.png)
Frameworks, core principles and top case studies for SaaS pricing, learnt and refined over 28+ years of SaaS-monetization experience.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Net Dollar Retention the Critical SaaS Metric for Sustainable Growth and Investor Confidence
A SaaS company can report strong new-logo growth while quietly losing the economic argument inside its installed base. Customers may renew at smaller volumes, resist price increases, or buy fewer add-ons than expected. Revenue can still rise for a period, but each quarter demands more selling expense to replace value that should have compounded.
That distinction matters more as SaaS companies add consumption, credits, and AI agents to established subscriptions. A pricing change is no longer simply a list-price decision. It is a bet that customers will stay, expand, and accept a larger share of spend over time. Investors look for evidence that this bet is working in the revenue already under contract.
Monetizely's position is clear: Net Dollar Retention should be the governing test for SaaS pricing decisions after launch. A pricing model that cannot sustain or improve NDR across defined customer cohorts may produce short-term bookings, but it does not create durable growth or investor confidence.
Investor confidence starts with the revenue base already won
Net Dollar Retention, often called Net Revenue Retention or dollar-based net retention, measures how recurring revenue from a starting group of customers changes over a set period. It includes churn, downgrades, and expansion. It excludes revenue from customers acquired after the cohort begins.
The measure answers a question that total revenue cannot: If the sales team stopped signing new customers for one year, would the existing base shrink, hold steady, or grow? A company at 110% NDR has grown the recurring value of its starting customer cohort by 10%. A company at 95% has lost 5%, even if new sales concealed that loss in the income statement.
Public companies define the calculation in similar ways. CrowdStrike begins with the ARR of subscription customers from 12 months earlier, then includes expansion and deducts contraction and churn, while excluding new subscription customers. Datadog uses the same cohort logic in its trailing-12-month dollar-based net retention calculation.
The formula should be simple enough for the chief product officer, chief revenue officer, and finance leader to use in the same meeting.
| Component | Formula treatment | Worked cohort example |
|---|---|---|
| Starting ARR | Revenue from the same customers at the start of the period | $10.0M |
| Less: churned ARR | Revenue lost when customers leave | ($0.6M) |
| Less: contraction ARR | Revenue lost through fewer seats, lower usage, or lower package levels | ($0.5M) |
| Add: expansion ARR | Revenue from added seats, products, usage, or price realization | $1.8M |
| Ending ARR from starting cohort | Starting ARR - churn - contraction + expansion | $10.7M |
| Net Dollar Retention | Ending cohort ARR ÷ starting cohort ARR | 107% |
A 107% NDR means the company retained $10.7 million from a $10 million starting cohort. New-logo revenue does not appear in the numerator, which makes NDR a disciplined test of whether customers see enough value to deepen their commitment. The formula follows the cohort approach described in public disclosures by CrowdStrike and Datadog.
Gross Revenue Retention should sit beside NDR, not beneath it. Gross retention removes expansion from the equation:
[ \text{GRR} = \frac{\text{Starting ARR} - \text{Churned ARR} - \text{Contraction ARR}}{\text{Starting ARR}} \times 100 ]
In the example above, GRR is 89%. Expansion lifted the final NDR to 107%, but the company still lost $1.1 million of recurring revenue before expansion. Leaders who celebrate NDR without watching gross retention can miss a weakening renewal foundation.
Each point of NDR changes the new-business burden
NDR is not merely a customer-success metric. It changes the amount of new ARR required to hit a growth plan. That arithmetic explains why investors give high-NDR companies more room to invest through cycles: the installed base carries more of the growth load.
Consider a company that starts the year with $100 million in ARR and wants to finish at $120 million.
The table makes the management point plain: on a $100 million opening base, one NDR point equals $1 million of cohort ARR. A pricing decision that improves NDR by four points can remove $4 million of new ARR from the annual selling burden, before considering the lower cost of expanding existing customers.
That does not mean every company should force NDR upward through aggressive price increases. Customers can react to a blunt increase by cutting seats, downgrading packages, or moving usage elsewhere. Durable NDR comes from a price and meter that match the value customers receive, then make expansion a natural next step.
Public disclosures set reference points, not interchangeable targets
Operators often ask for the “right” NDR benchmark. The better question is which companies face similar renewal behavior, customer concentration, usage patterns, and expansion paths. A cloud data platform, a cybersecurity suite, and a CRM platform can all report NDR, yet they may calculate it on different revenue bases and customer cohorts.
The following public disclosures provide useful reference points, not a universal target.
Across these four public-company disclosures, reported NDR ranges from 103.5% to 126% at the stated dates. The range is meaningful because each company has demonstrated expansion from its installed base; it is not an excuse to assign a Snowflake-level target to a company with a fundamentally different product, customer profile, or billing model.
The definitions matter. Snowflake's 126% rate reflects product-revenue behavior in a consumption-oriented platform. HubSpot reported 103.5% for 2025 under an adjusted calculation and later described seat and credit expansion as NDR drivers. Datadog's low-120%s coincide with broad multi-product adoption: 85% of customers used two or more products as of June 30, 2026, while 58% used four or more.
The management lesson is not “copy their number.” It is “build the conditions that make expansion credible.” Those conditions include a product customers use deeply, packages that reveal a next purchase, and a pricing metric that grows when customer value grows.
NDR makes pricing metric choice accountable after the contract is signed
Monetizely's 5-Step Pricing Framework puts decisions in the order that commercial results demand. First, clarify goals and segmentation: determine whether the company needs faster adoption, higher average revenue, stronger margin, or better retention, and identify the customer groups that matter. Second, build packages that fit those groups. Third, select the pricing metric - the unit the customer is charged for, such as seats, usage, transactions, or outcomes. Fourth, set price points. Fifth, operationalize pricing through product instrumentation, billing, sales rules, and customer communication. As discussed in Monetizing Agentic AI, this order matters because price cannot repair a package, segment definition, or metric that customers do not accept.
NDR belongs most directly in Step 3, the pricing-metric decision. A metric should not be judged only by whether it is easy to quote or familiar to sales. It should be judged by whether it produces healthy expansion without creating avoidable churn, contraction, or margin risk.
The table shows why a pricing metric is a product decision as much as a finance decision. A company that charges per seat may retain predictable revenue but fail to capture value from a customer that doubles its transactions. A company that charges per API call may capture expansion but create a bill customers cannot forecast. NDR exposes both failures over time.
HubSpot offers a useful public example of the connection between commercial design and NDR. For full-year 2025, the company reported 103.5% NDR, up from 101.8% in 2024 under the reported methodology. Its February 2026 earnings call linked the improvement to seat expansion and a pricing change, while its 2026 commentary also identified credit adoption as an expansion lever.
Agentic offerings require a meter that creates expansion without hiding cost
AI and agentic products make the NDR-pricing link more urgent because the cost of serving one customer can vary sharply. A per-seat plan may be simple for a buyer, yet become uneconomic when one user runs an agent continuously. A pure token model may protect margin, yet fail because buyers cannot connect tokens to business value.
The Agentic Monetization Spectrum, or AMS, helps resolve that tension. It assesses an agent on three dimensions: zero-human ability, or how much human work remains; operational domain, or whether the agent handles one task, one business function, or multiple functions; and output/cost ratio, or how fast customer value rises relative to compute cost. As autonomy, domain breadth, and the output-to-cost ratio rise, the pricing metric should move away from a human seat and toward the output or outcome delivered.
Consider a customer-support agent that independently resolves standard cases, updates the CRM, and sends customer responses, while humans handle exceptions.
| AMS dimension | Score for the support-agent archetype | What the score means for pricing |
|---|---|---|
| Zero-human ability | Large | The agent performs most of the work, so a human seat is a weak primary anchor |
| Operational domain | Medium | The agent runs an end-to-end workflow inside customer support |
| Output/cost ratio | Inflecting | Value can outpace compute cost, but quality and reliability still need proof |
| Recommended primary meter | Resolved cases | Charge on a defined, auditable output, supported by an annual platform commitment |
The score points to a clear architecture: resolved cases should be the primary meter, while an annual platform commitment provides budget predictability and funds the fixed value of integration, governance, and availability. The company should define a “resolved case” before launch, specify exclusions, and give customers a visible record of the outcomes billed.
NDR then becomes the proof that the meter works. Healthy agentic NDR should show that customers renew the annual commitment, expand case volume as trust rises, and do not contract because billing feels arbitrary. If expansion comes only from an unexpected usage spike followed by a renewal downgrade, the company has monetized activity rather than durable customer value.
NDR is powerful because it compresses a great deal of customer behavior into one figure. That same compression can mislead leaders who do not inspect the components.
The most common operator mistakes are straightforward:
Counting new-logo ARR in the cohort. New revenue can make a retention picture look stronger while hiding churn and contraction among existing customers.
Mixing ARR, recognized revenue, bookings, and billings. The denominator and numerator must use the same commercial measure, or the rate becomes uninterpretable.
Celebrating high NDR while gross retention falls. A large expansion from a handful of accounts can hide broad downgrades elsewhere.
Letting one large customer dominate the result. Report NDR for the whole cohort and for cohorts segmented by customer size, industry, package, and tenure.
Treating a price increase as product-led expansion. Break out expansion from added usage, added products, added seats, and price realization.
Changing the calculation without maintaining comparability. HubSpot disclosed an adjustment to its calculation for 2025 and restated prior-year comparison figures under the new approach, which is the standard leaders should follow.
A board should be able to ask one follow-up question after every NDR report: Where did the change come from? If management cannot answer in terms of churn, contraction, product expansion, price realization, and customer segments, the metric is reporting history rather than guiding action.
High NDR rarely comes from a heroic save at renewal. It comes from decisions made much earlier: which customers the company targets, what promise it makes at sale, how rapidly customers reach value, and whether the next purchase feels logical rather than forced.
CrowdStrike's disclosures show the commercial value of that discipline. Its fiscal 2026 dollar-based net retention rate was 115%, and the company reported that 43% of subscription customers had adopted five or more modules as of January 31, 2026. By July 31, 2026, 51% of customers had adopted six or more modules. The pattern is not simply retention; it is expansion through a growing platform footprint.
Monetizely's view is that pricing leaders should manage NDR as a designed outcome. Product teams must create expansion paths. Sales teams must sell customers into packages they can use. Finance must make the economics visible. Customer-success teams must know which adoption milestones predict a renewal at the same or higher level.
The following actions turn that principle into operating practice:
Set the annual growth plan from cohort economics first. Build the ARR plan by starting with expected NDR by segment, then calculate the new ARR required to close the remaining gap.
Assign joint accountability to product, sales, and finance. The CRO alone cannot own a metric shaped by product adoption, package design, discounting, and billing.
Report NDR by customer starting point. Track each cohort by first package, initial spend band, acquisition channel, and tenure so leadership can see where expansion begins and where contraction starts.
Fund roadmap priorities with retention evidence. Require every major product investment to state which customer cohort it should retain or expand, the expected ARR mechanism, and the time required for that effect to appear.
Make investor communication match operating reality. When reporting NDR externally, explain the underlying drivers and definition changes with the same rigor used in internal planning.
Footnotes
- https://www.amazon.com/Monetizing-Agentic-AI-Handbook-Transformation/dp/B0H7Z13VKJ/
- HubSpot, Form 10-K for the year ended December 31, 2025, filed February 11, 2026. (ir.hubspot.com)
- Snowflake, first-quarter fiscal 2027 results for the quarter ended April 30, 2026. (investors.snowflake.com)
- Datadog, second-quarter 2026 financial statement package for the quarter ended June 30, 2026. (investors.datadoghq.com)
- CrowdStrike, Form 10-K for the fiscal year ended January 31, 2026. (ir.crowdstrike.com)