Growth, AI, and Data Strategy

You walk into the budget review meeting with your CAC calculations ready. The CFO pulls up their spreadsheet. The numbers don't match. Neither of you can explain why with confidence. The conversation shifts from "How should we allocate next quarter's budget?" to "Wait, which number is actually right?"‍This isn't a spreadsheet problem. It's a data architecture problem masquerading as a finance disagreement.

Why Your CAC Calculations Keep Breaking

When customer data lives across Salesforce, your ad platforms, product analytics, subscription management systems, HubSpot, and billing tools, every stakeholder builds their own version of truth. Marketing calculates CAC using opportunity close dates from Salesforce. Finance uses actual cash collection dates from Stripe. Product attributes signups to referral programs that Marketing can't see. Sales claims credit for accounts that started as self-service trials.

The result: three different CAC numbers in the same meeting. Leadership debates whose methodology is "more correct" instead of making strategic decisions about where to invest.

Here's what breaks in fragmented customer data environments:

• Customer identity isn't unified. The same customer appears as separate entities across systems—one record in your CRM, another in your product database, a third in your advertising platforms. When someone converts from a free trial to paid, or when a single user represents a multi-seat enterprise deal, attribution fractures across systems that don't talk to each other.
• Timing mismatches create phantom discrepancies. Marketing counts a customer acquired when they enter the CRM. Finance counts them when payment clears. Product counts them when they activate. Each department is technically correct within their system—but the company has three conflicting acquisition counts and no way to reconcile them without manual spreadsheet archaeology.
• Attribution logic lives in BI tools, not your data layer. When critical business logic exists only in Tableau calculations or analyst notebooks, every report becomes a potential source of inconsistency. One analyst defines "marketing-attributed revenue" differently than another. The definitions drift over time. Nobody can trace back to understand why the numbers changed.

What "One Source of Truth" Actually Means in Practice

Building trustworthy unit economics isn't about picking the "right" CAC definition and forcing everyone to use it. It's about creating data architecture where customer identity, attribution logic, and financial calculations live in a shared, tested, documented foundation that all systems pull from.

One client came to us when their CMO and CFO presented conflicting revenue numbers to the board—a $2M discrepancy neither could explain confidently. Revenue data existed in separate silos: their CRM tracked opportunities, their subscription system managed billing, and their data warehouse had incomplete linkages between them. We unified these fragmented sources using Snowflake and dbt, rebuilding all revenue and conversion reporting with centralized business logic. Marketing and Finance finally had clear, reliable insights into what was actually driving revenue. The company proceeded with a platform rebuild and brand repositioning—confident that every strategic decision was backed by numbers both teams trusted.

Here's what changed at the architecture level:

• Unified customer identity resolution. We built identity graphs that resolve the same customer across every system—web sessions, CRM records, product usage, subscription data, and support tickets. When calculating CAC, everyone works from the same customer entity. No more "Wait, are we counting this as one customer or three?"
• Centralized attribution and metrics logic in dbt. Attribution rules, revenue recognition logic, and cohort definitions live in tested, version-controlled dbt models—not scattered across BI tools. When the CFO asks "How did we calculate this?", there's a single source of truth with full lineage documentation. Changes are tracked, tested, and propagated consistently.
• Time-series alignment for apples-to-apples comparisons. We built consistent date spines and event timestamps that align marketing spend, acquisition events, activation milestones, and revenue recognition on the same timeline. Marketing can calculate CAC by cohort using the same customer acquisition dates that Finance uses for revenue analysis.

The Data Models That Make CFOs Stop Questioning Your Numbers

Trustworthy unit economics require specific data structures that most companies don't build until they're already burned by discrepancies. Here's what production-ready customer and revenue data architecture actually looks like:

• Customer dimension tables with complete history. Every customer has one record that captures their complete journey: first touch, lead source, sales attribution, activation date, subscription start, expansion events, and churn status. This eliminates the "which system is right?" debates because all stakeholders reference the same underlying customer entity with its full historical context.
• Cohort analysis tables that match how Finance thinks. CFO teams think in cohorts—customers acquired in a specific period with defined start and end dates. We build cohort tables where each customer is assigned to their acquisition cohort, with cumulative revenue tracked over time. This lets you calculate LTV, payback period, and retention curves using the same cohort definitions that appear in board decks.
• Marketing spend tables joined to customer acquisition. CAC only makes sense when you can reliably connect spend to acquired customers. We build spend tables that roll up marketing costs by channel, campaign, and time period, then join them to customer acquisition events through tested attribution logic. The result: defensible CAC calculations at whatever granularity leadership needs—by channel, by segment, by cohort, by product line.
• Revenue tables that separate bookings, billings, and recognized revenue. Finance and Marketing often talk past each other because they're measuring different things. We build revenue tables that clearly distinguish between opportunity value (what Sales closed), invoiced amount (what was billed), cash collected (what hit the bank), and recognized revenue (what counts for GAAP). Each has its place in different analyses, but they're all derived from consistent, tested data models.

From Spreadsheet Battles to Strategic Conversations

After unifying customer and revenue data for a product-led growth company, marketing teams gained real-time visibility into which user behaviors predicted expansion revenue. Previously, PLG and sales teams couldn't agree on which accounts were expansion-ready—marketing targeted based on stale data, missing opportunities and wasting spend. With unified user and account engagement data feeding Hightouch for real-time audience targeting, the company reduced infrastructure costs by hundreds of thousands while improving PLG-to-SLG expansion sales. Marketing could finally target with confidence, knowing the data was accurate and current.

This is what shifts when customer and revenue data moves from fragmented chaos to unified foundation:

• Budget conversations become strategic, not defensive. When the CFO asks "Should we increase paid social spend?", you can show cohort-level CAC trends, payback curves, and LTV projections—all using data both teams trust. The conversation moves from "Is this number right?" to "What's our target CAC for this segment, and how do we optimize toward it?"
• Attribution becomes a solved problem, not an eternal debate. With attribution logic centralized in tested dbt models, marketing can prove which channels drive revenue at the granularity leadership needs. You're not arguing about methodology—you're discussing strategic allocation based on defensible, auditable attribution.
• Board decks write themselves from live data. Unit economics metrics flow directly from the same data foundation that powers operational dashboards. The CAC/LTV numbers in your board presentation match what Finance reported, which match what Marketing sees in daily dashboards. No more week-long scrambles to reconcile spreadsheets before board meetings.
• You make data-backed decisions instead of defending data quality. Leadership meetings shift from debating whose numbers are correct to confidently driving measurable growth. Your team stops firefighting data discrepancies and starts running sophisticated segmentation, experimentation, and optimization—because the foundation is finally trustworthy.

What It Actually Takes to Get There

The companies that successfully move from CAC chaos to confident unit economics don't just hire more analysts or buy better BI tools. They invest in the underlying data architecture that makes trustworthy metrics possible:

• Unified data warehouse with clear data models. Snowflake or BigQuery as the central source of truth, with dbt transformations that implement your business logic in tested, version-controlled code. Not BI-tool calculations or analyst scripts—production-grade data models that serve the entire organization.
• Customer identity resolution that works across your stack. Purpose-built identity graphs that resolve the same person across web, product, CRM, billing, and support. This is specialized work that most companies underestimate until they're drowning in duplicate customer records.
• Standardized metric definitions that all teams reference. CAC, LTV, MRR, ARR, churn—these aren't just formulas in spreadsheets. They're materialized in your data warehouse with documented logic, test coverage, and clear ownership. When definitions change, the changes propagate consistently.
• Fast, iterative delivery that doesn't disrupt the business. You can't afford a 6-month "big bang" data warehouse project that leaves leadership blind during the transition. The right approach delivers board-relevant metrics in weeks, then expands incrementally—your existing reports keep flowing while the new foundation gets built underneath.

‍

Most organizations realize they need this after a painful board meeting or budget battle. The CFO and CMO present different revenue numbers. Acquisition costs are climbing but nobody can explain why with confidence. Strategic initiatives stall waiting for data that should be straightforward but requires weeks of manual reconciliation.

The companies that fix this first—before the crisis meeting—are the ones making confident growth decisions while competitors are still arguing about whose spreadsheet is right.

Ready to move from CAC chaos to confident unit economics? We've built production-ready customer and revenue data foundations for companies like Etsy, Calendly, Nutrafol, and Replit—eliminating the trust gap between Finance and Marketing in weeks, not quarters. 

Let's talk about your specific metrics challenges.