AI is really good at parsing through a lot of data and identifying themes you might want to pay attention to. To test that out, I threw an export of the last two month’s of deficiencies from a demo Mechanical account into ChatGPT and asked it to help me prepare for the upcoming quarter by identifying any trends in the data.

As you can see, it had a little trouble analyzing the data, so I tried to help it out by giving it a link to our Support article on deficiency creation. But honestly, I think it would be better to give it a little more instruction on the deficiency data you consider to be the most important. That can be unique to your business processes, so try this out, but always remember that if you don’t get a quality answer the first time, give the AI a little help by explaining more about what you want, and also what you don’t want, and try again.

The best part about the relatively long analysis included below? It took about 30 seconds to generate and it cost me zero dollars. So tweak, iterate, and re-try, and when you finally DO get what you want, make sure to save it as a prompt template somewhere so that you and your team can use again next quarter!

The Prompt:

I work as the service manager for a commercial mechanical company. I need to plan out the next quarter by predicting repairs for customers. Analyze the attached data that shows all deficiencies reported in the past two months, and look for trends so you can generate a plan to address predictive issues. You can learn more about what deficiencies mean to me in this Help Center article: https://support.servicetrade.com/hc/en-us/articles/16143205860371-Creating-and-Managing-Deficiencies

The Output:

The data contains several columns that may be relevant for analyzing trends and predicting future repairs for your customers. Here are some key columns in the dataset:

• id: A unique identifier for each deficiency report.
• ref_number: Reference number for the deficiency.
• location: The name of the location where the deficiency was reported.
• location_street, location_city, location_state, location_postalcode: Address details of the location.
• location_phone: Phone number of the location.
• company: The company name related to the deficiency.
• asset: The specific asset reported as deficient.
• description: Detailed description of the deficiency.
• date_reported: The date when the deficiency was reported.
• reporter, reporting_company: Information about who reported the deficiency and which company they represent.
• office, tags, comments, regions: Additional metadata that might help in categorizing and analyzing the deficiencies.

To predict repairs for the next quarter based on these deficiencies, we can analyze trends such as the most common types of deficiencies reported, the frequency of reports over time, and if certain locations or types of assets are more prone to issues. We can also look into the specific issues reported in the description column to identify any recurring problems.

Let’s start by summarizing the data to identify the most common deficiencies, the frequency of reports over the past two months, and any patterns related to locations, companies, or assets. ​​

❗️Error analyzing
It seems there was an error in my attempt to analyze the data, specifically related to a ‘date_reported’ column that does not exist in the dataset as I initially thought. Let’s correct this by first reviewing the available columns to identify the correct one for analyzing the timeline of reported deficiencies. Then, we’ll proceed with the analysis based on the correct column names and data structure. Let’s quickly list all column names to ensure we use the correct one for time-based analysis and other trend identifications. ​​

After reviewing the column names in the dataset, it appears that the correct column for analyzing the timeline of reported deficiencies might be last_reported_on rather than a non-existent date_reported. I will use this column to perform a time-based analysis, along with analyzing the most common deficiencies by asset and common locations where deficiencies have been reported. Let’s proceed with this adjusted approach. ​​

Based on the analysis of the data provided, here are some key findings and trends identified from the past two months:

Deficiencies Timeline:

• January 2024: 120 deficiencies reported
• February 2024: 189 deficiencies reported
• March 2024: 8 deficiencies reported (up to the date of the data provided)

Most Common Deficiencies by Asset:
The most commonly reported deficiency involves the “Location – Building”, with 160 instances. This suggests a broad range of issues not specific to a single type of equipment but potentially indicative of widespread or recurring building-related issues.
Specific assets like chillers, rooftop units, and boilers have also been frequently reported for deficiencies, but far less frequently compared to location-based reports. Notably, certain chillers and rooftop units have been reported 2 to 3 times.

Most Common Locations with Deficiencies:
Denton County Public Facilities and Dave & Buster’s locations (among others) have seen a higher frequency of reports, with up to 14 and 10 reports respectively. This could indicate specific sites that may require closer attention or proactive maintenance efforts.

Predictive Repair Planning:
Given these insights, your repair planning for the next quarter might focus on:

Increasing inspections or preventive maintenance at sites with high deficiency reports to catch and address issues early.
Paying special attention to building-related concerns and ensuring that routine checks are in place to prevent common deficiencies.

Considering a targeted approach for assets that have been repeatedly reported for deficiencies, such as specific chillers and rooftop units, to understand if there are underlying issues that need more comprehensive solutions.

By focusing on these areas, you can potentially reduce the incidence of future deficiencies and improve overall customer satisfaction through proactive service management.