BEST Practices Seattle 2012 Live Updates

December 6, 2012


This is  LIVE BLOG – we’re updating in real-time from Analytics Pros BEST Practices Seattle Conference 2012 at the Seattle Art Museum in Seattle WA.


Follow us on Twitter as well @analyticspros for shorter updates throughout the day, and every day!



3:14 pm, Building a Successful Optimization Program, Brendan Regan, Analytics Pros

Brendan Regan presenting at BEST Practices in SeattleHow to be thinking of optimization and build a successful optimization program.

The art and science of increasing the efficiency of a web property.

Why Optimize?

  • Efficiency = More outputs and less inputs
  • Outputs = Sales, leads
  • Higher efficiency = Better marketing ROI

Optimization Philosophy

How to approach optimization and what to do:

  • Data driven – data drives testing, not opinion (quantitative and qualitative)
  • Be agile – a small lift next week is better than larger lift next year
  • Lift + learn
  • Gamble – some tests will lose, greater risk=greater potential
  • Baby steps – do what’s feasible asap, and build to more complex tests
  • A process, not project – long term goal is testing as apart of marketing operations

The people and skill set needed:

  • Business/Marketing Acumen
  • Data Analysis
  • Heuristic/Usability/US
  • Visual Design
  • Optimization SME
  • Copywriting
  • Technical SME
  • Project Management

An easy approach:

  • Define the problem
  • Research/observe the problem
  • Form a hypothesis
  • Conduct experiment
  • Analyze experiment results
  • Form a conclusion
  • Socialize results

Technologies needed:

A list of technologies needed to get started.

  • Coordinate resources
  • Design variations
  • Document requirements
  • Conduct experiment
  • Analyze results
  • Socialize experiment results

Do these 4 things tomorrow!

A quick list of action items to get you started tomorrow.

  • Conduct “gap analysis” on people, skills, and tools. Do we have enough resource allocation? Do we need new tools? Do we have all the necessary skills in-house?
  • Get a testing tool in place. Do we have IT support? build vs. buy? Price?
  • Build a business case. Opportunity cost? Tying to corporate goals? Who is scared of testing?
  • Get resources and/or outside help. Agency partners? Consultants/specialists? Testing vendors? Training?

1:20 pm, 9 Google Analytics Best Practices, Caleb Whitmore, Analytics Pros

Caleb Whitmore presenting at Best Practices in SeattleImplementing GA, what to do and what NOT do to!

Plain Vanilla Tag – GA is very easy to get started, but there are many, many things that should be changed in that tag.

10% – 30% of your Google Analytics tracking tag needs to be changed.

There are 55 documented settings for the tag! The three most important (that cause 80% – 85% of the pain).

  • Cookie Domain => setDomainName
  • Campaign Anchors => setAllowAnchor
  • X-domain Linking => setAllowLinker

Audit your data for quality and look for internal referrals, suspicious “direct” landing deep within your site, or odd exit pages with high rates.

Data Quality Checking Tools: Analytics Health CheckAnalytics Checkup, ObservePoint.

Implementation Tips:

  • Plan your business requirements
  • Implement the GA tag correctly
  • Take a course, read a book

Real World Example: Inaccurate tracking data in GA from paid AdWords campaign.

The Breakdown (here’s what needs to technically be happening):

  • AdWords Auto-tagging
  • Tagging must make it to landing page
  • GA tag must work on the site
  • GA tag must work for all parts of the site

External Data Integration with Google Analytics

3 things to keep in mind:

  • Import data into GA
  • Export data out of GA into a 3rd party tool – Example: Data exported with a unique session key
  • Push external data into GA – Example: Data from your CRM surfaced when a user is signed in

How it works – know your data model

Where will the date come from? How will it be structured? Will it need transformation? What are you going to do with it?

Data Types in GA:

  • PageViews
  • Events
  • Commerce
  • Social
  • Custom

Rules for Data: Pageviews

  • Use for tracking page loads
  • Can be customized to record a custom “url” or “page”

Rules for Date: Events

  • Use for everything between the pageview
  • Supports 3-dimensional data model (category, action, label, value) – can be re-ordered
  • Two types: Interaction vs. non-interaction

Rules for Date: Commerce

  • For transactional and product purchase information
  • Use only when money changes hands

Rules for Date: Social

  • Only for measuring social interactions – like/tweet/share, clicks to follow, clicks to social sites

Rules for Date: Custom Variables

  • Develop a data model around custom variables – pages, session, visitor

How many users performed this action this many times on the site?

Understand the type of information you are trying to capture and set a custom variable accordingly:

  • Session level vs visitor level custom variables


Know how your audience is using the site (researchers vs. bookers) and make sure to define the score and value. Then use the goal value to define a goal value framework (our workbook) to help you setup and analyze performance. Benefits include:

  • No longer blind to what happened before the conversion
  • See “funnel” view for traffic sources
  • Use “score” to weigh goals

Setting a goal value using a dollar amount won’t bleed over to ecommerce value.

Goal Flows

Some tips for using goal flow:

  • Flow reports are instant
  • They are retro-active
  • Use flow to “group” pages and see flow between them

Campaign Tagging

Need to be segmenting traffic sources. Google Analytics can track any kind of marketing, not just Google AdWords – email marketing, Microsoft AdCenter, display. If you don’t tag email, they will show up as direct. If you don’t tag paid search it will show up as organic.

If you don’t tag all of your traffic all of your data gets mixed up! Need to tag everything.

GA’s order of operations:

  • Looks for campaign tags
  • Looks for referrer
  • Otherwise direct/(none)

>> Free URL Builder from Google Analytics. <<

Multi-channel Funnels

Quick tips:

  • A “big-picture” tool so don’t get lost in the weeds.
  • Look beyond the last source (look at assists and first)
  • Avoid the path tool

Look at “Assisted/Last Interaction Conversions”. Awareness at the top, direct at the bottom.

Multi-tasking the UI

Duplicate your tab in Chrome, will give you a duplicate and retain browser back button information.

Technology Analysis

Look at browser versions to glean user experience / compatibility issues. Use browser-based analysis to find technology problems.

Analyze goal experience and screen resolutions.

Entry Point Analysis

Your website doesn’t have a front door. Your website is a Yurt with no walls! Treat multiple areas of your website like they were your home page (blog for example).

How porous is your site? Where to they come from? Where do they leave?

Real-Time Reports

Use this for instantaneous (1-2 seconds) verification of  data.


  • Is the website online?
  • Is GA working?
  • Did the campaign launch?
  • What’s trending?
  • Anyone see the TV Ad?

Remarketing Tool

Nifty aspect of the remarketing tool in GA. You can create a couple of filters of users that can be analyzed in a lookback window. (Example: how many users upgraded from iOS 6.0 to 6.0.1). Useful to find out information across sessions.

11:10 am, Google Analytics New Features, Ian Myszenski, Google

Next Generation Measurement.Ian Myszenski presenting at BEST Practices

Enable Better Decisions and Excel in Speed, Usability, and Reliability.

  • Fast – spend time in areas that add more value
  • Trusted – a reliable platform
  • Easy to Use – easily measure all touch points (Zero Moment of Truth)

Need to make sure your platform can be administered, scaled, and integrated with other systems. New feature updates to Google Analytics every week (beware: feature overload).

Feature Highlights

  • Google Tag Manager: First free tag management solution. Gives marketers speed and flexibility to better manage and track campaigns.
  • Real-Time: What is happening on your site right now and get instant feedback so you can adjust quickly.
  • Social Reports: Getting a sense of which sources drive visits and how your content is being shared – and how that impacts your goals.
  • Mobile: Device proliferation is going faster than expected. Know which devices your customers prefer. Are you measuring? What devices do your users prefer? How do mobile ads perform?
  • Remarketing: In 2010, 5 sources before making a decision in 2012, that increased to 10 sources. Continue the conversation after they’ve visited your website.
  • Attribution: No long a last-click world. See which sources drive the results you want.
  • Universal Analytics (coming soon!): Moving away from web analytics to digital analytics and see the whole picture with (1) simple open measurement protocal, (2) user ID control to centralize the information, and (3) the ability to add offline conversion data (longer purchase cycles).
  • Premium: Enterprise solution with more access to data, tools and support. More information.

Google Analytics Premium Case Study: Gilt. Flash deal business focused on premium products  Needed to improve self-service use of data across the site to get a holistic view of their user. Used insights to understand customer behavior and get attribution insights.

Benefits: (1) more holistic view of audience, (2) data driven culture and (3) cross-channel measurement and attribution.


9:42 am, Practitioner Keynote, Jesse Nichols, Google

Jesse Nichols from Google at BEST Practices in SeattleHow do you determine if you analytics solutions is producing value? How do we achieve higher ROI?


Ask: What are we measuring?

Then ask: So what?

Where do we typically begin?

  1. I have a tool
  2. I have chosen a few metrics to monitor
  3. I check every once in a while

So what? What is the connection between those metrics and your business goals? Measure things that matter! A way to start measuring what matters:

  • Ultimate Goal
  • Key Actions – what do users do to complete the business goal
  • Audiences – whom is each action relevant to
  • Steps – what is required to complete these actions
  • Items – what must be tracked in order to measure

Key Actions

Start with the ultimate goal and work back to items that must be tracked. But don’t just look at 5% conversion, look at the other “slices” of activities done on your website (doing research, signing up for email, locating a store, etc) and try to find true failure rate (might just be 20% as opposed to 95%).

Optimize to reduce those failure rates.


Break down audiences into relevant segments and what actions they hope to take. Some examples:

  • Researchers
  • Converters
  • Customers

How can I influence those actions? What are the steps that each audience member takes? What keeps someone from moving along this process? What guides them seamlessly through it?

Start optimizing based on answers to these questions.

From Strategy to Reality:

A step-by-step approach:

Plan > Design > Implement > Refine > Optimize

Planning Tool: Digital Marketing & Measurement Model (Avinash Kaushik). Link:

Design: Identify measurable items. What elements can be measured?

  • Key actions
  • Features
  • Characteristics of visitors

Keep some key questions in mind to help stay focused:

What am I willing to change? What is worth changing? What should I not measure?


Not only do you need core page tracking but every site should be tracking the following comprehensive items as well (the hard part):

  • Campaign Tags
  • Custom Variables
  • Event Tracking
  • Social Plugin Tracking
  • Ecommerce Tracking
  • Virtual PageView Tracker


Revisit your design and implementation to make sure your data is usable.

  • Context
  • Governance
  • Cleanliness
  • Integrations


Making the insight easier to gather.

Fast: Custom Dashboards, Custom Alerts

Easy: Custom Advanced Segments, Custom Reports

Once we have clean data, then we can do proper analysis that leads to action.


Analytics has an attribution problem. What is the ROI of analytics – aka ROA (Return on Analytics)?

Can we quantify the impact? Can we account for long term? Can we factor in all elements?


ROA = [(Ra – Rm) * (d)] / Ia

Rm – return on media

Im – investment media

Ra – return on media post analysis

Ia – investment in analysis

d – duration


Example: If a $10K spend on “analytics” improves a $40K per month return to $60K per month  for a 12+ month campaign then,

ROA = [($60K – $40K) * 12] / $10K = 2,400%




9:13 am, Welcome Keynote, Caleb Whitmore, Analytics Pros

Off to a great start, going over agenda and how to build an optimization strategy based on digital analytics.

Cause of Failure:

The following short list is

  • No strategy
  • No plan
  • Bad implementation
  • Data Overload – analysis paralysis
  • Distrust of data
  • Barriers to reaching data
  • Lack of action after insights

Strategic Digital Analytics Framework

Once you understand the main causes of failure, you can address those  challenges through the following framework. See also Caleb’s blog post on Strategic Digital Analytics: A Framework for Success.

  • Strategy – ask the right questions, aligning your data with your business
  • Capability – get the right team coverage (build the dream team – see below), get the right tools
  • Insight – actively glean the right information and data
  • Action – prioritize and make changes that improve performance


Capabilities: The Dream Team. The dream team is composed of the following key players mapped back to the team capabilities framework:

  • Strategist – Business In
  • Super Developer – Tech In
  • Data Miner – Tech Out
  • Analysis Ninja – Businees Out

Team planning model for Digital Analytics services

Understanding this can help you grow the right team but you also need to know which Tool Capabilities to invest in.