Automated analytics workflow optimizes order management

Automated analytics workflow optimizes order management

Automated analytics workflow optimizes order management

A multinational company’s regional order fulfillment team was bogged down by data silos across its systems. The team spent more than 10 hours every week manually analyzing orders, on hand and incoming stock, and inventory from channel partners. Without an efficient process to gather all key metrics in one dashboard, nor the time or know-how to encode logic to automate more than half the cases that were simple, the team’s analysts urgently needed a solution.

We built an automated analytics workflow to extract, transform and broadcast the daily insights through a set of self-service Tableau dashboards. The solution enables the order fulfillment team to make better informed order release decisions with lower effort than previously, while more than 50% of decisions flow with automated logic.

The value proposition:

The solution reduced the time spent analyzing and processing orders by 80%, nearly 500 hours in annual time savings. The consolidated dashboards improved order management, paying off through stronger partnerships, greater visibility into the upstream supply chain, and a 5% reduction in unfulfilled orders, representing more than $75K annually.

Our solution roadmap consisted of:

  • Leading workshops to understand the business process and its shortcomings, clarify requirements and define business logic.
  • Mapping business requirements to technical requirements.
  • Automating data extraction from internal systems.
  • Transforming, enriching and preparing powerful data models.
  • Designing a set of interactive, self-service dashboards.
  • Scheduling periodic checkpoints to offer support and plan enhancements.

Technical and analytical environment:

  • Python
  • Task Scheduler
  • Tableau
Self-service analytics powers intelligent real estate investments

Self-service analytics powers intelligent real estate investments

Self-service analytics powers intelligent real estate investments

A leading Canadian commercial real estate developer wanted more accessible insights, and a deeper understanding of its market, competitive landscape, and consumer profiles, to enable quicker and more data-driven decision-making companywide.
Our team built a cloud-based and open-source one-stop shop dashboard application, intended to reduce the time spent on ad-hoc querying, improve access to data across departments, deliver deeper insights by way of advanced analytics and new metrics, and empower business users with true self-service analytics.

The value proposition:

More than 2 full-time employees in annual time savings. Better decision-making in investment planning, leasing negotiations and marketing initiatives yield the potential for tens of millions of dollars in revenue and investment gains.

Milestones and learnings:

A key milestone came in navigating through a variety of the company’s existing data sources, and tying the raw data to business outcomes in the form of user stories. As in many analytics projects, data discovery and cleanup were the backbone of our solution.
The creative process of determining which internal data sources to focus on, and which external data sources would help enrich the insights, was crucial. We initiated a clear line of communication with the client early on, iterating and building on an established feedback loop.

Our solution roadmap consisted of:

• Interviewing stakeholders to identify the current state, pain points, and target state.
• Assessing available data and exploring new data to enrich analytics.
• Cleaning and aggregating the data for analytics.
• Evaluating a variety of BI tools and analytical techniques.
• Building an open-source, cloud-based dashboard app.
• Embedding advanced analytics and creating new metrics and visualizations.
• Leading client workshops, receiving feedback and agilely actioning those learnings.
• Testing and documenting to ensure quality and encourage user adoption.

Technical and analytical environment:

• R, JavaScript, CSS and HTML programming
• RShiny cloud applications
• Fuzzy clustering
• Text analytics
• Social media engagement analytics

Forecasting with external data untaps new paths to market

Forecasting with external data untaps new paths to market

Forecasting with external data untaps new paths to market

A market leader in the Canadian Coffee and Beverage industry required a substantial increase in visibility on their 2nd largest business channel to drive strategic commercial decisions. The client had previously outsourced an analysis of this channel to a major consulting firm for 6 figures, which quickly became outdated and obsolete. Our task was to create a modern data-driven solution to augment strategic insights for years to come.
To solve this issue, we developed an automated, cloud-based solution to source and process external data, integrating it with internal data to model channel projections, identify high value segments, and highlight untapped opportunities.

The value proposition:

With brand new KPIs, the strategic insights for this channel have enhanced decision-making. An estimated annual revenue gain of 1% equates to $1.2M. Additionally, 70% of a full-time employee’s workload is removed as ad-hoc queries are no longer required.

Challenge: starting from scratch

As with any analytical solution, data is the foundation. Our client had very little internal data applicable to the business objective. We needed to thoroughly explore available internal and external data sources to develop new data models and enable success drivers for the business case.

Our solution roadmap consisted of:

  • Working with stakeholders to identify desired metrics to drive decisions.
  • Evaluating tools and visualizations to maximize project ROI.
  • Assessing the data points required to achieve the target state outcomes.
  • Sourcing industry data from the web, and reorganizing the company’s internal data.
  • Defining and implementing a production analytics environment.
  • Developing forecasting models to highlight new business opportunities.
  • Sharing insights through self-service dashboards and email alerts.

Key to success: sharing knowledge

This kind of solution was brand new and needed to be understood thoroughly by the end users to trust the insights. We worked closely with the client from the start to ensure the vision was clear and that every metric was designed in line with strategic objectives

Technical and analytical environment:

• Microsoft Azure
• Tableau
• Web scraping
• Statistical forecasting

AI on guest reviews enhances vacation rental host insights

AI on guest reviews enhances vacation rental host insights

AI on guest reviews enhances vacation rental host insights

Accommodation sharing platforms provide hosts with pricing features and general guidance on how to improve their chances of booking out their listing. However, the platforms do not offer hosts value-added data products based on guest reviews to help them understand guest profiles, improve listings, and ultimately generate more business.
Our project, focused on mining guest reviews for sentiment and similarity, was built to show how text data can be used to enhance user experience. Our Airbnb host dashboard outputs insights on common themes mentioned in the most positive and negative reviews, key selling attributes, key points of concern, and guest profiles.

The value proposition:

The product helps hosts manage and optimize listings at scale, especially those with many listings and guests. It provides hosts with an actionable overview of who their guests are and what they are saying about their listings. This solution is an important lesson for any business to consider the full, untapped potential of its data. This exercise requires creativity, boldness and expertise in implementing cutting-edge analytics.

Our solution roadmap consisted of:

  • Understanding the Airbnb business model, its data points, and the shortcomings in how data is used to enhance user experience.
  • Building the value proposition for our proposed solution
  • Preparing the data for analytics.
  • Creating a pipeline to feed advanced text analytics into intuitive dashboards.
  • Designing the host dashboard to be brief and actionable.

Technical and analytical environment:

  • Python
  • Advanced text analytics