50 things I believe about data

A new trend is rising in the analytics and web marketing community! Lists of 50 things you believe about your industry. My inspiration was the following great lists of “50 things I believe…” from Tim Wilson and Cory Underwood:

• 50 Things I Believe About Analytics (Tim Wilson)
• 50 Things I believe about MarTech (Cory Underwood)

(Bonus: It’s also worth checking out this article from Jan Exner, which breaks down Tim Wilson’s list into more detail.)

The list below was created based on my experience working in digital analytics projects since 2008 and from numerous conversations I had while organizing the Digital analytics meetup, Thessaloniki since 2017.

The list

In no particular order:

1. Have detailed understanding of data before reporting on them
2. Documentation changes less often than data implementations
3. Use the KISS (Keep It Stupid Simple) principle when providing requirements to data developer teams
4. Encouraging a close relationship between data analysts and data developers can create bridges between these 2 different worlds
5. Bugs must be expected in data collection implementations
6. Document and be aware of bugs impacting data
7. Automated 24/7 testing is necessary
8. Manual testing can create a blueprint for automated testing
9. Using the Pareto principle (80/20) during testing can save embarrassment
10. Data is not important to most business owners. Information is.
11. Tracking everything doesn’t make sense
12. Building a data tracker is much easier, than making sure it always works right
13. Data security and privacy are not optional in 2020
14. Getting the trust and love from your data subjects will get you more data
15. Understand the power of data notifications and use it very wisely
16. Work-life of data professionals is lonely, communicate with similar-minded people will unlock value
17. Having clear business requirements ready before data collection implementation is efficient
18. Only a handful of people 🦄 can understand data collection AND appreciate data accuracy AND translate data into insights
19. Accuracy and consistency is necessary to get respect when working with data
20. Don’t expect accurate & consistent data from manual data collection
21. Be agile & flexible when working with data
22. Data collection can have a negative effect on performance in most cases
23. Be more like a thermostat (act on data), rather than a thermometer (observe data)
24. Real-time data analysis makes sense if you can have real-time actions
25. Real-time data analysis brings considerable overhead
26. Machine learning is not the solution to most data challenges
27. Good understanding of statistics and math is necessary when working with data
28. Spreadsheets might have limitations but are still a lifesaver when working with data
29. Fluency with VLOOKUP, INDIRECT, MATCH functions & pivot tables in Excel can work wonders
30. Questionable insights & inconclusive data will do more harm than good
31. A mild case of OCD is a very useful trait when working with data
32. Just like with food, data reports must provide value but they should also look nice
33. Just like with food, nice reports that don’t give any value will leave a bad taste
34. When updating the data collection process, make sure you can monitor newly collected data closely
35. Data systems can be complicated can be like chaos, even a small change can cause a butterfly effect, if you have poor understanding of them
36. Basic programming skills can bring joy to tedious tasks for every data analyst
37. Combining different data systems is usually painful but also rewarding in value
38. When combining different data systems, value formats and time zones are common traps to look out for
39. A fast system for storing and retrieving data makes a ton of difference
40. Immutable data will do more good than harm
41. If you experience a data loss you understand that data backup & data recovery plan should be top priority
42. Repetitive tasks must be automated
43. A reusable and expandable system should be available for automating tasks
44. Automated tasks can break, always have a fallback plan
45. Don’t use technical resources for the work of data analysts
46. Data can help find answers, but right questions come from skilled people
47. Understand the power of rounding numbers and use it respectfully
48. Understand the power of grouping granular values and use it respectfully
49. Good data professionals copy, bad data professionals steal
50. Combining qualitative with quantitative data can be eyeopening