Join Robin Hunt for an in-depth discussion in this video Data best practices, part of Learning Data Analytics.
- There are a million best practices for starting projects, but I thought I would give you a few to consider. Never guess at any of your findings. Independent verification of your results can be critical. It's also very important to have a verification process so that when you release your results, everyone is in agreement with what you discovered. Outcomes with no verification are not nearly as reliable as those that are verified. Checks and balances are important when you're working with millions of records.
Define ways that you can check your work within the processes that you're working with inside your organization. Look, no one loves to have meeting after meeting when your head is still stuck in the million lines of data you just got through looking at. But they are necessary for you to understand the data, and also for your team to understand what you're working with. These should be regular, and for them to be the most productive, always have an agenda. Take a million notes on where everything came from. You don't want any guesswork on what you were thinking later.
You'll always need to be able to justify where you got your information from and the thoughts behind it. If you're questioned in one area of data, then all of your data becomes questioned. Building documentation is one of the least fun things that people will do. But explaining something over and over again to support your findings that could have been documented is way worse. It's really easy to become comfortable with your old ways and habits. They're like Bruce Willis. They die hard.
I've been a professional technical hard skills trainer for 15 years at the time of this video, and I can say without doubt that most people need training that is so specific to them that it is hard for their corporations to provide it all. Just because they do not provide it, doesn't mean that you don't need it. I'll say it again. There's a million suggestions for best practices on projects. But trust me, I know for myself and the hundreds of people that I've worked with over the years that these are some great starting points.