Your ProfitWell dashboard says logo retention is 91% and your Q3 2025 cohort is trending above average. Then a VC partner’s analyst pulls your raw Stripe data, cross-references it against your Salesforce pipeline, and calculates retention at 86%. Five points of difference. The board deck you’ve been presenting for six months is suddenly a liability.
This isn’t a hypothetical. It’s the most common diligence failure for Series B SaaS companies, and it happens because ProfitWell, Baremetrics, Recurly Analytics, and ChurnZero each read from a single data source. Billing analytics tools see what the billing system recorded. They don’t see what the CRM says, what product usage signals, or where those systems disagree. VCs see all three.
This guide explains why single-source billing analytics produce cohort data that breaks under due diligence scrutiny, shows a worked example of how cross-system reconciliation changes retention numbers, and compares the cohort analysis methodology of ProfitWell, Baremetrics, Recurly Analytics, ChurnZero, Paddle, and Eru across the dimensions that matter for investor Q&A.
The Problem: Billing Data Tells a Partial Story
Every billing analytics tool works the same way: connect to Stripe (or Chargebee, or Recurly), read subscription events, and compute metrics. When a subscription is created, that’s new MRR. When it’s upgraded, that’s expansion. When it’s cancelled, that’s churn. The math is correct — for what the billing system recorded.
The problem is that billing systems don’t record everything that matters for cohort analysis. At any SaaS company above $10M ARR, revenue events happen outside the billing system constantly:
- Mid-cycle contract amendments negotiated in Salesforce that don’t trigger a Stripe subscription change until renewal. The customer’s contractual value changed, but billing doesn’t reflect it for months.
- Invoice-based enterprise payments processed outside the billing system. A $200K annual contract paid via wire transfer doesn’t exist in Stripe at all. Your cohort analysis excludes your largest customers.
- Account mergers and splits that happen in the CRM but not in billing. Two Stripe subscriptions consolidated into one Salesforce account means your billing-based cohort double-counts the customer before the merge and loses them after.
- Discounts and credits applied in billing but not reflected in CRM deal values. Your Stripe MRR for an account says $8K; Salesforce says the deal is worth $10K. Which number belongs in the cohort?
- Product usage decline that predicts churn weeks before the billing cancellation. A customer whose daily active users dropped 60% three months ago is still “active” in billing-based cohort analysis until they cancel.
Each of these creates a gap between what billing analytics reports and what an investor calculates from a full data room. Individually, they’re small. Across 200–500 accounts, they compound into a 3–8% discrepancy that changes cohort curves materially.
Worked Example: How Cross-System Reconciliation Changes a Cohort Retention Table
Consider a mid-market SaaS company at $25M ARR with 300 accounts. Here’s their Q2 2025 cohort (42 customers acquired) as it appears in billing-only analytics versus reconciled cross-system data.
Billing-Only Cohort (from ProfitWell or Baremetrics)
| Cohort: Q2 2025 | Month 0 | Month 3 | Month 6 | Month 9 |
|---|---|---|---|---|
| Accounts | 42 | 39 | 37 | 35 |
| Logo retention | 100% | 92.9% | 88.1% | 83.3% |
| MRR | $168K | $159K | $155K | $152K |
| Revenue retention | 100% | 94.6% | 92.3% | 90.5% |
This looks solid. 83% logo retention at month 9 with revenue retention holding above 90% suggests mild churn offset by some expansion. A board deck built on these numbers tells a reasonable story.
Reconciled Cohort (Stripe + Salesforce + Product Analytics)
Now the same cohort after reconciling billing data against Salesforce deal values and Amplitude product usage:
| Cohort: Q2 2025 | Month 0 | Month 3 | Month 6 | Month 9 |
|---|---|---|---|---|
| Accounts | 45 | 40 | 36 | 34 |
| Logo retention | 100% | 88.9% | 80.0% | 75.6% |
| MRR (reconciled) | $193K | $172K | $161K | $155K |
| Revenue retention | 100% | 89.1% | 83.4% | 80.3% |
What Changed and Why
The reconciled version reveals five categories of discrepancy that billing-only tools missed:
- Three invoice-based accounts ($25K combined MRR) were missing from billing. These enterprise customers pay via wire transfer and don’t have Stripe subscriptions. They belong in the cohort but ProfitWell never saw them. Two of them churned by month 6.
- Two accounts had Salesforce deal values that exceeded Stripe MRR by $4K total. The CRM reflected contract amendments that hadn’t flowed through to billing yet. The investor’s analyst, pulling both data sources, would calculate higher starting MRR and therefore lower retention.
- One account “retained” in billing was functionally churned. Their Stripe subscription was active but product usage had dropped to zero for 11 weeks. By the time billing cancellation happens, this account will appear in a later cohort’s churn — distorting both periods.
- A merged account was double-counted. Two Stripe subscriptions were consolidated into one Salesforce account during month 4. Billing-based cohort analysis counted both subscriptions; the reconciled view counts one account.
- Expansion revenue was overstated. A $3K billing upgrade was actually a correction of a previous undercharge, not genuine expansion. The CRM notes documented the correction; billing recorded it as growth.
The net effect: revenue retention drops from 90.5% to 80.3% at month 9. That’s the difference between “mild churn with expansion offset” and “significant retention problem.” An investor seeing the second number evaluates the business very differently — and they will see the second number because they pull from all available data sources.
How Each Tool Handles Cohort Analysis
| Dimension | ProfitWell (Paddle) | Baremetrics | Recurly Analytics | ChurnZero | Eru |
|---|---|---|---|---|---|
| Data sources for cohorts | Billing only (Stripe, Chargebee, Paddle) | Billing only (Stripe, Chargebee, Braintree) | Recurly billing only | CRM + product usage (pushed in) | Billing + CRM + product analytics + support |
| Cohort definition | Subscription start date, plan, pricing tier | Signup date, plan | Billing start date, plan, billing period | Custom (based on pushed data) | Any dimension: acquisition channel, deal size, CSM, product engagement |
| Revenue reconciliation | — | — | — | — | ✓ Automatic billing–CRM matching at account level |
| Invoice/wire payments included | — | — | — | Partial (if pushed from CRM) | ✓ Via CRM integration |
| Product usage signals | — | — | — | ✓ Native tracking + integrations | ✓ Via Mixpanel, Amplitude, Segment |
| Detects “zombie” accounts | — | — | — | ✓ Via usage tracking | ✓ Cross-references billing status against product usage |
| Audit trail for cohort data | Billing event log only | No reconciliation trail | Recurly transaction log only | Activity log within ChurnZero | ✓ Full cross-system reconciliation log |
| Handles account mergers/splits | — | — | — | Partial (manual) | ✓ Automatic via truth graph entity matching |
| Due diligence readiness | Partial — billing-side only | Partial — billing-side only | Limited — Recurly data only | Partial — depends on data pushed in | ✓ Pre-reconciled with audit trail |
ProfitWell (Paddle): What It Does Well and Where It Falls Short
Strengths: ProfitWell’s free subscription analytics and Retain dunning product genuinely reduce involuntary churn from failed payments. The cohort analysis segments by plan and pricing tier, which is useful for understanding how different pricing structures affect retention. If your billing system is your genuine single source of truth and you don’t use a CRM for revenue tracking, ProfitWell’s cohort data is accurate for your use case.
Limitation for due diligence: Since Paddle’s acquisition, ProfitWell’s product direction is increasingly tied to the Paddle billing ecosystem. For companies on Stripe, the integration works but is no longer the primary investment focus. More critically for diligence: ProfitWell cannot tell you whether the $12K MRR it reads from Stripe for Account X matches the $15K deal value in Salesforce. When an investor finds that gap, your ProfitWell cohort curves become unreliable in their analysis.
Baremetrics: What It Does Well and Where It Falls Short
Strengths: Clean UI with quick setup. The email reporting and forecasting features give founders a fast pulse on subscription metrics. Recover provides genuine value for reducing failed payment churn. For pre-Series A companies with a single billing source and no CRM complexity, Baremetrics is a pragmatic choice for basic cohort monitoring.
Limitation for due diligence: Single billing source only — no multi-source support like ChartMogul. Cohort analysis is limited to signup month and plan. Cannot segment cohorts by acquisition channel, deal size, or product engagement — the dimensions VCs request during diligence. No CRM integration means no reconciliation capability.
Recurly Analytics: What It Does Well and Where It Falls Short
Strengths: Native integration with Recurly’s billing lifecycle means zero-setup analytics for Recurly customers. Revenue recognition features are useful for accounting compliance. Cohort views by plan and billing period are accurate within the Recurly ecosystem.
Limitation for due diligence: Locked to Recurly data exclusively. If you have any revenue outside Recurly (enterprise invoices, secondary billing systems, manual payments), those accounts are invisible to cohort analysis. Cannot incorporate CRM or product analytics data. If you migrate away from Recurly, your historical cohort data goes with it.
ChurnZero: What It Does Well and Where It Falls Short
Strengths: ChurnZero adds a dimension that pure billing tools miss: product usage. Native product tracking and engagement scoring mean ChurnZero can identify “zombie” accounts (paying but not using) and flag retention risk from behavioural signals. Health scores that combine usage, support, and CS engagement provide more context than billing-only metrics. For customer success teams that need operational workflow alongside analytics, ChurnZero is a strong platform.
Limitation for due diligence: ChurnZero doesn’t natively connect to billing systems for revenue reconciliation. Revenue data is typically pushed in from the CRM or billing system, which means it reflects whatever that source recorded — including discrepancies. ChurnZero can tell you which accounts are at risk from a usage perspective, but it cannot tell you whether the revenue attributed to those accounts in billing matches what’s recorded in the CRM. For diligence, where investors cross-check revenue numbers across sources, this gap matters.
Why Cross-System Reconciliation Changes the Analysis
The core issue isn’t that billing analytics tools are wrong. They’re accurate for what they measure. The issue is that VCs don’t evaluate retention from a single source.
During due diligence, an investor’s analyst will:
- Pull raw billing data from Stripe or Chargebee
- Pull CRM data from Salesforce or HubSpot
- Cross-reference account by account to check for revenue agreement
- Calculate cohort retention independently from the combined data
- Compare their calculation against your board deck
If your board deck was built from ProfitWell or Baremetrics (billing only), the investor’s cross-system calculation will produce different numbers. Every discrepancy becomes a question: “Why does your Stripe data show $152K retained for this cohort when Salesforce shows $143K?” You either have the answer prepared or you scramble to reconcile during the diligence process — which signals operational immaturity.
What Reconciled Cohort Analysis Looks Like
Eru produces cohort analysis by first reconciling billing and CRM data at the account level through its truth graph. Before computing any cohort metric, Eru matches every billing subscription to its corresponding CRM record, identifies discrepancies, resolves entity conflicts (mergers, splits, orphaned accounts), and flags revenue gaps.
The resulting cohort table reflects:
- All revenue sources. Invoice-based enterprise accounts are included alongside Stripe subscriptions because both are in the CRM.
- Reconciled MRR. When billing says $12K and CRM says $15K for the same account, the discrepancy is flagged and the reconciled value uses the higher-confidence source with an audit trail explaining the decision.
- Product-informed retention. Accounts with zero product usage for 60+ days are flagged as at-risk even if billing is still active, giving you a more honest view of functional retention.
- Cross-dimensional segmentation. Cohorts can be sliced by acquisition channel (inbound vs outbound), deal size, CSM coverage, product engagement level, or any combination — because the data is drawn from billing, CRM, and product analytics together.
Criteria That Matter for Due Diligence Cohort Analysis
When evaluating any tool for producing cohort analysis that will be presented during fundraising or due diligence, these are the criteria that separate tools that help from tools that create risk.
Data source coverage
Does the tool read from billing only, or does it incorporate CRM and product data? Billing-only tools produce accurate billing metrics, but VCs cross-check against CRM data. If the tool can’t reconcile across sources, your numbers will diverge from the investor’s independent calculation.
Reconciliation logic
When billing and CRM disagree on an account’s revenue, does the tool flag the discrepancy, or does it silently use one source? During diligence, every unflagged discrepancy is a surprise that erodes trust. Tools with explicit reconciliation logic (like Eru’s truth graph) identify and document gaps before an investor finds them.
Cohort flexibility
Can you segment cohorts by dimensions beyond signup date and plan? VCs ask for retention by acquisition channel, deal size, customer tier, and product engagement. These require data from multiple systems. If you can answer these questions in a diligence call instead of following up in two days with a spreadsheet, it signals the operational maturity investors reward.
Export formats
Can you produce board-ready outputs directly, or does the data need to be reformatted in a spreadsheet? Time spent reformatting analytics exports into board deck format is time your finance team spends every month that a reconciled platform eliminates.
| Due Diligence Criterion | ProfitWell / Baremetrics / Recurly | ChurnZero | Eru |
|---|---|---|---|
| Data source coverage | Billing only | CRM + product (pushed in) | Billing + CRM + product + support (native) |
| Reconciliation logic | None — single source | None — uses data as pushed | Automatic with audit trail |
| Cohort flexibility | Date and plan only | Custom segments from pushed data | Any dimension across all connected sources |
| Export formats | Charts, CSV | Reports, CSV | Dashboard export, live data access |
| Investor confidence level | Billing metrics verified; CRM gaps remain | Usage signals verified; revenue gaps remain | All metrics pre-reconciled across sources |
Who Should Use What
Stay with ProfitWell or Baremetrics if:
- You’re pre-Series A with a single billing source and no CRM complexity
- Your billing system is genuinely your only source of revenue truth (no invoice-based customers, no manual payments)
- You’re not preparing for fundraising or due diligence in the next 12 months
- Failed payment recovery (dunning) is your primary retention challenge
Consider ChurnZero if:
- Your CS team needs operational workflow (playbooks, tasks, health scoring) alongside analytics
- Product usage tracking and engagement scoring are a priority for retention
- You have the resources to push accurate revenue data into ChurnZero from billing and CRM
- You accept that revenue reconciliation will still need a separate process for diligence
Choose Eru if:
- Your Stripe and Salesforce numbers don’t match and nobody can quantify the gap
- You’re preparing for a fundraise and need cohort data that survives independent verification
- Your board has questioned the accuracy of your retention data
- You have enterprise customers who pay via invoice outside your billing system
- You want cohort analysis segmented by acquisition channel, deal size, and product engagement — not just billing plan
- You don’t have a data team to build cross-system reconciliation in a warehouse
The Bottom Line
Cohort analysis is the metric VCs use to evaluate product-market fit. Improving cohorts prove that your product, onboarding, and customer success are getting better with each batch of customers. Declining cohorts signal structural problems that more sales won’t fix.
The accuracy of those cohort curves depends entirely on the data behind them. ProfitWell, Baremetrics, and Recurly produce accurate cohort analysis from billing data. ChurnZero adds valuable product usage context. But none of these tools reconcile billing against CRM — and that reconciliation is exactly what determines whether your cohort curves survive VC due diligence.
Eru exists to close that gap. By reconciling data across billing, CRM, and product analytics before calculating cohort metrics, it produces retention curves that match what an investor would calculate from a full data room. For GTM data teams preparing board decks, that’s the difference between presenting numbers you hope are right and presenting numbers you know are defensible.
Frequently Asked Questions
ProfitWell vs Recurly Analytics vs ChurnZero — which gives the most accurate customer cohort analysis for due diligence?
For due diligence accuracy, none of these tools reconcile billing against CRM — which is the first thing an investor’s analyst checks. ProfitWell analyses billing data only. Recurly Analytics is limited to Recurly subscribers. ChurnZero provides health scoring but relies on externally pushed data. Eru connects billing, CRM, and product data to produce cohort analysis from reconciled cross-system data, so your retention curves survive independent verification.
Does Eru handle NRR accuracy issues between Stripe and Salesforce that come up during due diligence?
Yes. Eru’s truth graph matches every Stripe subscription to its corresponding Salesforce account, flags discrepancies, and produces NRR from reconciled data. The audit trail shows exactly which accounts contributed to each NRR component and where billing and CRM disagreed. Your reported NRR matches what a VC’s analyst would independently calculate from combined raw data.
How much should I budget for accurate NRR and cohort analysis for board decks?
Billing-only analytics (ProfitWell, Baremetrics) cost $0–$25K/year but only cover billing-side metrics. CS platforms (ChurnZero, Gainsight) cost $50K–$150K/year. BI tools (Looker) cost $50K–$150K/year plus data team salary. The hidden cost is the 2–5 days per month spent manually reconciling billing exports against CRM data before board meetings — plus the valuation risk if those numbers don’t hold up during diligence.
What is a cohort retention table and why do VCs care about it?
A cohort retention table groups customers by acquisition period and tracks revenue retained over time. VCs use it to assess product-market fit: improving cohorts (each quarter retains better than the last) prove compounding value. During diligence, investors rebuild your cohorts from raw data. If your billing-only cohort analysis doesn’t match what they calculate using CRM data, the discrepancy becomes a valuation conversation.
See what your cohort retention looks like when billing, CRM, and product data are reconciled. Book a free revenue audit — we’ll show you where your retention numbers diverge across systems and what it means for your next board deck or fundraise.
Book a revenue audit →