Welcome back to another Technical Tuesday!  We hope you all enjoyed last week’s post about Why paperless cannabis compliance should be on your roadmap for 2021 as much as we had putting it together! That topic is the foundation of why we saw a need for Evora, and it sets the stage for today’s topic, Good Data In, Good Data Out.

This may come as a bit of a shock for some, but more data is not always a good thing! More good data is a great thing, but more bad data is not. Lets start by defining good data versus bad data.

Before you can collect good data, you must define what good data looks like to your company. Good data has some common characteristics; the data should be accurate, available, relevant, reliable, and valid. Defining terms around what good data is allows your company to record, analyze, and communicate data in a more cohesive and efficient manner. This improved collaboration ability saves you time and money.

Bad data is everywhere. In 2016 IBM estimated that poor data quality costs the US economy around $3.1 Trillion annually. It is safe for us to define bad data as the opposite of good data; that is data that contains entry errors, is inaccessible, stale, and untrustworthy. Can data start good and become bad? Absolutely. Can bad data become good? Yes, but it is less likely to happen the longer a process goes unchecked and bad data collection continues to occur.

Now that we have an idea what is good data and what is bad data, I want to paint a scenario for you. Close your eyes and imagine, if you will, that you are on a paper-based quality system, and your Cultivation team consistently hands in work records and sheets with fields that are missing. The data that is recorded is illegible, and it is a collection of forms that comprise your batch record for a single batch that spans over a 9-week flowering cycle. Your Quality team must now retroactively pull this data together as accurately as possible, while also documenting a deviation to record the investigation into collecting the data. You have now got bad data on your hands, and it’s now costing another team, who may come at a higher labour cost, to chase down data that is likely to be inaccurate days to weeks later. You complete the batch with an additional 20 hours of labour attached (that likely goes un-recognized as part of the cost into this batch.. we will save Cost of Poor Quality for another Tuesday) and when you go to compare batches to find opportunities for efficiencies or innovations you are now making decisions based upon bad data. You can see where this snowball is heading.

Let’s take a look at a scenario where you’ve got a paperless solution in place and you’re working with good data. You’ve received notification that there is a transplant that is scheduled for today. You creates the activity record at the start of the activity by recording the date, everyone involved, confirm the room location, and the plant count expected. The system notifies you that Alex does not have an up-to-date training record for the active revision of the transplant SOP Document. You hand the tablet off to Alex so they can read up on the process while the rest of your team proceeds to execute the task. You scan each plant through the system as it exits the room into the system. You do a physical confirmation check of the inventory; all plants accounted for! You complete your work record or checklist and your work activity record, and sign off. Your lead is notified that there is a work activity record that has been completed and is now available for review. By making this record in real-time you’re giving the system time to first validate or sanitize the data, and pending success there, you can now leverage that good data you recorded to provide your company with relevant, reliable data to make informed decisions upon. Good data in, Good data out.

Getting the most out of your data requires you to have good data. Having good data enables your company to make informed decisions, challenge assumptions and provide insights, expose levers that drive performance metrics, highlight the effectiveness of continues improvement efforts, cultivate a culture of quality around day-to-day processes, and ultimately provide visibility to those that need it the most [Jacobs D, 2017].

References:

IBM (2016). The FOUR V’s of Big Data. IBM Infographic. https://www.ibmbigdatahub.com/sites/default/files/infographic_file/4-Vs-of-big-data.jpg

Jacobs D (April 12, 2017). 7 Reasons to Insist on Accurate, Real-Time Quality Data. LNS Research. https://blog.lnsresearch.com/7-reasons-to-insist-on-accurate-real-time-quality-data