Predictive Modeling

Leverage Data Patterns for KPI Projection and Marketing Efficiency

Forecast your organization’s growth and achieve marketing efficiency by leveraging E-Nor’s expertise in predictive modeling. We’ll make your data work for you by detecting and harnessing patterns to predict opportunities for improving marketing ROI and customer retention.

A model is only as strong as the assumptions it makes. E-Nor has the experience to guide the critical assumptions and execute any predictive model.

Predictive Modeling Methodology & System Architecture

The predictive solutions for our clients start with this core methodology and corresponding system architecture.

Forecasting

Problem

The marketing management/customer acquisition team was having difficulty setting and meeting their goals for new revenue generation. They were consistently missing weekly and monthly KPI targets, management was unable to identify root causes, create accountability, or, most importantly, answer questions like “Did our SEO efforts 2 months ago drive the increase in number of micro and macro conversions we’re seeing?”

Approach

E-Nor performed a through discovery by reviewing the client’s marketing strategy, marketing measurement model, data analytics infrastructure, and reporting workflows. We then executed an exploratory data analysis (EDA), selected and tested a machine learning forecasting model, and integrated the forecasted KPI targets into weekly and monthly reporting.

Solution

Predict future KPIs based on segmented, historical data and implement predicted targets into existing reporting workflows. Predict impact on KPIs (e.g., sales) by comparing

  1. prediction or estimate of sales if no SEO efforts/action are taken and
  2. actual KPI performance after SEO changes are made.

Results

E-Nor was able to drive a 12% increase in quarterly organic revenue by properly attributing short-term improvements in KPIs to SEO optimization efforts and reallocating resources towards related activity areas in the long-term. Also, the client’s KPI target hit rate rose to 90% by using E-Nor’s forecasted KPIs.

We shared all source code and documentation on the model/solution used and continued supporting the client by monitoring and fine tuning the model and respective KPIs on a periodic basis.

Anomaly Detection

Problem

The website/product team wanted a better way to monitor their website performance, specifically their eCommerce website checkout funnel. They were continually chasing down “false leads” for spikes and dips in conversions and were slow to identify issues, wasting precious time and resources.

Approach

E-Nor performed a through discovery by reviewing the client’s marketing strategy, marketing measurement model, data analytics infrastructure and reporting workflows. We then executed an exploratory data analysis (EDA), selected and tested an anomaly detection model using R, and implemented the visualizations in Tableau.

Solution

Create an anomaly detection dashboard in using R and Tableau.

Results

The client was able to identify three technical, website-related issues from development bugs within the first two months, saving an estimated 3% in total monthly revenue.

With predictive modeling methodologies, E-Nor can help you:

Set attainable goals and precisely plan for growth by forecasting KPIs based on historical data and accounting for historical trends, growth, seasonality, and events.

Improve marketing efficiency by forecasting expected inputs (ad spend) and results (revenue) to answer questions like “Assuming we make no changes to our marketing budget allocation, what can we expect our revenue be?” ”

Improve marketing ROI by engaging with customer segments created from machine learning techniques leveraging your data, including raw Google Analytics data in BigQuery. Focus efforts on high-value customers, reduce spend on low-value customers.

Increase profitability with improved customer retention by predicting when customers are likely to churn and identifying the preventive action to take. Retaining existing customers is far more cost effective than acquiring new ones.

Ready to harness patterns in your data to improve marketing ROI and customer retention?

Or do you already have specific use cases for predictive modeling?

We can support your effort with the right tools and best strategies for leveraging historical data. Contact us.