Has an Executive Sponsor Got Your Back?
- Without executive sponsorship it’s hard to overcome organizational inertia
- Senior executive should within a key stakeholder group for WA (e.g. ecommerce, marketing, etc)
- They should have enough power and influence
- Depending on the WA maturity the sponsor executive may vary (tactical to strategic)
- Executive sponsor responsibilities
- Align WA program with corporate strategy
- Protecting the WA team from other initiatives/corporate politics
- Solving problems like budget constraints
- Promoting the success of the WA program
- Effective executive sponsor should be committed and involved
Online Accountability: Are You Data-Driven or Merely Data-Informed?
- Without accountability your organization is just data-informed
- Establish clear goals
- Regularly talk about the performances
- Give feedback and maybe rewards – can be adversarial
- Accounting should start at the top. Lead by example
- Expand beyond web KPIs
Soft vs Hard Bounces: A Closer Look at Bounce Rate
- Hard bounce rate = Bounce Rate on new visitors
- Soft bounce rate = Bounce Rate on returning visitors
Switching to a Data-Driven Culture
- You have appeal to the rational and emotional sides
- Sometimes resistance is lack of clarity
- Laziness can be exhaustion
- Tactics:
- Look at what works well instead of was is not working well – easier to promote
- Provide actions that have to be taken to change
- What does this mean for the near future?
- Surprise people – testing stuff is a great tool
- Try to achieve lots of small goals instead of one big one
- See failures in execution as learning not as failing
- Provide a data-driven environment
- Build habits – repeat, repeat, repeat
- Provide workshops for homogeneous groups
Five Times to Test: 4 — When you spot an opportunity in your analytics
- Often hypotheses drive testing
- But you can generate hypotheses with analytics
- They took the ~50 top-selling products and plotter conversion rate and avg. selling price
- Look for outliers
- Positive outliers: try to promote them more prominently
- Negative outliers: Check at least the page – is it broken? No content?
Is Your Data-Driven Organization Heading into a Lake?
- Data should inform and shape not dictate or control
- It’s like science: intuition helps to understand and inspire, data helps to check and reject
- It helps you question your assumptions
- Data-informed: nice to know this information
- Data-driven: acting on the information
Are You Using Web Analytics To Power & Improve Your Testing?
- Don’t test randomly, test with hypotheses
- Benefits of WA:
- Helps you to understand your testing efforts in context
- Helps you to prioritize testing areas
- Helps you to improve your decision maing
- Provides insights that help you to make even better tests
- Things you should do:
- Analyze your conversion funnels
- Start higher up the funnel (note: in contrast to previous article)
- Check your top landing pages that have high bounce rates
- Check heat maps for your testing pages – what is the customer intent?
- Set alerts for new highly visited pages
- Improve your test plans with analytics insights
- WAs and testing people should work together
- Testing is a long-term strategy
- Does every test answer a clear business question?
- Do you know before the test what you do, depending on the results, afterwards?
- Negative lift is even good lift – you learned something!
- Do you think testing is valuable or risky?
- Do you have to hit the big wins? – This can interfere with your learning ambitions
Why we do what we do: Garbage in and Garbage Out — Congruence Bias
- “the tendency to test hypotheses exclusively through direct testing, in contrast to tests of possible alternative hypotheses”
- Example: Hypotheses: Button 1 opens the door, not Button 2; Test: Just press Button 1 and check if it opens the door
- In web testing: Test picture against no picture, or CTA against no CTA
- It’s easy to get big results but not great answers
Great summary/overview: Digital Governance: Best Practices from the Trenches
