Most businesses think they need better dashboards. They usually need better data architecture. The CRM has leads, contacts, accounts, deals, tasks, notes, owners, and stages. Zoho Books has customers, invoices, payments, credits, taxes, and receivables. Zoho Inventory has items, SKUs, stock, purchase orders, COGS, and adjustments. Zoho Creator has custom operational workflows. Shopify or WooCommerce has orders, refunds, discounts, shipping, and customer activity. Marketing tools have campaigns, UTMs, clicks, forms, and lead source data. Spreadsheets fill whatever gaps the systems missed. Then leadership asks: where is revenue actually coming from? That question should be easy. But for many businesses, it turns into a reporting war. Sales has one number. Finance has another. Marketing has another. Operations has another. Analytics has another. Nobody is fully confident which number is right. That is not a dashboard problem — that is a data architecture problem. Medallion architecture solves this by organizing data into progressive layers: raw data, cleaned data, and business-ready data. In simple terms: Bronze is what happened. Silver is what it means. Gold is what leadership should trust. For Zoho-centered businesses, this architecture can transform disconnected CRM, finance, ecommerce, inventory, and operational data into a true Revenue Engine — not just reports, not just dashboards, but a governed data foundation for sales, finance, operations, margin, attribution, forecasting, and eventually AI.
Quick Answer: What Is Medallion Architecture?
Medallion architecture is a layered data design pattern that organizes data into Bronze, Silver, and Gold layers. Each layer improves the quality, structure, and business usefulness of the data.
For a Zoho-centered business, the Bronze layer holds raw source data from Zoho CRM, Zoho Books, Zoho Inventory, Zoho Creator, ecommerce platforms, marketing tools, spreadsheets, APIs, and external systems. The Silver layer holds cleaned, standardized, deduplicated, joined, and validated business entities such as customers, accounts, deals, invoices, payments, products, orders, campaigns, and support tickets. The Gold layer holds analytics-ready tables and dashboards for leadership, including pipeline, paid revenue, margin, attribution, customer lifetime value, product profitability, sales performance, finance visibility, and operational bottlenecks.
The goal is not to make data more complicated. The goal is to make it trustworthy. A business cannot build advanced analytics, AI, forecasting, attribution, or executive dashboards on top of messy raw data. It needs layers. That is what medallion architecture provides.
Medallion Architecture for Zoho: Simple Comparison Table
This is how advanced Zoho analytics should be framed — not as we build dashboards, but as we build the data architecture underneath the dashboards.
| Layer | What It Contains | Zoho Example | Business Purpose |
|---|---|---|---|
| Bronze | Raw source data | Raw CRM deals, Books invoices, Inventory items, Creator records | Preserve what came from each system |
| Silver | Cleaned and standardized data | Deduped customers, standardized products, linked deals and invoices | Create reliable business entities |
| Gold | Business-ready analytics | Revenue dashboards, margin models, attribution reports | Help leadership make decisions |
| Optional AI Layer | Feature-ready data | Lead scoring features, churn risk, demand forecasting | Support predictive analytics and AI |
Why Medallion Architecture Matters for Zoho
Zoho is powerful because it can run many parts of the business. But that also means Zoho data can become complex. A business may use Zoho CRM for sales, Zoho Books for finance, Zoho Inventory for stock and COGS, Zoho Creator for custom workflows, Zoho Desk for support, Zoho Projects for delivery, Zoho Campaigns or Marketing Automation for outreach, Zoho Analytics for dashboards, and Zoho Flow or custom APIs for integrations.
If every app is treated as a separate reporting source, the business will struggle to build one trusted view of revenue. CRM may say a deal is Closed Won. Books may say the invoice was never sent. Inventory may say the product cost is missing. Creator may say onboarding started. Analytics may show revenue, but not margin. The sales team may think the customer is live, but finance may still be waiting for payment.
A Revenue Engine is not just a CRM. It is the connected system that shows how revenue moves from lead, to deal, to invoice, to payment, to delivery, to retention, to margin. Medallion architecture gives that system structure.
The Problem With Building Dashboards Directly on Raw Data
Many businesses connect Zoho CRM or Books directly to a dashboard and assume they are done. That works for simple reporting. But as soon as the business grows, raw data starts creating problems.
Raw CRM data may include duplicate contacts, duplicate accounts, old deals, missing lead sources, unclear stage names, inconsistent owner fields, manual revenue values, test records, stale pipeline, and bad close dates. Raw Books data may include duplicate customers, invoice timing differences, credits, refunds, tax differences, payment delays, voided invoices, partial payments, and unapplied payments.
Raw Inventory data may include missing SKUs, old item costs, incorrect stock adjustments, missing landed cost, product variants, bundles, supplier changes, and incomplete COGS. Raw ecommerce data may include guest checkout customers, refunds, discounts, shipping charges, failed orders, cancelled orders, payment gateway delays, and marketplace fees.
If dashboards are built directly on this raw data, leadership gets fragile reports. They may look good. But the numbers are not governed. That is why Bronze, Silver, and Gold layers matter.
Bronze Layer: Preserve the Raw Truth
The Bronze layer is the raw data layer. This layer captures data as close to the source as possible. For Zoho and revenue operations, Bronze may include raw data from Zoho CRM leads, contacts, accounts, deals, tasks and activities, Zoho Books customers, invoices, payments, and credit notes, Zoho Inventory items, purchase orders, and adjustments, Zoho Creator records, Zoho Desk tickets, Shopify and WooCommerce orders, Amazon orders, QuickBooks records, Stripe payments, Google Ads and Meta Ads campaigns, website forms, spreadsheets, and API data.
The Bronze layer should not try to be perfect. Its job is to preserve what each system said at the time it was extracted. That matters because source systems change — a deal amount may be edited, a contact may be merged, an invoice may be voided, a product name may change, a workflow may update a field, a user may overwrite lead source, or an integration may fail and resend data. If raw history is not preserved, the business loses visibility into what changed. The Bronze layer protects the audit trail.
Bronze Layer Checklist
A strong Bronze layer should capture source system name, source table or module, raw record ID, external ID, created date, modified date, sync timestamp, raw field values, deleted or inactive status if available, API extraction metadata, error or sync status, batch ID or load ID, source file name if imported, and historical snapshots when needed.
For Zoho data, that might mean preserving raw CRM deal fields before cleaning, raw Books invoice records before joining to CRM, and raw Inventory item values before standardizing SKUs. The Bronze layer answers: what did the source system say? That is the starting point.
Silver Layer: Clean and Standardize the Business
The Silver layer is where raw data becomes usable business data. This is where data scientists, analytics engineers, and serious Zoho architects earn their money. Silver is not just cleaned data — Silver is where the business model starts to appear.
In the Silver layer, raw data is cleaned, standardized, joined, deduplicated, validated, and structured into reliable entities: clean leads, contacts, accounts, customers, deals, invoices, payments, products, SKUs, orders, refunds, campaigns, support tickets, projects, owners, dates, and source fields.
Raw CRM may have contacts and accounts. Raw Books may have customers. Raw Shopify may have buyers. Raw Desk may have support contacts. The Silver layer creates a customer identity model that links them together. That is not dashboard work — that is data architecture.
What Happens in the Silver Layer?
The Silver layer answers: what does this data mean in business terms? That is the layer most businesses skip. They go straight from raw Zoho data to dashboards. That is why the dashboards break.
- Deduplication, field standardization, data type cleanup, date normalization
- Picklist standardization, null handling, invalid value handling
- Customer identity matching, account-contact relationships
- Deal-to-invoice matching, invoice-to-payment matching
- Product-to-SKU matching, order-to-customer matching, refund-to-order matching
- Lead source cleanup, campaign normalization, owner mapping
- Region mapping, currency handling, timezone handling
- Data quality flags, business rule validation
Silver Layer Example: Customer Identity
Customer identity is one of the most important Silver layer jobs. A single real customer may appear as a lead in Zoho CRM, a contact in Zoho CRM, an account in Zoho CRM, a customer in Zoho Books, a buyer in Shopify, a requester in Zoho Desk, a portal user in Zoho Creator, a payment customer in Stripe, and a row in a spreadsheet.
If these records are not linked, leadership cannot trust customer reporting. The business cannot accurately answer how much revenue a customer generated, which source created them, how many support tickets they submitted, what they bought, what the margin was, whether they renewed, whether they churned, or which account manager owns the relationship.
A Silver customer model creates a trusted customer identity using email, phone, domain, company name, CRM account ID, Books customer ID, Shopify customer ID, Stripe customer ID, external reference fields, manual match review, and confidence scoring. This is the kind of work that separates a basic Zoho consultant from a real data team.
Silver Layer Example: Deal to Invoice to Payment
One of the most important revenue models is the relationship between deal, invoice, and payment. CRM tells you what sales closed. Books tells you what was invoiced. Payment records tell you what was collected. Those are different things.
A Silver layer should model CRM deal ID, account ID, customer ID, Closed Won date, deal amount, invoice ID, invoice date, invoice amount, payment ID, payment date, payment amount, payment status, credit notes, refunds, and outstanding balance.
This allows leadership to separate pipeline revenue, Closed Won revenue, invoiced revenue, paid revenue, outstanding receivables, refunded revenue, and net revenue. Without this model, revenue reporting becomes vague. A Revenue Engine must know the difference between a deal that was won, an invoice that was sent, and money that was actually collected.
Silver Layer Example: Product and Margin Model
For ecommerce or inventory businesses, the Silver layer must also clean product and cost data. This may include SKU standardization, product name standardization, variant mapping, category mapping, supplier mapping, purchase cost, landed cost, COGS, refunds, discounts, shipping, marketplace fees, channel data, and inventory adjustments.
This allows leadership to answer which products are profitable, which products sell but have weak margin, which channels generate revenue but destroy profit, which discounts reduce margin too much, which suppliers create cost risk, which SKUs have high return rates, and which inventory categories deserve more investment. This is where data architecture becomes business strategy. Revenue is not enough — the business needs margin intelligence.
Gold Layer: Build Trusted Business Outputs
The Gold layer is the analytics-ready layer. This is the layer leadership sees. Gold data should be clean, modeled, governed, and ready for decisions.
Gold tables and dashboards may include executive revenue dashboard, sales pipeline dashboard, source-to-revenue dashboard, finance and receivables dashboard, product margin dashboard, ecommerce performance dashboard, customer lifetime value dashboard, sales rep performance dashboard, marketing attribution dashboard, inventory and COGS dashboard, support and customer health dashboard, revenue leakage dashboard, and forecasting dashboard.
The Gold layer should not expose messy raw logic to users. It should answer business questions clearly: how much pipeline do we have, how much revenue did we close, how much did we invoice, how much did we collect, what is still unpaid, which sources create profitable customers, which products have the best margin, which customers are at risk, where is revenue getting stuck, and what should leadership do next. Gold is where analytics becomes decision-making.
Gold Layer Checklist
Gold should be simple for users because the complexity has already been handled underneath. That is the entire point of medallion architecture.
- Official metric definitions and governed dashboards
- Role-specific reporting and executive KPIs
- Source-of-truth documentation and refresh schedules
- Data quality indicators, clear filters, and drill-down paths
- Trusted date logic, customer logic, revenue logic, margin logic, and attribution logic
The Revenue Engine Medallion Model
Medallion architecture turns Zoho from a collection of apps into a Revenue Engine data system. A Revenue Engine data model connects Lead Source → Lead → Contact → Account → Deal → Proposal → Invoice → Payment → Product → COGS → Margin → Support → Retention. That chain is difficult to model directly from raw systems, but it becomes manageable with layered architecture.
Bronze Revenue Data
Raw CRM, Books, Inventory, Creator, ecommerce, marketing, and support data preserved as close to source as possible.
Silver Revenue Model
Clean customer identities, standardized deals, linked invoices, normalized products, matched payments, source-of-truth rules, deduped records, and validated relationships.
Gold Revenue Intelligence
Dashboards showing pipeline, revenue, margin, paid revenue, customer lifetime value, attribution, retention, bottlenecks, and executive KPIs.
Why This Makes CloudStream Different
Most Zoho consultants think in apps: CRM setup, Books setup, Creator app, Analytics dashboard, integration, workflow, automation. That work matters. But high-level data teams think in systems: data sources, raw ingestion, transformation layers, entity models, customer identity, metric definitions, data quality rules, lineage, source of truth, data marts, executive decision layers, and AI-ready features.
A basic consultant connects Zoho apps. A serious data team designs the architecture that makes those apps trustworthy. CloudStream should own that position: the Zoho data team that understands CRM, finance, ecommerce, inventory, analytics, and medallion architecture well enough to build the revenue intelligence layer underneath growth. That is how you move from implementation vendor to strategic data partner.
Medallion Architecture for Zoho Analytics
Not every business needs BigQuery immediately. Some businesses can apply medallion thinking inside Zoho Analytics.
Bronze inside Zoho Analytics means imported tables from Zoho CRM, Books, Inventory, Creator, Desk, Shopify, WooCommerce, spreadsheets, and APIs. Silver inside Zoho Analytics means query tables that clean, join, dedupe, standardize, and model business entities. Gold inside Zoho Analytics means dashboards and reporting tables for leadership, sales, finance, inventory, ecommerce, and operations.
This can be a strong approach for small and mid-sized businesses. The architecture pattern matters even if the tool is lighter. The discipline is the same: do not build executive dashboards directly on messy source data. Create modeled layers.
Medallion Architecture With BigQuery
For more complex businesses, Zoho Analytics may not be enough as the modeling layer. That is where BigQuery can become important.
A more advanced architecture may include a Bronze layer with raw data from Zoho CRM, Books, Inventory, Creator, ecommerce platforms, ad platforms, payment processors, APIs, and spreadsheets; a Silver layer with cleaned business entities built with SQL, Python, dbt, notebooks, or transformation pipelines; a Gold layer with business-ready marts for sales, finance, marketing, ecommerce, inventory, support, and executive reporting; and a BI layer with Zoho Analytics, Power BI, Looker Studio, Tableau, or another dashboard tool.
This gives the business more scalability and control. It also creates a stronger foundation for AI, forecasting, attribution, and advanced analytics.
When Zoho Analytics Is Enough
In this case, CloudStream can still apply medallion thinking without overengineering the architecture. The goal is not to force enterprise tools on an SMB. The goal is to build the right data architecture for the business's stage.
- Most business data is already in Zoho and data volume is manageable
- Reporting needs are mostly operational and dashboards are needed quickly
- Transformations are not extremely complex and the business does not need advanced data science yet
- Leadership needs trusted KPIs more than machine learning
- The team wants faster time to value
When You Need BigQuery
This is where CloudStream can move beyond dashboards into real data engineering and analytics architecture.
- Data comes from many systems outside Zoho and historical snapshots are important
- Data volume is high and transformations are complex
- Advanced SQL or Python is needed and multiple BI tools need the same governed data
- The business wants machine learning, AI, or complex attribution modeling
- Customer lifetime value modeling is needed and ecommerce and inventory margin models are advanced
- Data quality checks need to be systematic and the business needs a long-term warehouse or lakehouse
Medallion Architecture and AI
AI does not fix bad data. AI amplifies whatever data foundation you give it. If your CRM has duplicates, missing lead sources, unclear stage definitions, unreliable invoice links, and inconsistent product data, AI will not magically create trustworthy predictions — it will generate confident confusion.
Medallion architecture helps prepare data for AI by creating cleaner layers. Bronze preserves raw history. Silver creates clean entities and reliable features. Gold creates business-ready metrics.
A lead scoring model may need lead source, industry, company size, form intent, time to first response, activity count, deal stage movement, sales owner, historical close outcome, deal value, and customer segment. A churn model may need payment history, support tickets, renewal date, usage data, customer segment, product adoption, account manager activity, prior complaints, and invoice delays. A margin forecast may need SKU, channel, discounts, COGS, landed cost, refund rate, shipping cost, supplier changes, and inventory movement. All of that depends on clean modeled data. Medallion architecture makes AI possible because it creates the trusted data layers AI needs.
Data Quality Rules by Layer
Bronze Data Quality
Bronze should preserve raw data but still track source system, load status, sync timestamp, record ID, extraction errors, schema changes, and missing required metadata. Bronze should not over-clean data — it should preserve the raw truth.
Silver Data Quality
Silver is where data becomes trustworthy.
- Deduplication rules and standardized IDs
- Valid email formats, clean dates, standardized picklists
- Customer matching logic and deal-to-invoice relationships
- Product-to-SKU relationships and payment-to-invoice relationships
- Null handling and business rule checks
Gold Data Quality
Gold is where data becomes decision-ready.
- Metric definitions and source-of-truth rules
- Approved KPIs and reconciled values
- Dashboard-ready fields and executive reporting logic
- Clear filters, consistent date ranges, role-specific outputs
Medallion Architecture Example: Zoho CRM to Paid Revenue
Bronze
Raw CRM deals, Books invoices, Books payments, CRM accounts, CRM contacts, and activity records are loaded as-is.
Silver
The data team cleans account names, dedupes contacts, standardizes deal stages, links deals to customers, links invoices to CRM accounts, matches payments to invoices, and creates a clean revenue lifecycle model.
Gold
Leadership sees dashboards for pipeline revenue, Closed Won revenue, invoiced revenue, paid revenue, Closed Won but not invoiced, invoiced but unpaid, revenue by source, revenue by owner, sales cycle length, time from Closed Won to invoice, and time from invoice to payment. Now leadership can see where revenue is stuck. That is the Revenue Engine in action.
Medallion Architecture Example: Ecommerce Margin
Bronze
Raw Shopify orders, WooCommerce orders, Zoho Inventory items, Zoho Books invoices, refunds, payments, discounts, shipping, and COGS records are loaded.
Silver
The data team standardizes SKUs, maps variants, dedupes customers, links orders to invoices, models refunds, maps landed costs, applies channel labels, and creates a clean order-line margin model.
Gold
Leadership sees dashboards for revenue by SKU, gross margin by product, contribution margin by channel, refund rate by product, discount impact, shipping impact, COGS by category, customer lifetime value, channel profitability, and inventory risk. This is what separates basic ecommerce reporting from real margin intelligence.
Medallion Architecture Example: Lead Source to Lifetime Value
Bronze
Raw lead source data, UTM fields, CRM records, deals, invoices, payments, support tickets, and renewal data are preserved.
Silver
The data team models customer identity, normalizes campaign names, preserves original source, tracks latest source, links deals to invoices, connects support data, and creates a clean customer revenue table.
Gold
Leadership sees revenue by original source, paid revenue by campaign, margin by lead source, retention by acquisition channel, customer lifetime value, sales cycle by source, support burden by source, and churn risk by channel. This is where marketing attribution becomes serious — not just leads, but customers, revenue, margin, and retention.
The 30-Day Medallion Architecture Audit
Before building a warehouse or advanced analytics system, audit the current data environment.
Week 1: Source System Inventory
Document every system that creates revenue-related data. Review Zoho CRM, Books, Inventory, Creator, Desk, Analytics, Shopify, WooCommerce, Amazon, QuickBooks, Stripe, payment processors, Google Ads, Meta Ads, website analytics, spreadsheets, internal databases, and custom apps. The goal is to identify all Bronze data sources.
Week 2: Business Entity Mapping
Define the core business entities: leads, contacts, accounts, customers, deals, invoices, payments, products, SKUs, orders, refunds, campaigns, support tickets, projects, renewals, and churn events. The goal is to design the Silver layer.
Week 3: Metric and Source-of-Truth Review
Define official metrics: pipeline revenue, Closed Won revenue, invoiced revenue, paid revenue, net revenue, gross margin, contribution margin, customer lifetime value, lead source, close rate, sales cycle, retention, churn, support cost, and product profitability. The goal is to identify what belongs in the Gold layer.
Week 4: Architecture Roadmap
Decide whether the business needs Zoho Analytics only, Zoho Analytics with query-table modeling, BigQuery warehouse, hybrid warehouse plus Zoho Analytics, phased medallion architecture roadmap, data quality and governance plan, or AI-readiness roadmap. The deliverable should not just be a dashboard list — it should be a data architecture plan.
Common Medallion Architecture Mistakes
The first mistake is overengineering too early — Not every business needs advanced warehouse modeling immediately. The second mistake is underengineering too long — if the business has complex data and still relies on raw dashboard joins, reporting will eventually break.
The third mistake is skipping the Silver layer. Most reporting failures happen because raw data was pushed directly into executive dashboards. The fourth mistake is treating Gold dashboards as the source of truth — Gold is an output layer; the modeled data underneath is the source of trust.
The fifth mistake is ignoring customer identity. The sixth mistake is mixing revenue definitions — pipeline, Closed Won, invoiced, paid, net, and recognized revenue are not the same. The seventh mistake is ignoring margin. The eighth mistake is building AI before cleaning the data foundation. The ninth mistake is failing to document metrics.
What a Healthy Medallion Architecture Looks Like
That is the standard — not just data movement, not just dashboards, but a governed revenue intelligence architecture.
- Raw data is preserved and clean business entities exist
- Customer identity is modeled; deals link to invoices; invoices link to payments
- Products link to SKUs and COGS; refunds and credits are modeled
- Source and campaign data are preserved; revenue definitions are separated
- Margin definitions are clear; dashboards use Gold tables, not messy raw exports
- Leadership trusts the numbers; data quality issues are visible
- The architecture can support analytics, AI, and forecasting
Final Thoughts
Medallion architecture is not just an enterprise data buzzword. It is a practical way to make business data trustworthy. For Zoho-centered companies, it can be the difference between disconnected dashboards and real revenue intelligence.
Bronze preserves the raw truth. Silver cleans and models the business. Gold gives leadership decision-ready metrics. That layered approach is how a business turns CRM, Books, Inventory, Creator, ecommerce, marketing, and support data into a trusted Revenue Engine.
Most companies do not need more charts. They need a data foundation that tells the truth. That is what medallion architecture helps create.
Need Help Building a Medallion Architecture for Zoho Data?
CloudStream Software Solutions helps businesses design advanced Zoho data architectures that connect CRM, Books, Inventory, Creator, ecommerce, finance, marketing, and analytics into one trusted revenue intelligence layer.
If your dashboards do not match, your CRM and finance numbers conflict, your ecommerce revenue does not reconcile, your margins are unclear, or leadership still depends on spreadsheets, the problem is not just reporting — the problem is data architecture.
A Zoho Systems Review can help identify which systems should feed the Bronze layer, which raw data should be preserved, how to model customers, deals, invoices, payments, products, and campaigns in the Silver layer, which Gold dashboards leadership should trust, whether Zoho Analytics is enough, whether BigQuery should become the warehouse or lakehouse, where source-of-truth definitions are unclear, where revenue is being double-counted, where margin and COGS reporting are incomplete, whether the business is ready for AI or predictive analytics, and what architecture should be built first.
The goal is not just to connect data. The goal is to build a Revenue Engine with a trusted data foundation — one that shows where revenue comes from, what it costs, where it gets stuck, and what to do next.
Frequently Asked Questions
What is medallion architecture?
Medallion architecture is a layered data design pattern that organizes data into Bronze, Silver, and Gold layers. Bronze stores raw data, Silver stores cleaned and modeled data, and Gold stores business-ready data for reporting and analytics.
How does medallion architecture apply to Zoho?
For Zoho, Bronze can include raw data from CRM, Books, Inventory, Creator, Desk, and ecommerce systems. Silver can clean and model customers, deals, invoices, payments, products, and campaigns. Gold can power executive dashboards and revenue intelligence.
What is the Bronze layer?
The Bronze layer stores raw source data as close to the original system as possible. For Zoho, this may include raw CRM deals, Books invoices, Inventory items, Creator records, ecommerce orders, and marketing data.
What is the Silver layer?
The Silver layer cleans, standardizes, deduplicates, validates, and joins raw data into reliable business entities. This is where customer identity, deal-to-invoice relationships, product mapping, and payment matching are modeled.
What is the Gold layer?
The Gold layer contains business-ready tables and dashboards. It supports leadership reporting, revenue analytics, margin analysis, attribution, customer lifetime value, forecasting, and executive decision-making.
Do small businesses need medallion architecture?
Small businesses may not need a full enterprise warehouse, but they can still benefit from medallion thinking. Even inside Zoho Analytics, separating raw tables, modeled query tables, and leadership dashboards can improve reporting trust.
Do I need a warehouse for medallion architecture?
Not always. Medallion architecture can be applied in BigQuery, Snowflake, or even lightweight Zoho Analytics models depending on business size and complexity.
Is Zoho Analytics enough for medallion architecture?
Zoho Analytics can be enough for many SMBs if the data volume and transformation needs are manageable. More complex businesses may need BigQuery underneath Zoho Analytics as the warehouse or lakehouse layer.
How does medallion architecture help AI?
AI needs clean, reliable, modeled data. Medallion architecture helps by preserving raw history, creating clean business entities, and producing trusted metrics and features that can support lead scoring, churn prediction, forecasting, and advanced analytics.
What is the first step to build medallion architecture for Zoho?
The first step is a data architecture audit. Identify source systems, raw data, business entities, metric definitions, source-of-truth rules, reporting gaps, and whether the business needs Zoho Analytics, BigQuery, or a phased architecture.