Cost Modeling

Background

Cost modeling is key to business planning and budgeting. Cost modeling reveals inefficiencies, enabling cost-cutting measures. Further, multi-year budgets and forecasts, in particular, will be far more accurate with better cost modeling. Moreover, cost modeling makes it possible to make accurate cost comparisons with other industry members.

Indeed, major corporations like Walmart and McDonald’s owe their success largely to effective cost modeling, enabling them to squeeze supplier margins to the minimum, and satisfy customers with impossibly low pricing.

To achieve the best pricing strategy, accurate cost modeling is key. As Wharton writes, changes in pricing can boost profitability far more than a reduction in fixed or variable costs, or an increase in volume. Pricing is key, yet many organizations don’t consider the best possible pricing strategy. One common strategy is cost-based pricing, but doing this well requires an intricate understanding of the costs of your customers or products, which can be highly variable and complex.

Cost modeling depends on more than just price, such as shipping expenses, quality costs, and inventory-carrying costs. Moving beyond products, cost modeling only gets more complex. For an insurance provider, for instance, cost modeling for a person’s medical care includes factors like their age, gender, BMI, where they live, and more.

These numerous factors are interrelated in non-linear ways, making it difficult to accurately predict with traditional statistics. Machine learning is the answer, which lets you comb through historical data, and accurately predict any costs. With Akkio’s no-code machine learning, you can easily build and deploy cost models, and improve your bottom line.

Cost Modeling With AI

To get started, sign up to Akkio for free. As with any machine learning task, step one is getting historical data and picking a column to predict.

Historical Data

In this case, our historical data is a Kaggle Dataset called “Medical Cost Personal Datasets.” This file has 1,338 rows of patients and seven columns, or six features after excluding the target column. That final column is called “charges,” which is simply the medical cost. This is what we’re trying to predict.

We’ll upload that Kaggle Dataset by simply clicking “Upload Dataset,” then “Table,” as our CSV dataset is in a tabular format.

Building the Model

Now, we can click on the second step in the AI flow, which is “Predict.” Scroll all the way down under “predict fields,” and you can select the column to predict, named “charges.”

Then just hit “Create Predictive Model,” and you’re done! It takes as little as 10 seconds to train a model from scratch. You can also select a longer training time—from 1 to 5 minutes—for potentially more accurate models. Keep in mind that longer training times aren’t always better, and may lead to overfitting.

Also note that you don’t pay for model training time, unlike with many typical automated machine learning tools, so feel free to build as many models as you’d like.

Analyzing the Model

After a model is created, you get a simple overview, which highlights the top fields, model prediction quality, sample predictions, and more.

We ended up with a very high-quality model, with the prediction usually within around 23% of the true charge. The RMSE, or Root Mean Squared Error, was $6,619, while the MAE, or Mean Average Error, was $3,761. With charges ranging from around $1,000 to $64,000, these errors are relatively slim.

By selecting 1-minute training time instead of 10 seconds, we can get the error to within 20% and the MAE under $3,000. 

There’s also the option to “See Model Report,” which lets us easily collaborate with others and share these model details with anyone, even if they don’t have an Akkio account.

In short, we’ve discovered that it’s very easy to build a highly accurate cost model.

Deploying the Model

Now that we’ve built a model, it’s time to deploy it in the real-world.

With Akkio, it becomes trivial to deploy complex machine learning models, through a wide range of deployment options, including Salesforce, Google Sheets, Snowflake, API, and a web app, with many more methods coming soon.

Deployment via Zapier is an easy way to serve predictions in a number of settings. Zapier is a no-code automation tool that connects tools with “Zaps,” and you could use a Zap to send new product (or patient) data to the Akkio model automatically.

Zapier allows you to link up thousands of different tools, so no matter where your data is coming from, it’s easy to connect it to Akkio to get a prediction, even if you don’t have any technical expertise.

If you’re looking for more technical power, you can also use Akkio’s API, which is formatted as a Curl command, allowing you to send a GET request that contains your flow key, API key, and input data, and get a prediction back.

Cost Modeling on New Patient Data

In insurance, it’s standard practice that different people get different pricing. If you’re a 20-year-old urban male, with a BMW as your daily driver, you’ll pay drastically more for car insurance than a middle-aged rural female driving for pleasure. Some insurers go even further, taking driving behavior into account.

However, not all insurers, let alone all companies, have the technical and financial resources to implement intelligent cost modeling. 

Indeed, there are around 8 million businesses in the United States—according to the most conservative figures—and just around 6,300 data scientists in the US. We hear all about AI at industry-leading companies, but the reality is that the vast majority of normal companies have no AI talent.

With Akkio, any firm can effortlessly build and deploy cost modeling systems. Let’s walk through a simple example in Zapier. First, we create a “trigger” to pull in new patient data.

For this example, we’re using a Google Sheets file as the source of new patient data. Note that any database source could be used.

We’ll then use the cost model we’ve built to predict the cost of a new patient. To activate Akkio in Zapier, visit this link, and then select “Make Prediction in Akkio.” Make sure that the data you input exactly matches up with what was used to build the model.

Given our prediction, we can add it back to our patient data with another step. Simply add a step to “Update Spreadsheet Row in Google Sheets,” and then we’ll add the prediction to a final column named “charges.” 

Summary

We’ve explored how cost modeling is key to business planning and budgeting. It’s used by leading corporations like Walmart and McDonald’s to drive down prices, and even by insurers to create more intelligent pricing strategies. However, these efforts are too costly and time-intensive for millions of regular firms.

Using Akkio’s no-code AI, you can effortlessly build and deploy cost models.