Ep. 648 Your Dashboards and AI Tools Are Only as Good as Your Data

Client Technology Specialist Brenda Colby interviews Media Communications and Client Services Partner Alyssa Burley and discusses best practices when implementing data-dependent tools.

Show Notes: ⁠⁠⁠⁠Subscribe to Rancho Mesa's Newsletter⁠⁠⁠⁠

Host: ⁠⁠⁠⁠Brenda Colby

Guest: ⁠⁠⁠⁠Alyssa Burley

Editor: Megan Lockhart

Music: "Home" by JHS Pedals, “Breaking News Intro” by nem0production

© Copyright 2026. Rancho Mesa Insurance Services, Inc. All rights reserved.

Transcript

Brenda Colby: You’re listening to Rancho Mesa’s StudioOne™ podcast, where each week we break down complex insurance and safety topics to help your business thrive.  I’m your host, Brenda Colby, and I’m joined by Alyssa Burley, Media Communications and Client Services Manager with Rancho Mesa, and today we’re going to talk about best practices when implementing data-dependent tools.

Alyssa, welcome to the show.

Alyssa Burley: Thank you. I’m rarely on this side of the interview. But, I’m happy to be here.

BC: That’s true! And we’re happy you’re here.

Okay, so you recently wrote an article where you emphasize that data is only as valuable as the accuracy, completeness, and consistency of what's entered. Can you explain why data quality is becoming such a critical issue as companies adopt more AI and analytic tools?

AB: Yeah so, I just spent several days at a conference with other business leaders and many of the discussions were focused around applied AI and data analytics and the Keynote Speaker at this event was Sol Rashidi, who’s been described as the World’s First Chief AI Officer. And, she described AI as essentially being just mathematics. Which makes sense to me. If you have a formula and one number, and one piece of data is inaccurate or missing, the answer you get will be wrong. I think her comments were more meant to describe how large language models actually work, but I think it applies here, too.

Now every day, employees at all levels enter information into mobile forms, safety observations, incident reports, and training records. When fields are consistently used as intended, companies can generate meaningful reports, review reliable dashboards, and leverage AI-powered insights to guide future safety training and business decisions.

However, when fields are used inconsistently or not as they were intended by the platform developer, the value of that data is significantly reduced both today and in the future. For example, if you’re using a field labeled Employee ID, which the platform expects to be unique to each user, but your organization chooses to use it for the employee’s t-shirt size, you’ve significantly reduced your ability to utilize the tools and functions that use that specific data field. Now, you might think that a particular field isn’t currently tied to any logic or functionality, but that doesn’t mean it won’t be in the future. And, I think that’s where organizations who are just now implementing AI and data analytics are finding that they may have unintentionally restricted their ability to utilize this new technology to its fullest. They may have to go back and clean up their data in order to get the full benefits of AI and data analytics.

BC: Yeah that’s a really great point! So, in your article, you shared an example of organizations misusing the Project Name field in our proprietary safety platform, SafetyOne. So can you explain how entering information into the wrong fields can impact reporting, dashboards, and ultimately business decision-making.

AB: So, much like the Employee ID field I just mentioned, one of the most common examples we see in SafetyOne of not using a field as intended is the Project Name field on the Administrator website. This field is designed to identify a specific job, crew, facility, location, or program associated with a group of records. So, depending on your organization’s industry, it might make sense for the Project Name field to track records by each construction job, a service team or crew, a physical campus, or community program. It just depends on your industry. And, I’d recommend having a discussion with your client technology team member here at Rancho Mesa prior to adding all of your project names to the platform. Since every organization is different, they can give you recommendations on how to organize your data using the Project Name field.

Alternatively, when the Project Name field is incorrectly used for individual supervisors or employees, reporting capabilities become essentially ineffective.

The example I wrote about in the article explains management may want to analyze incidents, inspections, or training activities by Project to identify trends and allocate resources. If the Project field is used differently than intended, dashboards may not be able to display the aggregated data for jobs, crews, facilities, locations or programs accurately. Dashboards and AI tools rely on consistent data to identify patterns and generate meaningful recommendations.

So, while you may think you’re streamlining the documentation or organization of data by using the Project Name field for individuals, in reality, you’re just limiting your ability to utilize the AI and data analytics feature for reporting, dashboards and analysis.

BC: Valid points for sure. As we know, many organizations focus on collecting required information, but you stress the importance of completing optional fields and maintaining consistent naming conventions. So, how do those small details improve the quality of AI-generated insights and long-term safety analysis?

AB: Great question so for anyone who has ever had the task of manually updating data in a platform in order to get a report to be accurate, you know the importance of entering as much information as you can at the beginning and making sure that everything is spelled the same way. It’s much easier to enter data correctly the first time than to go back to fix it later. Now, there are tools designed to clean up data for you, but not every organization is going to have it at this point.

So, completing all available fields within a platform can save your organization significant time and frustration down the road. Although many fields in SafetyOne and other platforms are required, optional fields often contain valuable information that may become critical later. What seems unimportant today could become a key data point when leadership wants to analyze trends, evaluate performance, or leverage AI-powered insights.

One of the greatest challenges organizations face is realizing they need data that has never been collected. If information is not captured at the time an event occurs or the record is entered, historical insights often cannot be recreated after the fact.

As data analytics and AI become increasingly integrated into business operations, the completeness of your data may ultimately determine the quality of the decisions you can make from it. 

And, consistency is just as important as completeness. For example, if an employee signs and types their name on a toolbox talk record as “John Smith,” then the next week signs it "Johnny S,” and the following week "J. Smith," the platform may interpret those as three separate people. So, establishing standardized naming conventions throughout the platform helps ensure records are grouped correctly and that reports, dashboards, and AI tools can accurately analyze the data. As I mentioned earlier, there are third-party tools designed to clean up data if there are inconsistencies, but it’s a best practice to establish naming conventions early on and stick to them. 

As technology continues to evolve, data quality will only become more important. Clean, standardized data is the foundation that supports accurate reporting, effective dashboards, and meaningful AI-generated insights. By taking a few extra moments to ensure information is entered correctly into the proper fields, organizations can significantly improve the accuracy and usefulness of the data they collect. 

Unlike a human reviewer who can often infer meaning from incomplete information, AI tools depend on structured, standardized data to identify trends and generate insights. The more organized and complete the data set, the more accurate and useful the resulting recommendations will be.

BC: Yeah all of that is really important information to consider. I have personally seen how employees have signed reports with different naming conventions across various toolbox talks or other forms, and later trying to generate a report analysis, such as safety meeting participants, makes it nearly impossible to provide accurate data.

Well Alyssa, thank you so much for providing all of this insightful content for our listeners. With that being said, if listeners have questions about Rancho Mesa’s AI tools and data analytics, how can they get in touch with the team?

AB: Clients can reach out to our client technology team either through our safetyoneapp@ranchomesa.com email or by reaching out to their assigned client technology coordinator or specialist.

BC: Great. Well Alyssa, thanks for joining me in StudioOne.

AB: It was a pleasure.

BC: Thank you for tuning in to our latest episode produced by StudioOne. If you enjoyed what you heard, please share this episode and subscribe. For more insights like this, visit us at ranchomesa.com and subscribe to our weekly newsletter.