Skip to main content

Blog Archives

Older posts
Newer posts

ADA helped a leading digital first “House Of Brands” gain competitive market intelligence by building a modern data platform on Databricks

Case Study

ADA helped a leading digital first “House Of Brands” gain competitive market intelligence by building a modern data platform on Databricks

The Results
Eliminated manual data collection requirements
Eliminated data discrepancy issues
Enabled data scientists to make dashboards and ML models
The Challenge

The customer is a leading digital-first “House of Brands” venture focusing on fostering the growth in Direct-To-Customer (D2C) in the region.

The company was previously collecting data from their managed brands manually which resulted in data in accuracies and delays in getting business insights for decision making activities. They also had no visibility on data from their managed brands ecommerce platforms and marketplaces.

The Solution

ADA ran several proof-of-concepts(POC) to help evaluate the most suitable cloud platform and technologies. After consideration of price and performance, we helped the customer build an end-to-end data pipeline and data warehouse on Databricks and AWS.

The data pipelines allowed us to gather insights from many different data sources for sales, returns, inventories, and finance.

ADA also built bespoke API connectors for various ecommerce platforms to enrich collected data. We also implemented web scrapping to extract market intelligence on competitors pricing, ranking, and patterns to help drive the customers’ product strategy.

Content
Results
Challenge
Solution
Get in touch

Let’s turn your data and AI into measurable outcomes!

Book a Demo

ADA elevates Boost’s brand growth with business & consumer insights

ADA elevates Boost’s brand growth with business & consumer insights

Results
0 %
in new users
0 %
in re-engaging lapsed users
0 x
increase in weekly transactions
The Challenge

As the COVID-19 pandemic struck, eWallet adoption skyrocketed in Malaysia. With businesses shuttered and the government advocating for digital transactions, Boost found itself in a fiercely competitive landscape. To stay relevant, Boost embarked on a multifaceted mission:

  • Engage and retain their existing user base
  • Acquire new users
  • Drive Gross Transaction Value (GTV)
  • Facilitate the shift of MSMEs online
Solution
1 Derive insights from XACT, ADA’s proprietary data DMP

We managed to identify specific B40 COVID-19 crisis personas and worked towards meeting their needs. These personas were: Adaptive Shoppers, The Bored Homebody, The Health Nut, and The WFH Professional.

2 Using GWI (Audience Insight Tools)

We found that endorsements by celebrities are 1.5 times more likely to persuade our users to use our brands. This insight was central to our idea of getting KOLs to promote the brand and engage with our users.

Results
Challenge
Solution
Get in touch

Let’s turn your data and AI into measurable outcomes

Book a Demo

Enhancing branch digital transformation for an Indonesian commercial bank with location analytics

Enhancing branch digital transformation for an Indonesian commercial bank with location analytics

Results

  • 100% of underperforming branches identified and prioritized for relocation to higher-potential trade areas

  • New high-affluence opportunity zones mapped, uncovering untapped locations with strong concentrations of wealth managers, businesses, and salaried professionals lacking banking presence

  • Location Planner insights embedded into 5-year transformation roadmap, guiding long-term branch strategy and investment planning

The Challenge

As Indonesia’s banking landscape grew more competitive, the client needed to find their competitive edge to continue growing in times of critical uncertainty.

The client turned to ADA to leverage our data analytics expertise to evaluate several key locations to determine the presence and distribution of Syariah banking competitors. In addition, our heatmap enables our client to optimise the blueprint of their branches for the next 5 years.

The Solution
1 Derive insights from XACT, ADA’s proprietary DMP

We were able to draw fundamental information to assess location opportunities for customer profiling: Audience Movement, Area Density, Competitive Intensity.

2 Identify low / non-penetrated hotspot location

We identified unpenetrated “hotspots” by using the customer and Point of Interest (POI) filters to validate the branch model and identify a new location with minimal competition.

3 Identify underserved hotspot locations

Besides identifying the underserved areas for a new branch concept, we also assessed competitive intensity and uncovered underserved hotspots for the right branch mix.

4 Determine whether to retain or relocate branches

Using heatmap, we determined whether its several existing branches can be retained, or if a new branch or relocation would be necessary.

Execution
5 Regional Location Planner

Split into 8 different regions with multiple filters, e.g: home vs. footfall density, segment, age, affluence, and more were applied to verify the customer behaviour of a particular location.

6 Commercial Dashboard

The Commercial Dashboard enables the client to choose multiple variables on province and industry type, with colour-concentration intensity.

7 Customer Zip Dashboard

Displays the density of the client’s customer in particular areas based on customer segments such as company/merchants/premier customers with activity status (dormant vs. active).

Results
Challenge
Solution
Execution
Get in touch

Let’s turn your data and AI into measurable outcomes

Book a Demo

We helped a leading ecommerce retailer trial a centralised data operations team to improve efficiencies and reduce incident rates

We helped a leading ecommerce retailer trial a centralised data operations team to improve efficiencies and reduce incident rates

Results
0 %
The average Time-To-Response for frequent data issues was reduced from 20 days to 5 days
0 %
​Average weekly incidents was reduced from 30 to 10
Customer’s engineering teams could focus on improving the source data integrations
The Challenge

A leading Indian e-commerce platform in fashion, beauty, and lifestyle operated with fragmented data ownership across departments such as marketing, finance, and insights. Each team requested different datasets, commercial metrics, clickstream data, storefront analytics, creating a constant stream of ad hoc data pipeline requests.

Engineering teams were overwhelmed with building, maintaining, and troubleshooting these pipelines, leading to long turnaround times for resolving data issues and a high volume of recurring incidents. The lack of centralized triage and standardized processes created inefficiencies, duplicated effort, and inconsistent data quality across teams.

The client needed a more structured data operations model to reduce incident volume, speed up response times, and free engineering resources to focus on improving core data integrations and platform reliability.

The Solution

ADA helped define and trial a centralized data operations team that would triage and manage all departmental data issues, perform initial assessment and route analysis findings to the relevant engineering teams.​

ADA also improved the data process and minimized incident rates by building self-healing data pipelines and reporting on data accuracy. ADA also put in place data pipelines and business logic for derived and aggregated datasets orchestrated using Azkaban; a batch workflow job scheduler that allows for tracking and monitoring of the data workflows.​

Results
Challenge
Solution
Get in touch

Let’s turn your data and AI into measurable outcomes

Book a Demo

ADA helped a leading online pharmacy optimize data storage costs and performance by migrating from AWS Redshift to Apache Hive

ADA helped a leading online pharmacy optimize data storage costs and performance by migrating from AWS Redshift to Apache Hive

Results
0 %
Reduction in storage costs​
0 %
Performance reduction of query run time from ~50 minutes to ~15 minutes​
0 months
Reduction of query run time from ~50 minutes to ~15 minutes​
The Challenge

A leading online pharmacy and medical platform serving 1,000+ cities and 22,000+ pin codes was experiencing rapid data growth from streaming transactions, customer activity, and operational reporting. Their existing AWS Redshift data warehouse began facing scalability limits, with storage costs rising sharply and query performance slowing down.

Long-running queries—often taking nearly an hour—delayed reporting cycles and impacted decision-making across operations, supply chain, and customer analytics teams. At the same time, the business needed to maintain uninterrupted reporting while modernising its data infrastructure, making migration and optimisation complex and high-risk.

The client required a scalable, cost-efficient data warehouse solution that could reduce storage expenses, improve query performance, and support continued growth without disrupting critical analytics workflows.

The Solution

The customer is the region’s top online pharmacy and medical care platforms with doorstep delivery service available in 1000+ cities and towns across 22000+ pin codes. They provide a wide range of over-the-counter products and medical equipment across a broad budget spectrum.​

The customer was facing growing storage costs and scalability bottlenecks on their existing AWS Redshift data warehouse as their volume of streaming and transactional data was increasing. They needed a more cost-effective data and customizable data warehouse for their data growth.​

ADA helped to evaluate different technologies and developed a phase-by-phase plan to migrate the existing data to a Hadoop based Apache Hive data warehouse. We ensured that existing data reporting operations remain unaffected during the migration process. ​

ADA established best practices, performed necessary data validation, and deployed optimized codes that would power multiple business-wide dashboards and reports.

Results
Challenge
Solution
Get in touch

Let’s turn your data and AI into measurable outcomes

Book a Demo

We helped a leading ecommerce retailer automate their financial month end closing and reduce their turn around time​

We helped a leading ecommerce retailer automate their financial month end closing and reduce their turn around time​

Results
0 %
Reduced month end closing effort from 7-10 man days to only 4 hours.
Reduced the turn-around-time for resolving data discrepancies.​
The Challenge

A leading Indian e-commerce platform in fashion, beauty, and lifestyle was struggling to close its financial books efficiently at month end. The finance team relied on manual calculations and reconciliations, taking 7–10 man days to complete reporting and increasing the risk of errors.

Data discrepancies across multiple systems covering taxes, commissions, stock transfers, and goods received, caused further delays, slowing down dependent processes such as accounts receivable, accounts payable, and inventory reconciliation.

The client needed a faster, more reliable way to automate calculations, improve data accuracy, and streamline month-end closing without disrupting ongoing finance operations.

The Solution

An eminent Indian e-commerce entity specializing in fashion, beauty, and lifestyle products, faced significant challenges to expedite their financial month end closing. ​

The customers finance team were manually calculating and reporting their month end closure taking between 7-to-10-man days to complete. Compounded with the high turnaround time to resolve data discrepancies and provide clean data, this further impacted other dependent finance activities such as accounts receivable, account payable, and inventory.​​

ADA helped to reduce the turn around time by building automated data pipelines that calculated the various figures required for their financial reporting (e.g., tax, commissions, stock transfer notes, good received notes).​

ADA built data completeness checks and split the month end closure activity into separate phases (Phase 1: start of month till 24th of the month, Phase 2: 24th till month end) allowing us to quickly identify and resolve any data discrepancies early on.​

Results
Challenge
Solution
Get in touch

Let’s turn your data and AI into measurable outcomes

Book a Demo

Improving treatment effectiveness for an Indian hospital chain with patient-based forecasting model

Improving treatment effectiveness for an Indian hospital chain with patient-based forecasting model

The Results
  • Patient education was enhanced, enabling individuals to have a deeper understanding of their ailments and treatment options
  • Support and interventions were tailored to each patient
  • Seamless communication and feedback loop
0 %
treatment adherence increased
The Challenge

Headquartered in India, our client is an established chain of eye specialty hospital that operates within a network of 103 centres in India and 15 centres internationally.

However, the chain has been experiencing a significant drop in patient adherence at various stages of treatment, leading to suboptimal treatment outcomes. The client’s objective is to enhance patient adherence throughout the treatment process, including post-procedures, to achieve improved treatment outcomes.

To address this challenge, the client aims to develop a robust patient-targeting module that can effectively remind patients of their appointments with ophthalmologists and other critical checkpoints, thereby promoting and facilitating improved adherence to the treatment plan.

The Solution
1 Patient segmentation

We started by mapping out the end-to-end patient journey to identify key stakeholders and areas of patient dropouts during that journey. Patient segments were created based on the patient’s demographic details, medical history, and segment types. For example, a healthy patient with a generic eye ailment or a patient with a history of complex eye ailments.

2 Defining the line of treatment

We worked with the hospital SMEs to define the line of treatment for various ailments, such as refractive errors, glaucoma, cataract, retinal detachment, macular degeneration, and amblyopia. These involved a range of interventions, including corrective lenses, medication, surgery, or vision therapy, depending on the specific condition. A complete treatment requires long-term adherence.

 

3 Patient targeting

Patient-level targeting was done based on the segment to which the patient belongs, and the ailment assessed by the physician through diagnosis. This was achieved through a series of reminder protocols.

Results
Challenge
Solution
Get in touch

Let’s turn your data and AI into measurable outcomes

Book a Demo

ADA helped a global retailer create a modern unified data platform on Databricks to reduce operational costs and streamline data workflows

Case Study

ADA helped a global retailer create a modern unified data platform on Databricks to reduce operational costs and streamline data workflows

The Results
0 %
Costs reduction from streamlined workflows and optimized code
0 Dashboards
Medallion architecture on Databricks enabled faster dashboard refresh rates.
Data validation measures ensured data accuracy and integrity post migration.
The Challenge

The customer is a leading global sporting goods retailer with a footprint in 56 countries and selling products from over 20 brands.

The Solution

The customer built their existing data platform on disparate cloud services (Amazon Redshift for data storage, Python/SQL for data processing, and Jenkins for workflow automation). The existing platform introduced many integration challenges that led to increased costs and reduced operational efficiencies.

ADA helped design, implement, and migrate their existing data to a new unified data platform on Databricks. We successfully resolved many technical challenges including handling and migrating diverse data sources, overcome limited support for converting Python code to PySpark using innovative solutions, and ensuring data alignment between the old platform and the new Databricks platform.

The project was executed seamlessly, preserving data integrity and security while avoiding disruptions to ongoing operations.

The Approach

Our goal is to establish a gold layer within the Databricks system, utilizing data from the silver layer for business intelligence reporting. Our approach to achieve this involves:

  • Data Source Analysis: Understand current data sources and associated codes for a clear migration foundation.
  • Code Transformation with Optimization and Quality Checks: Convert Python/SQL to Pyspark/Spark SQL, adding data quality checks for reliable data integrity and optimize it for space and speed.
  • Data Validation and Comparison: Verify Databricks data against Redshift data, ensuring seamless alignment.
  • Data Orchestration: Schedule jobs using Databricks workflows, implementing monitoring and alerts for a reliable workflow.
  • Documentation and Knowledge Transfer: Document the migration process for reference and conduct knowledge transfer sessions for enhanced team adoption.
Content
Results
Challenge
Solution
Approach
Get in touch

Let’s turn your data and AI into measurable outcomes

Book a Demo

Gaining 6x ROAS for an Indonesian dairy-based nutrition company with CPAS campaign

Case Study

Gaining 6x ROAS for an Indonesian dairy-based nutrition company with CPAS campaign

The Results

Successfully established itself on Lazada in Indonesia

0 M
people reached
0 x
growth in ROAS
The Challenge

Our client, a prominent global dairy-based nutrition specialist, known for providing essential milk products for children and families, had been a market leader in Indonesia for generations. However, with increasing competition in the category, they faced a challenge in converting their market awareness into sales. Despite multiple engagement and sales campaigns, their efforts were falling short.

The Strategy

To address this issue, we shifted our focus towards channel selection and product platform optimization to drive tangible sales growth with trackable and optimizable conversions. This involved a comprehensive business expansion strategy on the eCommerce platform, particularly targeting major sales events like Singles’ Day and 12.12. We employed broad audience targeting, emphasizing Indonesian women in the 20 to 35 age range, particularly those in the motherhood phase who value quality products. Additionally, retargeting was implemented for users who had previously shown interest in products but had not completed a purchase. Automatic placements and campaign budget optimization were utilized to ensure cost-effective ad delivery across various platforms.

Awards
  • Bronze in Best Use of Gamification category at Loyalty & Engagement Awards 2020
  • Bronze in Best Use of Technology category at Loyalty & Engagement awards 2020
  • Bronze in Best Use of Mobile ‚Äì Food & Beverage category at Mob-Ex 2020
  • Bronze in Best Use of Mobile ‚Äì Gaming category at Mob-Ex 2020
  • Silver in Innovation category at MMA SMARTIES APAC 2020
  • Silver in Innovation category at MMA SMARTIES Indonesia 2020
  • Bronze in Most Engaging Mobile Creative category at MMA SMARTIES Indonesia 2020
Content
Results
Challenge
Strategy
Awards
Get in touch

Let’s turn your data and AI into measurable outcomes!

Book a Demo

Acquiring deep understanding of consumer behavior for a major Japanese real estate developer with customer profiling

Case Study

Acquiring deep understanding of consumer behavior for a major Japanese real estate developer with customer profiling

The Results
  • The client obtained a comprehensive overview of the top 10% (high affluence) of the Hanoi province population to be the potential target audience.
  • In addition, key consumer profile, persona of high-end mall visitors, key district (region) of interest, and the right tenant mix were identified to support the client’s strategic development goals
The Challenge

A prominent Japanese real estate company aimed to expand their roots in Vietnam to build a luxury mall in Hanoi. To drive the decision-making process in developing the new shopping complex, the client turned to ADA to understand data-driven insights and deep dive into their most important questions:

  1. Who and where are the affluent shoppers located?
  2. When are they shopping?
  3. What are the key interests and behaviour patterns of shoppers in Hanoi, Vietnam?
The Solution
1 Derive insights from XACT, ADA’s proprietary DMP

We presented clear and comprehensive insights which allow the client to understand the demographic profile and persona of the consumers of interest.

2 Comparison analysis

Five competitor malls were selected based on consumer footfall and existing retailers and online sentiments, to feed into their own strategic decision-making processes.

3 Gather key information

We discovered the distribution, density, the cross-visitation, and travel patterns of visitors to competing malls. Then we built a customer profile and persona of highly affluent visitors to develop the client’s brand and high priority segments.

4 Formulate a precise plan

We planned accordingly for the client’s tenancy mix and merchant partnership in all their existing and future malls.

The Execution
5 Consumer Profiler Dashboard

We identified the density of visitors at different malls during specific times of day. Home and work locations were identified to visualise only visitors of interest (region of origin), with demographic filters such as age group, gender, and affluence level applied.

6 The Insights: Target Market & Competitors

The insights were broken down into two sections. The first section covers an overview of each of the affluent tiers through demographic and psychographic comparison. The second section involves diving into the five selected competitor malls.

7 Spatial and temporal changes of audience mobility and digital footprints

ADA tracked and analysed pre-COVID vs. post-COVID movement restriction patterns to forecast future consumer behaviour in selected points of interest. With a deep understanding of competitors’ value propositions, they can make strategic decisions to garner reach and maintain their competitive edge.

Content
Results
Challenge
Solution
Execution
Get in touch

Let’s turn your data and AI into measurable outcomes!

Book a Demo
Older posts
Newer posts