• Data accuracy is critical today but still a struggle for even the most important tasks and analysis
  • To maximize its effectiveness, AI requires accurate, real-time data
  • The right data structure will enable the focus to shift from today to the future

At this moment, the most visible data in the world is the number of positive COVID-19 cases. Because of this data’s importance, you would expect those who report the numbers would all refer to the most accurate source available. However, when I reviewed the numbers from three reputable U.S. government agencies, the total cases reported varied by close to 10%.

Even with the whole world focusing on this important data, it still suffers from the same issues that most B2B companies experience with their own reporting. We have all experienced variations when analyzing key company metrics. This problem must be solved if companies want to leverage data, knowledge, and insights to create a competitive advantage. 

Most organizations recognize the impact AI and machine learning (ML) is having today and will have on future growth. However, AI/ML models are only as valuable as the accuracy, quantity, and frequency of the data running through them. These models read data and look for insights that can be used to improve business results. The more frequently accurate data is available, the faster the improvements.

For example, say you’re trying to improve close rates for sales over the next year using the analysis of sales rep activity data. Running the analysis monthly only provides 12 opportunities to improve and adapt the model. Running the analysis daily provides 365 improvement opportunities in the same period, and 30 times more opportunities for improvement. The same company running the same AI model with monthly data would take 30 years to improve at the same rate as the company analyzing the data daily. Accurate, high-frequency data is a point of advantage when competing with AI.

Here are three things you can do to increase the speed and accuracy of your organization’s data.

  • Agree on one source of truth. Many meetings have been wasted debating the accuracy of data only to find out that someone used different sources to report on the same data. Eliminate this issue by only using one source that provides accurate, real-time updates.
  • Automate updates. A report that requires human intervention is not valid. Many adjustments may be needed to enhance data; however, only those that can be automated should be used to ensure maximum speed and accuracy. If that’s not possible, a different data structure is needed.   
  • Eliminate sending copies of reports. When using important data, don’t export and send.  Instead, use a link to the data and take advantage of the many visualization tools available to display your data; many allow for embedded reports within emails.  Exported real-time data is out of date and no longer reflects reality.  If the data is still valid when sent, then the data isn’t being updated fast enough.

When working from the most accurate and timely data source, we will find more value in future projections rather than where our data is today. Stock prices are an example of a data source that has made this shift. For example, Tesla earned $24 billion in 2019 compared to Ford’s $156 billion — however, it is three times more valuable. Investors pay more for Tesla because of its projected future value. 

Finally, when we have this level of data for pandemics like COVID-19, AI models will allow us to focus on where we will be and how to maximize our resources to minimize the impact.