A comparison analysis is one of the most important tools to make an informed decision. Whether you’re using it as a business owner, as part of a team, or making a personal choice, comparison analyses simplify the task of picking between closely matched alternatives.
Choices between alternatives can get complicated in a hurry if you don’t apply a thought framework to your analysis. In this article, you’ll learn how to build a comparison framework to guide your analysis. We’ll be using an example from the supply chain industry and build a framework for comparing vendors.
All analyses rely on data, and in this case, temperature notification systems provide us with the raw information we need to arrive at insights.
Before analyzing anything, you have to define the boundaries of your analysis and the frame of reference you will use. Typically, this is a benchmark that can guide you. In the supply chain industry, you could use a research report that classifies typical vendor response times and delivery success rates to benchmark the vendors you’ll be comparing.
Given the complexity and scenario-specific nature of this type of data, it’s likely that such a study doesn’t exist. In these cases, developing your own benchmark using temperature notification data is the way forward. For instance, these datasets can highlight which of your logistics vendors deliver goods in optimal conditions and can be used to create a rating system. Vendors that routinely deliver goods in conditions that are borderline are less reliable than ones that deliver them optimally.
Seasonality biases are also revealed through these data. Some vendors might experience greater demand at certain times of the year, which lowers the average condition of the goods they deliver. Others may specialize in local regions where the natural temperatures are more hospitable to your product condition requirements, making the vendor’s vigilance less of an issue at certain times of year.
An analyst might want to consider creating season-specific benchmarks to account for such scenarios. When collecting this data, make sure your sample size is large and diverse enough to avoid biases.
Once you’ve defined a benchmark, you’ll need to define which variables you’ll use to rank the alternatives. Temperature notification data provides a readymade dataset, along with other condition-related data. Using a temperature logging system, you can feed your database with readings on packages throughout their journeys. Then, mining this database, you can create a simple rating system by measuring the proximity of the product’s average in-transit temperature to its required thresholds, and multiply that by the frequency of each occurrence to yield a number measuring performance.
The point of this first step is to establish a strong foundation so that you always know which way your analysis is headed and how you can rectify issues when you encounter them. This way, you can stay on track and avoid conducting a SWOT analysis instead.
Once you’ve defined the base of your analysis, it’s time to define your alternatives. This task is tougher than it looks since you need to ensure you’re comparing apples to apples. Comparative analyses breakdown when comparing alternatives with access to different resources.
For instance, comparing a vendor that has a fleet of 30 vehicles to another that has a 100 strong fleet is not a fair process. The large vendor might perform better with big loads in all weather conditions, while the former might work better in specific conditions with small loads. Examining variance in temperature-related data will reveal some trends you can dig deeper into.
The larger vendor might have steady performance throughout the year while the smaller vendor might be nearing temperature thresholds more often when they’re heavily in demand. Fleet size could be a possible issue. Route planning might also offer an explanation.
Can you tolerate some temperature deviations from required thresholds, or does that result in a total loss of your shipment and render that vendor less trustworthy? As an analyst who can take the specifics of your situation into account, data from temperature notification systems provides you a good starting point from which you can dig deeper.
The aim here is to categorize your alternatives appropriately. You can do this by measuring the attributes connected to each vendor. In our example, the size of the fleet, number of drivers employed, the presence of dispatchers on staff, and the diversity of the vehicles used (trucks, cold chain vehicles, airplanes, etc) are good variables to use.
You can compare choices across different categories but make sure you explicitly note their differences. Also, make sure you restrict bias towards the less advantaged categories when drawing conclusions. Those categories will have their own appeal, so make sure you note them in your report.
Conducting an analysis can become a tedious task if you don’t prepare beforehand. Note the steps you’ll take to execute the analysis. For instance, note that you’ll begin by defining your categories of vendors and the names of all vendors appropriately. You’ll then examine their delivery success rates while keeping an eye on the variables used to measure success.
Temperature notification systems provide a range of effective variables you can use. For instance, a vendor might have a high success rate, but does the data reveal that they occasionally overcool their products? This will increase costs. Do their drivers often need assistance? Excessive assistance might indicate a problem with vendor infrastructure and a lack of adequate cooling mechanisms. These variables also allow analysts to assume a predictive stance.
For instance, a rising rate of driver assistance despite a constant delivery success rate indicates a brewing problem. A vendor’s rating can account for this with analysts weighting ratings according to trends revealed by temperature-related data and the alerts these systems offer.
In addition, alternative variables such as the change in insurance premiums the vendor pays from one year to the next also offer a good measure of success. Insurance rates indicate how often the vendor has been found liable for damage, with higher rates potentially indicating unsafe delivery practices. Defining the analysis process allows you to look for non-traditional variables such as these.
Often, you’ll run into discrepancies between data sources. Defining your error handling process is critical. Not only does the definition give you a structure to follow, but it also establishes a benchmark for future studies. It also forces you to think about handling missing data and drawing conclusions in such cases.
A good comparison analysis is a constant process. While initial analysis will support one conclusion, it’s important to return to your variables and assumptions after a while to re-examine them. Variables change over time, and your steps will change with them.
For instance, some vendors will grow bigger and change categories. They might upgrade their delivery infrastructure. In such cases, you’ll have to conduct your analysis from scratch, reexamining the conditions and temperatures of the delivered goods. Either way, constant iteration keeps your analysis fresh and relevant. A good comparative analysis is a great asset for an organization, and the person conducting it is a valuable resource.
Follow this framework, and you’ll ensure your analysis is always relevant and addresses the right issues.