← Back to blog NRR

NRR Forecasting Methodology for Series B SaaS: From $10M to $50M ARR

The methodology, metrics, and tool criteria you need to forecast net revenue retention accurately as you scale past $10M ARR.

At $15M ARR and a fresh Series B, your board expects retention forecasts you can defend. Not a spreadsheet with last quarter’s churn rate applied to next quarter’s renewal cohort. A methodology — one that separates gross from net retention, accounts for expansion and contraction dynamics, and produces a number that holds up when actuals come in.

This guide covers the forecasting methodology that works at the Series B stage: how to calculate and present gross and net retention, how to build forecasts your board and investors will trust, what to look for if you’re evaluating NRR forecasting tools, and how those tools are typically priced.

Why NRR Forecasting Breaks at the Series B Stage

At Series A, most founders track retention in a spreadsheet. It works when you have 30 customers and know each one by name. At Series B — 100 to 300 accounts, multiple segments, enterprise and mid-market mixed together — the spreadsheet stops being reliable.

Three things change at the $10M–$50M ARR stage:

The methodology below is designed for this stage — complex enough to be accurate, structured enough that your VP of RevOps or finance lead can own and present it.

Gross vs Net Revenue Retention: How to Calculate and Forecast Each for Board Presentations

Gross Revenue Retention (GRR)

Formula: (Starting MRR − Churn − Contraction) ÷ Starting MRR × 100

GRR measures how well you keep the revenue you already have, ignoring expansion entirely. It answers the question: how leaky is the bucket?

Series B benchmarks:

How to forecast GRR: Start with your renewal cohort for the forecast period. Segment accounts by risk tier (based on usage, engagement, support history, and contract signals). Apply tier-specific churn and contraction rates rather than a single average. Sum the expected retained revenue and divide by starting MRR.

Board presentation tip: Always present GRR alongside NRR. GRR shows the health of your existing base without the mask of expansion revenue. A company with 120% NRR but 80% GRR is running on a treadmill — expansion is hiding a churn problem. Boards that see both metrics understand the true dynamics.

Net Revenue Retention (NRR)

Formula: (Starting MRR + Expansion − Churn − Contraction) ÷ Starting MRR × 100

NRR is the single most important metric for Series B SaaS companies. It tells you whether your existing customer base is growing, stable, or shrinking — independent of new logo acquisition.

Series B benchmarks:

How to forecast NRR: Take the same risk-tiered cohort used for GRR, but add expansion projections. Identify accounts showing expansion signals — seat utilisation nearing limits, feature adoption indicating upsell readiness, active conversations about additional products. Apply tier-specific expansion rates alongside the retention and contraction assumptions.

Board presentation tip: Break NRR into its components — gross retention, contraction, and expansion — as a waterfall chart. This shows the board exactly where value is being created or lost. A single NRR number tells you the outcome. The components tell you the story.

Presenting Both Metrics to Your Board

Your board slide should include:

  1. GRR and NRR side by side for the current quarter and trailing four quarters, showing the trend.
  2. NRR waterfall breaking down starting ARR, expansion, contraction, and churn to ending ARR.
  3. Cohort view showing how retention evolves for each quarterly acquisition cohort — improving cohorts over time signals that your product and onboarding are getting better.
  4. Forward-looking forecast for the next quarter with the methodology described above, including the confidence range based on your risk tiers.

This structure gives your board what they need to assess retention health without requiring a 30-minute explanation of your data pipeline.

Building a Reliable NRR Forecasting Model

Step 1: Define Your Cohort

Choose a cohort definition and stick with it. Most Series B companies use a trailing 12-month cohort for annual NRR and a monthly cohort for operational tracking. The key is consistency — changing your cohort definition between board meetings erodes trust.

Step 2: Segment by Risk

Not every account has the same likelihood of churning, contracting, or expanding. Segment your renewal cohort by risk using signals from across your stack:

At the Series B stage, you don’t need a machine learning model. A weighted scoring system using 4–6 signals that your team can explain and defend is more valuable than a black box.

Step 3: Apply Segment-Specific Assumptions

For each risk segment, apply different assumptions for retention, expansion, and contraction. Calibrate these against your historical data. If your “high risk” segment has historically retained at 55%, use 55% — not a generic industry number.

Step 4: Build the Forecast Range

Present a base case, upside case, and downside case. The base case uses your segment assumptions as-is. The upside assumes your intervention on at-risk accounts succeeds (e.g., 20% of red accounts saved). The downside assumes a negative scenario (e.g., one large account churns unexpectedly).

A range communicates sophistication. A single number communicates false precision.

Step 5: Validate Quarterly

After each quarter, compare your forecast to actuals at the segment level. Which segments were you right about? Which were off? Adjust your scoring weights and segment assumptions based on what actually happened. A forecast that improves each quarter is more valuable than one that’s perfect once.

What to Look for in NRR Forecasting Tools

If you’re evaluating tools to help with NRR forecasting, here are the criteria that matter at the Series B stage. This is what separates tools that actually improve forecast accuracy from those that just add another dashboard to check.

1. Multi-System Data Integration

Your retention signals live in 4–6 different tools. Any forecasting platform that only connects to one system — billing only, CRM only, or product analytics only — gives you a partial picture. Look for native integrations with your billing system (Stripe, Chargebee, Recurly), CRM (Salesforce, HubSpot), support (Zendesk, Intercom), and product analytics (Amplitude, Mixpanel, or direct database connections). Eru connects across all of these sources in minutes with read-only OAuth, and its AI agent maps your data automatically — no schema configuration or SQL required.

2. Account-Level Granularity

Aggregate metrics hide the signal. You need account-level risk scores, not just portfolio-level retention rates. The tool should show you which specific accounts are driving your NRR up or down, and let you drill into the signals behind each score. Eru scores every account daily based on cross-system signals, giving you the account-level detail behind every number.

3. Automated Data Reconciliation

If your Stripe MRR doesn’t match your Salesforce ARR, your retention metrics are wrong from the start. The tool should reconcile data across systems automatically and flag discrepancies — orphaned accounts, billing mismatches, and MRR gaps. Eru’s data integrity layer catches these discrepancies automatically, ensuring your NRR calculations start from accurate, reconciled data.

4. Forward-Looking Signals, Not Just Lagging Metrics

Historical churn rate tells you what happened. You need signals that predict what will happen — declining usage, support escalations, champion departures, billing warnings. The tool should surface leading indicators, not just report trailing metrics. Eru correlates signals across billing, usage, support, and CRM data to surface at-risk accounts before they churn.

5. Clear Segmentation and Cohort Analysis

You should be able to segment by account size, industry, acquisition cohort, plan type, and custom attributes without building SQL queries or waiting for a data team. The tool should make segmentation self-serve.

6. Board-Ready Outputs

The forecasting tool should produce outputs that go directly into a board deck — NRR waterfall, cohort charts, GRR and NRR trends, and risk-tiered account lists. If you need to manually rebuild every chart in slides, the tool is creating work instead of saving it.

7. Implementation Speed

At the Series B stage, you don’t have 3 months for a tool rollout. Look for time-to-value measured in days, not quarters. If the tool requires a dedicated implementation project, data engineering support, or extensive configuration, the cost and risk may outweigh the benefit. Eru is designed for teams without dedicated data engineering — most teams are up and running in under a day.

How NRR Forecasting Tools Are Typically Priced

Pricing in the revenue intelligence and customer success platform space varies widely. Here’s how the most common models work, so you can budget accurately when evaluating options.

ARR-Based Pricing

Many platforms charge based on your company’s ARR rather than a flat fee. Typical pricing ranges from 0.5% to 2% of your ARR. For a $15M ARR company, that means $75K to $300K per year. This model is common among enterprise customer success platforms like Gainsight and Totango. It aligns the vendor’s revenue with your growth but can become expensive as you scale — a $50M ARR company could be paying $250K–$1M per year.

Per-Seat Pricing

Some tools charge per user seat, typically $100–$500 per user per month. This is common among CRM-adjacent and analytics tools like ChurnZero and Vitally. For a RevOps team of 5–10 users, that’s $6K–$60K per year. Per-seat pricing is predictable and scales with your team size rather than revenue, but it can limit adoption — you end up restricting access to save costs, which reduces the tool’s value.

Platform or Flat-Fee Pricing

Some platforms charge a flat monthly or annual fee based on features and tier. This is more common among newer tools and is typically the most predictable model. Flat fees for mid-market tools range from $1K to $5K per month ($12K–$60K per year). This model works well at the Series B stage because costs are predictable and don’t spike as you grow.

Usage-Based Pricing

A smaller number of tools charge based on the volume of data processed, number of accounts tracked, or API calls. Typical ranges are $0.50–$5 per tracked account per month. For a company with 200 accounts, that’s $1.2K–$12K per year. This model scales directly with your customer base and can be cost-effective at lower volumes.

What to Budget at Series B

For a $15M ARR Series B company evaluating NRR forecasting as part of a broader retention and revenue intelligence investment, a reasonable budget is $30K–$80K per year. Below $30K, you’re likely getting a reporting tool that lacks the multi-system integration and account-level scoring needed for accurate forecasting. Above $80K, you’re likely paying for enterprise features you won’t use for another 2–3 years.

When comparing vendors, look beyond the list price. Factor in implementation costs (some platforms charge $10K–$50K for onboarding), time to value (weeks vs months), ongoing maintenance burden (do you need a dedicated admin?), and the cost of the data team hours the tool replaces.

Bringing It Together

NRR forecasting at the Series B stage isn’t about a perfect model. It’s about a defensible methodology that improves over time. The steps are:

  1. Separate gross and net retention in every analysis and board presentation.
  2. Segment accounts by risk using cross-system signals, not gut feel.
  3. Apply segment-specific assumptions calibrated against your own historical data.
  4. Present a range, not a single number.
  5. Validate against actuals every quarter and adjust.

The companies that get NRR forecasting right at this stage aren’t the ones with the most sophisticated models. They’re the ones with the most connected data — because forecast accuracy is bounded by data accuracy, and data accuracy requires visibility across your entire stack.

Related Resources

Related reading: ChartMogul vs Baremetrics vs ProfitWell vs Eru for Board Reporting & NRR Due Diligence — a side-by-side comparison of NRR calculation accuracy and VC due diligence readiness across the leading subscription analytics tools.

Eru connects your billing, CRM, support, and product data to produce live NRR, GRR, and account-level risk scores — so your forecasts are built on reconciled data and leading indicators, not spreadsheet assumptions.

Book a churn audit →