Web Analytics Case Study: How to Achieve up to 98% Accuracy in Forecasting Models

Thanks to Web Analytics, marketers can effectively measure, examine, and adjust changes to various online campaigns. Modern tools like Google Analytics allow to carefully analyse the data from CRM and prepare handy dashboards for the executive team and stakeholders. But the most important is Web Analytics’ potential in forecasting and bringing your business forward over again. In this material, we share the experience that helped our Ukrainian client achieve up to 98% accuracy in forecasting models. See another story of digital success with Promodo further on this page.

If you haven’t seen a number of articles related to Web Analytics, check for the below.

How to Integrate CRM Data with Google Analytics Correctly
The Three-level Analytics Approach For eCommerce: Descriptive, Predictive and Prescriptive
Top 6 eCommerce Analytics Tools For Online Stores In 2019
9 Google Analytics Reports that Drive Ecommerce Marketing Decisions and Sales
How to analyse the shopping behaviour on your website: 5 easy steps
A Beginner’s Guide to Key Performance Indicators for Online Stores

Client

Stylus is one of the largest retailers of electronics, home appliances, and accessories in Ukraine.

Challenge

Initial data

In order to calculate the workload of the call centre managers and competently build logistics, the online retailer needed to prepare a forecast for the next month, i.e. the daily number of transactions and calls received by the call centre, taking into account which parts of the country majority of the orders are coming from.

‍

Based on the forecasts, the retailer expected to correctly allocate resources, marketing budgets, and make decisions about hiring new employees.

We analysed the existing methods of data analysis and settled on three models for forecasting time series:

1. ARIMA (autoregressive integrated moving average model)
2. Additive Regression Model
3. Snaive Model

Solution

To predict the time series, we used dedicated Python scripts, machine learning, various forecasting models, and Power BI for visualization.

Why Machine Learning?

Using forecasting models based on machine learning instead of the intuitive calculation methods allows to increase accuracy in:

• Advertising budget planning
• Calculation of the call centre workload
• Warehouse workload planning and logistics

Step 1. A Test Forecast

To forecast each metric in the most accurate way, we took a data set from the CRM system and Google Analytics account for the past 2 years.

After studying the data, we made test forecasts to predict the number of sessions and transactions for the next three months and two years. This allowed tracking all the abnormal periods that increase the error in forecasting.

Trend analysis showed that a significant increase in sales for the retailer started in November 2018.

Weekly and annual seasonality charts allowed us to identify the periods of growth and fall in the number of transactions and predict the frequency of orders by month with an error of 10-15%.

To solve the retailer’s tasks, we decided to achieve more accurate indicators, since the everyday data was needed.

Step 2. Achieving More Accurate Transaction Indicators

To achieve more accuracy, we refined forecasting models and created a dictionary with all the holidays, as well as periods of increased demand (Black Friday, Christmas, etc). This helped to reveal one more pattern: promo campaigns launched close to the period of high demand impact the dramatic trend changes. At the same time, this fact had almost no impact on the weekly and annual seasonality.

Step 3. Logistics Forecasting

To distribute the warehouse workload, calculate the number of daily shipments and forecast the delivery of orders, we worked out a data set from the CRM and predicted the total number of orders, dividing them by country parts. We also created a dictionary of locations by region.

For some locations, the data was insufficient, so we combined them into a separate sample “The other parts of the country”.

Step 4. Visual Representation of the Data Acquired

To make the perception easy, we uploaded all data on the predicted transactions, calls and orders to Power Bi and presented it to the Stylus team in the form of interactive dashboards.

Results

Such high accuracy indicators were achieved thanks to:

• The presence of sufficient volume of educational selection (historical data on the metrics of interest for 2 years)
• The presence of up to 50 values of a certain metric per time unit (hour, day, month)
• A relatively high level of variance during a certain time period

What's next?

‍