All case studies
Data & analytics
Bringing market data into a shared reporting system
Analysis that used to happen in spreadsheets now happens in the dashboards, on one set of KPI definitions every team reads the same way.
Go DesiQuick-commerce food brand
- The starting point
- Go Desi's competitor and category data sat in a vendor platform that produced dashboards but not decisions. KPI definitions varied between teams, the real analysis still happened in spreadsheets afterwards, and there was no route to anything more advanced.
- What we did
- Evaluated ingestion, warehouse and BI layers against explicit criteria, then built automated daily ingestion into BigQuery with a standard KPI framework and nine production dashboards.
- The outcome
- Analysis that used to happen in spreadsheets now happens in the dashboards, on one set of KPI definitions every team reads the same way.
The brief
They came to us asking about Zoho Analytics and Power BI.
We evaluated instead of building
Three layers, each assessed against explicit criteria rather than preference:
- Ingestion
- BigQuery native connectors, custom Python, and Apps Script were compared against third-party ETL tools including Airbyte, Fivetran and Meltano. We selected the native connectors with Python and Apps Script.
- Warehouse
- BigQuery against Snowflake and Databricks. We selected BigQuery.
- Business intelligence
- Metabase against Looker Studio, Power BI, Superset, and Zoho Analytics. We selected Metabase.
What we built
- Automated daily ingestion
- full BigQuery setup including project, datasets, IAM and billing
- bronze, silver and gold model layers
- a standardised KPI framework with reusable SQL models
- nine production dashboards covering SKU, competitor, pricing, ranking and category
- documentation
The engagement continued past handover with a full pipeline overhaul and resync, and periodic audits of the live data against source.
The outcome
One warehouse, one set of KPI definitions, and dashboards every team reads the same way. Analysis that used to happen in spreadsheets after the dashboards now happens in the dashboards.
“They were incredibly helpful in helping us build a dashboard to make sense of our quick commerce growth data. What I really appreciated was how easily they understood our data pipelines and helped us get the right systems and processes in place. … Their combination of technical understanding, patience and a very collaborative approach made the entire process much easier for us.”
Next case study
Managing invoicing across eleven entitiesFinance firmHave a similar challenge?
Tell us what your team needs to work through.