Customer retention in B2B SaaS: benchmarks, strategies and how to measure it
Customer retention is the share of customers, or customer revenue, that stays with you across a period. In SaaS Capital's 2025 survey of more than 1,000 private B2B SaaS companies, median gross revenue retention was 91% and median net revenue retention 101%. What separates teams is not whether they track the number but whether they know which actions move it.
What this guide covers
What customer retention is, and how it differs from churn and from loyalty
How to calculate retention rate, GRR and NRR
What a good retention rate actually is, from primary benchmark data
What retention costs, and which widely repeated claims have no traceable source
Six retention strategies built for B2B, not retail
How to build a retention program and report it to a board
What software supports it, and where AI helps
Share
What is customer retention?
Customer retention definition
Customer retention is the practice of keeping existing customers, and the metric that measures how well you do it.
In B2B software it is measured two ways. Logo retention counts accounts: how many of the customers you had at the start of a period are still customers at the end. Revenue retention counts money: how much of the revenue from those accounts you still hold. The two diverge whenever your customers are different sizes, which is almost always.
Customer retention vs churn
They are two views of the same event, and the arithmetic differs depending on what you are counting.
Logo retention and logo churn are complementary: 10% logo churn means 90% logo retention. Revenue retention needs more precision. GRR measures what remains after churn and contraction without expansion, while NRR adds expansion and can therefore exceed 100% even when some revenue is lost.
The practical difference is where each one points you. Churn analysis asks why accounts left and what the warning signs were. Retention work asks what keeps accounts and how to do more of it.
See customer churn and retention for churn causes, signals and the churn formulas.
Customer retention vs customer loyalty
Retention is behavioural and loyalty is attitudinal.
A customer can be retained without being loyal: locked into a contract, mid-way through an integration, or without a viable alternative. A customer can be loyal without being retained, if budget was cut or the buying team changed.
Retention data describes what happened. Loyalty measures can add forward-looking context that retention data alone does not provide.
Why customer retention matters in B2B SaaS
In a subscription business, the revenue from an existing customer continues after the first sale, and the share of new ARR coming from existing customers grows with scale.
The 2024 KeyBanc Capital Markets and Sapphire Ventures Private SaaS Company Survey asked 52 companies what proportion of new ARR came from existing-customer expansion:
2023 ARR | Expansion share of new ARR | New-logo share |
Under $10M | 30% | 70% |
10M–25M | 48% | 52% |
25M–50M | 49% | 51% |
Over $50M | 50% | 50% |
Overall | 45% | 55% |
At that scale, a weak expansion motion means relying more heavily on new-logo acquisition than the larger companies in KeyBanc's survey, which generated about half of new ARR from existing-customer expansion.
How to calculate customer retention rate
Customer retention rate formula
Customer retention rate = ((Customers at end of period − New customers acquired) / Customers at start of period) × 100
Subtracting new customers is essential; skipping it turns a retention metric into a growth metric.
Retention rate calculation example
An illustrative B2B software company starts the quarter with 240 customers. During the quarter it wins 35 new ones and loses 18. It ends with 257.
CRR = ((257 − 35) / 240) × 100 = 92.5%
Without subtracting the 35 new customers, the same figures produce 107%, which would describe a company that lost nobody.
Logo retention, GRR and NRR compared
Logo retention | Gross revenue retention | Net revenue retention | |
Counts | Accounts | Revenue, excluding expansion | Revenue, including expansion |
Ceiling | 100% | 100% | None |
Answers | Did we keep the relationships? | Did the revenue base hold? | Did the revenue base grow? |
Blind to | Contract size | Expansion | Erosion underneath expansion |
Gross revenue retention = (Starting ARR − Churned ARR − Contraction ARR) / Starting ARR × 100
Net revenue retention = (Starting ARR − Churned ARR − Contraction ARR + Expansion ARR) / Starting ARR × 100
Both measure the same cohort: customers who were already customers at the start of the period. New customers acquired during the period are excluded from both. The same formulas work on MRR for businesses that measure monthly.
A company can retain 95% of logos and 85% of revenue if the accounts that left were its largest. The reverse happens too. All three belong in a retention report.
How to choose your retention measurement window
Match the reporting window to both the contract cadence and the number of accounts exposed to a retention event in that window.
In businesses with staggered annual renewals, monthly retention can still be operationally useful, because a meaningful number of accounts reaches a decision every month. Where renewal opportunities are sparse or seasonal, quarterly or rolling twelve-month views are more stable.
With monthly contracts the opposite problem applies: annual retention hides churn happening continuously underneath it.
Why retention is high but revenue is falling
This pattern is common enough to name: logo retention looks healthy, revenue retention does not.
Three common situations can produce it:
Accounts renew at lower values. Seat reductions and downgrades hold the logo and lose the revenue.
Churn is concentrated in large accounts. A handful of departures outweighs many retained smaller customers.
Churn and contraction together exceed expansion. The base is eroding faster than it grows.
KeyBanc's data shows the balance between outright churn and downsell, not whether churn was concentrated in larger accounts. Among 48 respondents, 63% of lost ARR came from outright churn and 37% from downsell, and the mix varied by size: outright churn accounted for 75% of lost ARR below $10 million ARR, 53% between $10 million and $25 million, 62% between $25 million and $50 million, and 63% above $50 million.
Retention curves and cohort retention analysis
A cohort view groups customers by a shared starting point and tracks each group separately over time.
In B2B the useful cohorts are rarely calendar months. Group by first-renewal quarter, by implementation start date, by segment or by ACV, and ask whether each successive cohort retains more logos and more revenue at comparable lifecycle points.
That answers a question the aggregate cannot: is retention improving, or does it look better because recent cohorts have not yet reached their first renewal?
What is a good customer retention rate?
There is no single number, and the most common mistake is taking one from a table built on a different kind of business.
B2B SaaS retention benchmarks by contract value
SaaS Capital's 2025 survey covers more than 1,000 private B2B SaaS companies. Retention is measured by comparing December 2024 recurring revenue from customers who were already customers in December 2023 against those customers' December 2023 recurring revenue. Companies below $1 million ARR are excluded from the ACV breakdown.
Median NRR across the full sample was 101% and median GRR 91%.
Average contract value | Median NRR | Median GRR |
Under $12K | 98% | 90% |
12K–25K | 103% | 91% |
25K–50K | 102% | 91% |
50K–100K | 104% | 90% |
100K–250K | 102% | 91% |
Over $250K | 106% | 95% |
Source: SaaS Capital, 2025 B2B SaaS Retention Benchmarks. More than 1,000 private B2B SaaS companies; retention measured December 2023 to December 2024; survey conducted Q1 2025. NRR includes upsell, cross-sell and price expansion; GRR removes them.
The lowest-ACV cohort has the weakest NRR and the highest-ACV cohort the strongest, and retention does not rise through every intermediate band. Contract value is a useful segmentation variable, not a predictor.
An independent cross-check
Benchmarkit's 2024 report is based on 936 private B2B SaaS companies reporting CY2023 data, with different ACV boundaries.
Average contract value | Median NRR |
Under $5K | 95% |
5K–10K | 99% |
10K–25K | 106% |
25K–50K | 101% |
50K–100K | 103% |
Over $100K | 103% |
Source: Benchmarkit, 2024 B2B SaaS Performance Metrics Benchmark Report. 936 private B2B SaaS companies, CY2023. Participants supplied financial and KPI data; percentile benchmarks calculated with statistical-significance and outlier procedures applied.
Median CY2023 GRR was 89%, with the 25th percentile at 79% and the 75th at 95%.
Two independent surveys, different samples and different ACV boundaries, GRR medians of 91% and 89%. Treat them as separate readings instead of averaging them, and use SaaS Capital as the primary ACV citation with Benchmarkit as corroboration.
Why industry benchmark tables mislead in B2B
Search for a retention benchmark and most of what comes back is a table of rates by industry: retail at one number, media at another, hospitality at a third.
Those tables are usually built on consumer and e-commerce data, where a customer is an individual making repeat purchases. In B2B software a customer is an organisation on an annual or multi-year contract, and the behaviour behind the number has little in common.
Within B2B, contract value, contract length and segment appear more useful than vertical. A $250,000 enterprise contract in media and a $250,000 enterprise contract in logistics have more in common with each other than either has with a $500 self-serve subscription in the same vertical.
Why a single benchmark number misleads
Even inside one ACV band, the spread is wide.
SaaS Capital reports that companies with contracts under $12,000 range from 90% NRR at the 25th percentile to 106% at the 75th. Companies above $250,000 range from 102% to 110%. The median of 98% in the lowest band describes a middle point that many companies in it are nowhere near.
A benchmark tells you where you sit in a distribution. It does not tell you what is achievable for your product, your segment or your contract structure.
How retention relates to growth
SaaS Capital's 2025 data shows a strong association between net revenue retention and growth.
NRR | Median growth |
Below 90% | 15% |
90%–100% | 16% |
100%–110% | 21% |
110%–120% | 30% |
120%–130% | 38% |
Above 130% | 50% |
Median growth for the broader survey population was 24%.
Two cautions on reading that. The study is observational, and NRR contains expansion revenue, so part of the association is mathematical as well as economic. SaaS Capital separately reports that GRR showed much less apparent relationship with growth than NRR did.
The defensible formulation is that higher NRR was strongly associated with higher revenue growth in this sample, not that raising NRR causes growth.
The economics of customer retention
This is the area where marketing folklore most badly outruns the underlying evidence.
What the "five times cheaper" claim actually rests on
The claim that acquiring a new customer costs five times more than retaining one cannot be tied to a publicly inspectable primary study with a reproducible methodology.
The most thorough source-tracing work is Loyalty Myths by Keiningham, Vavra, Aksoy and Wallard. The authors investigated the origin and reported they could not determine one; the earliest sources they identified attributed it to work by TARP in the late 1980s. The claim gained authority when it appeared in the 1990 Harvard Business Review article "The Profitable Art of Service Recovery".
That does not make the underlying idea wrong. It makes the number unusable as evidence.
What contemporary data shows instead
Benchmarkit's CY2023 data gives a traceable comparison. The median new-customer CAC ratio was $1.76 of sales and marketing expense for every $1 of ARR acquired from a new-name customer, after removal of statistical outliers. Leaving outliers in produces a median of $1.85. The median expansion CAC ratio was approximately $1.00 per $1 of expansion ARR.
So generating a dollar of new-logo ARR required about 76% more go-to-market spend than generating a dollar of expansion ARR in that population.
This compares expansion with new acquisition, not retention with acquisition. A true retention cost would include renewal, support, success and account management activity per account retained, which Benchmarkit does not claim to measure. It is the closest defensible contemporary figure, and it should be described for what it measures.
How to calculate customer retention cost
There is no universally standardised CRC denominator, so the definition matters more than the arithmetic.
Customer retention cost = total direct cost of the retention motion during the period
And if you want a per-account figure:
Retention cost per active account = total retention cost / average active accounts in the period
Retention spend typically includes customer success salaries and tooling, support costs attributable to existing accounts, renewal and account management time, and programmes run specifically to retain.
Two decisions determine whether the number is comparable over time: whether support is counted in full or apportioned, and whether expansion-focused work sits in retention cost or in sales cost. Define the cost boundary and the denominator before comparing periods, because different vendors and benchmarks use different conventions.
Why expansion matters more as you scale
As SaaS companies scale, the post-sale commercial motion carries more of the growth load. In KeyBanc's sample, existing-customer expansion contributed about half of new ARR at larger companies.
Benchmarkit found the same direction independently: expansion ARR represented 35% of growth ARR at the median in 2023, and companies approaching $50 million ARR and above were close to obtaining half of their growth ARR from existing-customer expansion.
Expansion and retention are different activities with different costs. What the data shows is that the post-sale motion as a whole becomes a larger share of growth with scale.
See customer expansion.
The "5% retention raises profits 25–95%" claim
This one has a real research lineage, and the modern shorthand strips away its scope.
The foundational work is Reichheld and Sasser's "Zero Defections: Quality Comes to Services", Harvard Business Review, 1990. The exact 25–95% formulation appears in Reichheld and Schefter's 2000 article "E-Loyalty: Your Secret Weapon on the Web", which estimated that retaining 5% more customers could raise profits 25% to 95%.
The context matters. That passage is explicitly about early internet businesses, where acquisition costs were unusually high and customers could remain unprofitable for two or three years. It is a legitimate finding about a different kind of business, and it is not a universal SaaS coefficient.
Customer retention strategies for B2B SaaS
Why B2B retention needs a different playbook
Most retention advice available online is written for consumer and e-commerce businesses: loyalty programmes, points and rewards, omnichannel support, win-back email sequences, community building.
Many of those tactics do not transfer directly to B2B, where the user, the buyer and the budget owner may be different people and renewal decisions are usually contract-based. Three structural differences explain why.
The buyer and the user are often different people. A loyalty programme that delights daily users does not by itself give the budget owner a defensible value case.
Renewal decisions happen on a contract cadence. Many B2B SaaS customers decide once a year or on a multi-year cycle, so the work happens months before the moment and the tactics are about evidence more than persuasion.
Discounting can undercut the value case. A price concession to save a renewal can set a precedent that shapes future renewal conversations.
Some consumer tactics do adapt. Customer education, community and personalised engagement can all work in B2B when they are built around the account instead of the individual. What follows is built for the B2B case.
Strategy 1: Make first value happen faster
The conditions that influence retention are established earlier than the renewal conversation. An account that has not reached a meaningful outcome in its first quarter has little to defend at renewal.
Define what first value means for each segment, state it as something observable, and build the implementation plan around reaching it instead of around completing configuration.
See customer onboarding.
Strategy 2: Spread adoption beyond the champion
A common structural risk in a B2B account is a deployment that lives with one team or one person.
Broader adoption reduces the degree to which retention depends on a single champion, and it gives the relationship a footing that survives personnel change.
See customer adoption.
Strategy 3: Build the value case across the term
Renewal conversations are stronger when the evidence is accumulated across the term, instead of being reconstructed in the final six weeks.
That means recording what was committed at the sale, tracking delivery against it, and capturing outcomes as they happen.
See customer renewals.
Strategy 4: Make risk visible before it is a decision
A retention programme that responds to notice periods is working on accounts where the options are already narrower. The accounts worth working on are the ones where nothing has happened yet.
That requires a scoring model that fires on defined conditions rather than on whoever happened to look, and one validated against accounts that actually churned.
Strategy 5: Structure contract terms around the value cycle
Contract length is associated with large differences in observed churn. KeyBanc's 2024 survey, covering 52 companies on this metric, reports median churn of 14% for month-to-month contracts, 10% for one-year terms, 6% for two-year terms and 3% for terms of three years or longer, against 7% overall.
Read that as association, not mechanism. Customers willing to sign multi-year contracts are likely different from monthly customers to begin with, and contract value, product criticality and purchasing process may all contribute.
A longer term can also defer a churn event without improving the underlying relationship, and it reduces the frequency of formal renewal events, which can defer visibility into churn risk. The useful question is whether the term matches the time the customer needs to realise value.
Strategy 6: Run retention as a repeatable play
Retention work that depends on who noticed what produces inconsistent coverage, and the accounts nobody is watching are the ones the programme exists for.
A play with a defined trigger, a named owner and an exit condition makes it possible to apply the same standard to every account.
See customer success playbooks.
Retention is inherently cross-lifecycle, which is what makes it hard to operate. The promise is made in sales, the value is delivered in implementation, adoption shows up in usage, friction appears in support, and the commercial decision happens at renewal. Those signals often live across separate systems, which means the context used to assess retention risk may be disconnected from the record that acts on it. That is the logic behind what Planhat calls its One Commercial Brain: the sale, delivery, usage, support context and renewal held against the same customer, so a Workflow or Agent can pick up a retention signal where it appears. Trend Micro describes the before state: a customer success manager hunting through screens and tables in a sales system to work out which customers needed help.
How to build a customer retention program
Step 1: Define what retention means for your segments
An enterprise account with a services engagement and a self-serve subscription are both retained or lost, and almost nothing else about them is comparable.
Decide per segment what you are measuring, over what window, and what counts as a loss. A downgrade is a partial loss in revenue terms and a full retention in logo terms, and the programme has to say which it tracks.
Step 2: Choose the metrics you will hold yourself to
Two or three as the accountable set, even if you report more. In B2B SaaS, a practical accountable set is gross revenue retention as the health measure, net revenue retention as the growth measure, and logo retention as the relationship measure.
Reporting a wider metric set is fine. What causes problems is nobody being able to say which one they are accountable for.
Step 3: Assign ownership
Retention sits across customer success, account management, support and product, which is why it frequently sits nowhere.
What needs a named owner is not the outcome but the programme: who reviews the numbers, who decides what plays run, and who is accountable when a segment drifts.
Step 4: Set the review cadence
Monthly for execution, quarterly for the model itself.
A quarterly review should ask whether the scoring model still predicts, whether the segments still describe the business, and whether the plays are producing different outcomes from doing nothing.
Customer retention metrics and KPIs
Metric | What it tells you | Who reads it |
Gross revenue retention | Whether the base is holding | Board, leadership |
Net revenue retention | Whether the base is growing | Board, finance |
Logo retention rate | Whether relationships are holding | Customer success |
Customer retention cost | What the retention motion costs | Finance |
Retention by cohort | Whether retention is improving over time | Operations |
Time to first value | Operational signal to monitor alongside first-year retention | Onboarding |
Adoption breadth | Operational signal to monitor alongside renewal outcomes | Customer success |
Three of these carry accountability. The rest provide context.
Leading and lagging retention signals
Retention rate is a lagging metric. By the time it moves, the decisions behind it were made months earlier.
Common operational signals include time to first value, adoption breadth, stakeholder engagement and support friction. Their predictive value should be tested against your own renewal and churn outcomes, which is the same discipline that applies to a health score.
How to report retention to the board
Four things make a retention report useful at board level.
Report distribution, not averages. An average retention figure hides the segment that is failing. Retention by ACV band or by segment shows where the problem is.
Separate gross from net. A net figure above 100% can mask gross erosion underneath it. Showing both is the only way to make that visible.
Connect it to growth. SaaS Capital's data shows median growth rising from 15% among companies below 90% NRR to 50% above 130%. Framing retention as a growth input instead of a defensive metric changes the conversation.
Show what moved. A number without the accounts behind it invites questions nobody can answer in the room.
Customer retention software
AI for customer retention
Where AI helps with retention
Automation and AI contribute differently to retention work.
Coverage at scale. Automation or agents can evaluate every account consistently against defined criteria. AI becomes useful when the signal requires interpretation, not just reading a predefined field.
Signals in unstructured text. A concern raised on a call or in a support thread is a retention signal that structured fields miss.
Preparation. Account briefs, value summaries and renewal cases drafted from the record instead of assembled by hand.
Whether risk is rule-based or model-generated, the underlying signals should remain visible and the scoring should be tested against what actually happened to those accounts.
What still needs a person
The conversation, the judgement about timing, and the decision about what to offer when an account is genuinely at risk.
A model can rank a book by retention risk and show the signals behind each position. What it cannot reliably do is weigh those against what you know about a specific account's internal politics.
Sources and methodology
Three primary benchmark sources carry the modern B2B SaaS operating benchmarks on this page. None is published by a customer success, CRM, churn or analytics software vendor.
Source | Sample | Period | Methodology |
1,000+ private B2B SaaS companies | Retention measured Dec 2023 – Dec 2024; survey Q1 2025 | Annual company survey; same-customer recurring-revenue cohort compared year over year, expansion included in NRR and removed from GRR | |
Benchmarkit — 2024 B2B SaaS Performance Metrics Benchmark Report | 936 private B2B SaaS companies | CY2023 | Participants supplied financial and KPI data; percentile benchmarks with statistical-significance and outlier procedures |
KeyBanc Capital Markets + Sapphire Ventures — 2024 Private SaaS Company Survey | 104 companies; n=55 retention metrics, n=52 expansion share and contract-length churn, n=48 churn versus downsell composition | 2023 actuals, 2024 estimates | Annual survey of senior executives at privately held global SaaS businesses |
Two historical claims are discussed on this page for provenance, not as benchmarks. The "five times cheaper" acquisition claim is traced through Loyalty Myths, Myth #8, and the 25–95% profit figure comes from Reichheld and Schefter's 2000 article E-Loyalty: Your Secret Weapon on the Web.
We did not find a recent public primary dataset, meeting the same methodology criteria, that quantifies the causes of realised B2B SaaS churn. Survey evidence on software switching points to cost disappointment and better alternatives as triggers, but it measures intended changes, not observed subscription churn. Where this page describes churn causes, it describes patterns rather than measured proportions.
Customer retention FAQs
What is customer retention?
Keeping existing customers, and the metric that measures how well you do it. In B2B software it is measured two ways: logo retention counts accounts, revenue retention counts money, and the two diverge whenever customers are different sizes.
What is a good customer retention rate?
SaaS Capital's 2025 benchmark reports median NRR of 101% and median GRR of 91%; Benchmarkit's CY2023 benchmark puts median GRR at 89%. The useful comparison is against companies with a similar contract value and segment, since the spread within any band is wide.
Is 90% customer retention good?
For gross revenue retention in B2B software, 90% sits close to the median in both surveys: 91% in SaaS Capital's data and 89% in Benchmarkit's. For logo retention in a high-volume self-serve business, 90% would be strong. The number means different things depending on what you are counting.
How should I interpret a retention rate such as 80% or 44%?
Both only mean something once you know the population, the window and the definition. In revenue terms, 80% gross revenue retention sits near Benchmarkit's 25th percentile of 79%, meaning roughly three quarters of surveyed companies retained more. A 44% figure would be far below any B2B benchmark here, and could be normal for a consumer app measured on 30-day return behaviour.
How do you calculate customer retention rate?
Subtract new customers acquired during the period from customers at the end, divide by customers at the start, multiply by 100. Omitting the subtraction turns it into a growth metric.
Should I measure retention monthly, quarterly or annually?
Match the window to the contract cadence and to how many accounts reach a retention event inside it. With staggered annual renewals, monthly retention can be operationally useful. Where renewals are sparse or seasonal, quarterly or rolling twelve-month views are more stable.
What if retention is high but revenue is falling?
Common causes include accounts renewing at lower values, churn concentrated in large accounts, or churn and contraction together exceeding expansion. KeyBanc's data shows how lost ARR splits by mechanism: among 48 respondents, 63% came from outright churn and 37% from downsell.
What is the KPI for customer retention?
For B2B SaaS, GRR, NRR and logo retention form a useful core set: one shows gross erosion, one includes expansion, and one tracks accounts. Other metrics provide context rather than accountability.
What is a customer retention program?
A defined set of plays, owners, metrics and review cadences aimed at keeping existing customers. What separates a programme from an intention is that the plays have triggers and the metrics have an owner.
How do you calculate customer retention cost?
There is no universally standardised denominator. Total retention cost is the direct cost of the retention motion in a period; a per-account figure divides that by average active accounts. The two decisions that determine comparability are whether support costs are counted in full or apportioned, and whether expansion work sits in retention cost or sales cost.
What is the difference between acquisition and retention?
Acquisition brings in customers who have not bought before. Retention keeps the ones who have. Benchmarkit's CY2023 data puts the median cost of $1 of new-logo ARR at $1.76, against roughly $1.00 for $1 of expansion ARR, which is the closest traceable comparison available. It measures expansion, not retention.
What is the difference between retention and churn?
Logo retention and logo churn are complementary and sum to 100%. In revenue terms it depends which metric you use: GRR and gross revenue churn are complementary, while NRR adds expansion and can exceed 100% even when revenue is being lost underneath.
Does it really cost five times more to acquire a customer than to retain one?
The claim cannot be tied to a publicly inspectable primary study. The authors of Loyalty Myths investigated the origin and reported they could not determine one; the earliest attribution they found was to TARP work in the late 1980s. Benchmarkit's $1.76 against $1.00 comparison is the defensible contemporary figure.
What are the best customer retention strategies for B2B SaaS?
Six that account for the structure of B2B: reach first value faster, spread adoption beyond the champion, build the value case across the term, make risk visible before a decision point, structure contract terms around the value cycle, and run retention as a repeatable play. Many consumer tactics do not transfer directly, because the user, buyer and budget owner may be different people.
What software do I need for customer retention?
Look for software that combines customer context with a way to identify risk, coordinate a response and measure the result. The main distinction between options is whether a tool only reports retention risk or can operationalise it.