
• Query data directly - SQL or Data Modeling generally required.
How you decide to collect your data will help inform what tools to use.
18. There isn’t one tool that rules all across devices
You will need data from everywhere
STEP 5: CHOOSE YOUR TOOLS
•A/B Testing - Optimizely, VWO, Taplytics
•Attribution - AttributionApp, Branch
•Analytics - Kissmetrics, Keen, Woopra
•CRM - Salesforce
•Email - Customer.io, Sendgrid,
Mailchimp
•Referrals - Curebit, Extole
What tools you choose will factor in how you structure your data.
19. STEP 6: CREATE YOUR ANALYTICS SCHEMA
Should you fire events client side or server side?
20. REQUEST A DEMO
Start optimizing your marketing today with a free
personal demo of Kissmetrics
21. STEP 7: MANAGE YOUR EVENT INTEGRATION
Why Your Engineers (Secretly?) Hate You*
• They are really busy and don’t have time for “marketing”
• You asked for one too many tags
• You tell them to pull you a report of “stuff” from the database
* If this slide is not relevant to you, then simply Slack them a link to the
analytics schema and head to the beach/mountain etc. for some fun.
22. STEP 7: MANAGE YOUR EVENT INTEGRATION
Make Your Engineers Love You
• Wow them with your analytics schema
• Explain how the pieces fit together
• Triage the events so they can more easily fit them into their schedule
• Bribe them with beer or scotch
23. STEP 7: MANAGE YOUR EVENT INTEGRATION
** Statements like this are why your engineers (secretly?) hate you.
An experienced engineer can usually implement an
analytics schema in about one day.**
24. STEP 8: AUDIT YOUR DATA
•Audit in a Staging environment before going to
production - Identify and Track data is forever
•Confirm the events are collected from all
relevant places, on all devices.
•Confirm that identify is being called in the
appropriate places.
•Confirm anything you can by comparing it
directly to the data in your database.
25. STEP 9: GENERATE YOUR BASELINE REPORTS
•Using your relevant tools, generate the reports
that answer your initial business questions.
•Call a meeting. Make Presentation. Get Raise.
•Your initial reports will create even more
questions, for more analysis and reports.
•Using the data, develop optimization sprints,
test hypothesis and implementation plans.