Thinklytics

Enterprise SaaS Company · Technology & SaaS · Austin, TX · 14 weeks

Enterprise SaaS Company Data Foundation case study

A mid-market SaaS company was struggling with six conflicting ARR numbers across finance, sales, and the board, with a $1.4M gap between them. We developed a certified revenue metric layer that standardized ARR calculations and aligned all teams on one accurate figure for board reporting.

Challenge

Finance pulled ARR numbers from billing. Sales used the CRM. The CFO combined both in a spreadsheet. The board got a different figure from investor reports. The gap between the highest and lowest ARR was $1.4M on a $22M base. Meetings wasted time arguing over which number to trust.

Approach

We identified how each source system calculated ARR, focusing on upgrades, downgrades, churn, and multi-year contracts. We documented every difference and created one clear ARR definition. Then, we built a certified ARR calculation in dbt using the data warehouse and linked all four reporting tools directly to this standardized metric.

Outcome

By week 8, we identified and fixed the $1.4M ARR discrepancy. We aligned all four reporting systems to display the same ARR number. This eliminated metric disputes in board meetings. We also expanded the certified ARR framework to include NRR, GRR, and churn, creating a consistent and reliable revenue metrics layer across the business.

The data itself wasn’t the issue. The real problem? Everyone was using the same words but meant totally different things.

Here’s the deal: every system was crunching ARR in its own way. Finance skipped professional services. Sales mixed in pipeline numbers. The investor tool had this odd churn date thing going on. No surprise the numbers didn’t match. So, we brought everyone together and locked in one simple, clear ARR definition that everyone could rally behind.

How We Jumped Into dbt to Finally Sort Out Our Data Certification Headaches

Here’s the deal. We built the certified ARR calculation straight into dbt. This let us follow the data from the ground up to the final report without missing a beat. Every move we made was tracked, tested, and version controlled. The coolest part? When the ARR rules change, we just tweak it in one spot. Then, boom, all four reports update themselves. No extra work needed.

We tore apart every revenue number to spot what’s slowing us down

First, we locked in the ARR numbers. Then we spent around three weeks digging into NRR, GRR, and churn to make sure everything lined up. That deep dive gave us real confidence that their revenue metrics were solid and trustworthy. Now, the CFO, CRO, and the board actually rely on these numbers when making decisions.

Results

  • $1.4M Revenue discrepancy resolved
  • 6 to 1 ARR definitions unified
  • 4 Reporting surfaces on one number
  • 14 wks Full metrics layer delivered

For two years, we kept going back and forth over ARR in every board meeting. Thinklytics resolved that in eight weeks, and since then, we haven’t had any debates about metrics at the board level.

Chief Financial Officer, Mid-Market SaaS Platform

In the client's terms

The question

Where does this number come from, and can the path be trusted?

The situation

Finance took ARR from billing, sales took it from the CRM, the CFO combined both in a spreadsheet, and the board saw a fourth figure from investor reporting. The spread between highest and lowest was $1.4M on a $22M base.

What we found

Six definitions of ARR existed and all six were internally consistent. The disagreement lived entirely in the edge cases: upgrades, downgrades, churn timing and multi-year contracts. Nobody had written those down, so every meeting relitigated them.

What it produced

The $1.4M discrepancy was identified and closed by week 8. All four surfaces now return the same ARR. Board meetings stopped arguing about which number was real. The certified framework was extended to NRR, GRR and churn.

The roadmap

Each source system was traced to establish how it actually calculated ARR, with every difference documented rather than averaged away. One definition was agreed and the edge cases were written into it. A certified ARR calculation was built in dbt on the warehouse, and all four reporting surfaces were pointed at it.

The work behind this result

Related case studies

Frequently asked questions

Why do finance and sales report different ARR figures?

Because each defines it correctly for its own purpose. At this enterprise SaaS company, finance took ARR from billing, sales from the CRM, the CFO combined both in a spreadsheet and the board saw a fourth figure from investor reporting. The $1.4M spread on a $22M base lived entirely in upgrades, downgrades, churn timing and multi-year contracts.

How do you fix conflicting revenue metrics?

Document every difference rather than averaging them, agree one definition with the edge cases written into it, then implement it once where every tool reads it. This company's certified ARR calculation was built in dbt on the warehouse and all four reporting surfaces were pointed at it.

How long does a certified metrics layer take to build?

Fourteen weeks here, with the $1.4M discrepancy identified and closed by week 8. The remaining time extended the certified framework to NRR, GRR and churn so the next revenue metric does not repeat the pattern.