Analytics Horror Stories - BrightonSEO 2016
•
6 likes•6,871 views
The document discusses common issues and mistakes seen with Google Analytics implementations and use. It provides examples of scenarios where analytics were not being used properly or data was missing due to technical problems. The key recommendations are to work with analytics experts on implementations, have a data plan for any site changes, test all data, don't delete or duplicate analytics setups, and understand the limitations and error margins of analytics data.
Report
Share
Analytics Horror Stories - BrightonSEO 2016
- 1. Analytics Horror Stories Anna Lewis Polka Dot Data #BrightonSEO September 2016
- 2. Dotty About Data SEO Role - Discovered Google Analytics PPC and SEO role Analytics specialist within agency Lots of blogging, training and speaking gigs Analytics specialist role in ecommerce company Maternity Leave Started my own analytics consultancy business 2008 2016
- 4. THE DATA MAKES ME CRY! © Anna Lewis
- 5. #1
- 6. “I’m suddenly getting a lot more sessions”
- 7. “My banner is really successful, but I can’t see which traffic source generated this”
- 8. “We launched an app and it’s so successful we have almost double the traffic already”
- 9. Did you use campaign tracking codes on internal links?
- 10. Auto tagging Use the correct tracking method for each activity Event tracking Enhanced Ecommerce UTM campaigns EXTERNAL INTERNAL Internal search
- 11. #2
- 12. “Can you tell us how to improve the checkout? …”
- 13. “…it’s all in this one fantastic iFrame”
- 14. “…it’s hosted by a third party but I’m sure there’s data”
- 15. “…it only has a few 302 redirects in the journey, it should be tracking fine”
- 16. “…it sends users to a subdomain that we don’t have any control over but the devs say everything works fine”
- 17. (>_<)
- 19. Match journey to reports Get tracking code implementation experts involved in all new developments on site Plan Plan plan Test full journey Talk to experts Educate key players in the process Get more involved
- 20. #3
- 21. “We have this amazing set up, it should tell us exactly what we need to do, but there’s just too many numbers, so we don’t really use it.”
- 22. “Of course we have Google Analytics. Geoff knows how to access it. He checks it every month”
- 23. You spent £££’s installing a tool but don’t have anyone who uses it well?
- 24. Tools Analyst Remember Avinash Kaushik’s classic rule: Spend 90% on people 10% on tools Agencies UsersAnalyst Analyst Devs Devs Experts Experts
- 25. #4
- 26. “I didn’t need to look at the data, I know that this version will perform better”
- 27. “I analysed the data and it told me to do X, but I think doing the opposite to X would be better”
- 28. Did you do all that set up, analysis or testing just for fun?
- 29. If you don’t trust the data, test it Instil trust across the business Test the data before using it Agree success and fail criteria
- 30. #5
- 31. “We set up a new site but I think it’s performing worse than the old one”
- 32. “This account shows a decline but the other one shows improvements”
- 33. “Comparing to last year, the results are different in a way we didn’t expect them to be”
- 34. Did you forget to put in a plan to have consistent data?
- 35. Don’t forget the data! Get tracking code implementation experts involved in all new developments on site Document everything Plan plan plan Know the best practices Test all data
- 36. #EXTRA
- 37. Bounce Rate My new site’s bounce rate is so good, it’s less than 10% You may have added a new set of tracking code without removing old code Two codes = two pageviews = no bounces!
- 38. Other pages I can see (other) in my Pages report, that’s not one of my pages! You probably have too many unique page URLs and have run out of rows Exclude parameters in the admin settings to remove unique strings and group URLs together
- 39. Differences Why am I getting different results for the same thing? Check the sampling level, is it different on the two reports? When adding segments or date comparisons, GA doesn’t always have the accurate processed data to hand.
- 40. #BONUS
- 41. “I just hit delete. I wasn’t expecting everything to disappear.”
- 43. Restrict admin access to trusted people DO NOT DELETE Find out what the delete button deletes before hitting it
- 44. If you don’t trust it, test it! Write and stick to guidelines Top Takeaways DON’T delete or duplicate Trust the data Have a data plan for every website change Talk to experts Understand the error margins & limitations
- 45. Analytics Training Insights from your website anna@polkadotdata.com polkadotdata.com Anna Lewis Set up & Auditing Get the most from your website!