Retention: What to Measure and How to Do It

“Acquiring a new customer is 5 to 25 times more expensive than retaining an existing one."

One of those age-old stats nearly everyone hears at some point in business school, and depending on the professor, probably more than once. Regardless of its origins (or its accuracy, frankly), it remains true: retention is cheaper and more profitable than acquisition. 

This makes understanding and optimizing customer retention a crucial focus for any business, especially startups striving for sustainable growth. Retention, and its inverse, churn, measures how likely a customer is to remain a customer and is critical to a startup’s ability to scale. Beyond the financial impact and its role in improving unit economics, it is a strong indicator of how much users actually value the product. As the book The Mom Test puts it, the best way to know if users truly need your product is if they are willing to open their cheque books and part with their own cash. High retention and low churn are strong signals you are solving a real need and building a durable business.

So how do you quantify it?

If you’ve spent time in the startup ecosystem (or on our page, for that matter), you know there’s no shortage of venture metrics. Here are a few of the most important when it comes to retention.

Churn Rate: Churn is the cornerstone of retention analysis. It measures the rate at which customers cease doing business with you over a specified period - typically monthly, quarterly, or annually. There are two primary ways to assess it:

Customer Churn: The percentage of customers lost in a given period of time (typically annually or monthly).

Revenue Churn: The percentage of recurring revenue lost (e.g., MRR or ARR).

Like any metric in this article, analyzing churn is only productive with context. Bessemer notes that top SaaS companies keep customer churn below 7% and revenue churn below 5%. Monthly customer churn ranges from 3–7% for SMBs, while enterprise SaaS targets ~1% due to stickier contracts and deeper integration.

Net Dollar Retention (NDR): A holistic metric that captures how revenue from existing customers grows or shrinks over time. It accounts for upgrades, downgrades, and churn.

Formula: (Starting MRR + Expansion MRR – Contraction MRR – Churned MRR) / Starting MRR

An NDR greater than 100% means your existing customers are growing their spend, a necessary signal for scalability. While anything over 100% reflects growth from your current base, top-performing SaaS companies will target NDRs above 120%, especially at growth stage. According to Benchmarkit, the median NDR for SaaS companies in 2023 was 103%, with the 75th percentile at 111%. At IPO, the average NDR was around 110%. Also, OpenView Partners reports that NDR typically increases with scale, from a median of 100% at $1–3M ARR, to 105% for $3–15M ARR, 111% at $15–30M, and levelling off around 110% for companies between $30–100M ARR.

Gross Dollar Retention (GDR): Similar to NDR, but excludes expansion. GDR tells you how much of your existing revenue you’re able to retain without upselling. It is important because it strips out expansion revenue and focuses purely on how much recurring revenue a company retains solely from existing customers. While NDR includes expansion and can look healthy, it can sometimes mask deeper churn issues, especially if a few high-spending customers are rapidly expanding and offsetting losses elsewhere. GDR helps uncover whether your core customer base is stable, independent of any outsized growth from a handful of customers.

Formula: (Starting MRR – Contraction MRR – Churned MRR) / Starting MRR

This is a “true retention” number and. 95%-100% is generally accepted to be the goal for software businesses.

LTV/CAC Ratio: Your customer lifetime value / customer acquisition cost ratio (LTV/CAC) tells you how valuable a customer is relative to how much it cost to acquire them. Retention is a crucial factor in this ratio, since longer customer lifespans drive higher LTV. This is a massively important metric that investors will always ask for. So important, in fact, that A16z stated improving your LTV to CAC from 2x to 3x can nearly triple your valuation.

Formula: LTV/CAC Ratio = Lifetime Value (LTV) / Customer Acquisition Cost (CAC)

Lifetime Value (LTV) = (Revenue Per Customer × Gross Margin) / Churn Rate

Customer Acquisition Cost (CAC) = Sales and Marketing (S&M) Expense / Number of New Customers Acquired

Rule of thumb: healthy businesses aim for an LTV/CAC > 3x.

CAC Payback Period: The CAC Payback Period is how many months it takes for a company to earn back the cost of acquiring a customer through the gross profit that customer generates. It reflects how quickly your sales and marketing investments return capital, which is critical for cash flow and growth planning.

Formula: CAC Payback Period = Sales and Marketing Expense (S&M) / (New MRR × Gross Margin)

As a general rule, most viable SaaS startups have a CAC payback period of fewer than 12 months. A shorter payback period indicates a more efficient growth engine. OpenView Partners considers anything under 12 months strong, with 6 to 12 months ideal for SMB. In enterprise SaaS, 18 to 24 months is common due to longer sales cycles and higher deal sizes. As startups mature, these figures will shorten as sales and go-to-market engines dial in and become more efficient.

Active User Rate (DAU/MAU): Especially useful for products with high-frequency usage and companies building with product led growth in mind. Tracking the percentage of users who return daily, weekly, or monthly helps you understand stickiness and can be a leading indicator of long-term retention. 

A good DAU/MAU ratio is around 20-30%.

Improving Retention

So how do you actually improve retention? There’s no simple answer, but it starts with getting close to your users. Understand their pain points, how they use the product, and where the experience breaks down. Too many founders build in a vacuum and overlook the actual people, the humans using the product they’re building.

Cohort analysis is a critical tool for doing this. Cohort analyses pinpoint where and why users are churning, how different groups behave over time, and whether recent changes are driving improvement. Paired with customer conversations, it gives insight to both the what and the why.

Retention metrics show what is happening, but cohort analysis reveals why and to whom. It groups users by a shared trait, like signup month or acquisition channel, and tracks how long they stay engaged or generate revenue. When a VC asks for a cohort analysis, it is not just a box to tick. It is how they evaluate whether your growth engine is working. Are users sticking around beyond 90 days? Are newer cohorts improving? Are product changes compounding? Top-line growth might look good in a pitch, but cohort analysis shows if it is actually sustainable.

Cohort Analysis: The Foundation of a Retention Strategy

Cohort analyses groups users by a shared trait, in this case when they signed up. It can also be based on behavior, plan type, or acquisition channel, and tracks how those users behave over time. Rather than blending everyone into a single retention number, cohort analysis lets you isolate each wave of users and observe how they engage, drop off, or grow. It provides the kind of insight that simple averages overlook.

Let’s say 500 users signed up in January. You track how many are still active in February, March, and so on. If 300 stick around by March, that’s a 60% retention rate for the January cohort. Do the same for February, March, and April signups, and you start to get a retention matrix that tells you whether your product is genuinely improving or if you're just acquiring more users who behave the same (or worse) as before. The same can be done with MRR to get an idea of if existing customers are expanding or contracting.

Improving retention takes more than building new features. It requires understanding why customers stay. Cohort analysis reveals patterns that averages miss. It shows whether a product change actually improved retention or if growth is just coming from new users. It connects behavior to outcomes, showing whether actions like completing onboarding or inviting a teammate lead to longer engagement. When newer cohorts outperform older ones, it signals that your product is getting stronger. That is what gets investors’ attention.

While the default cohort is based on time, users grouped by signup month or quarter, more sophisticated operators go further. Behavioural cohorts like users who activated a key feature in week one, channel-based cohorts like paid versus organic, plan cohorts like basic versus enterprise, or even customer size cohorts like SMB versus mid-market versus enterprise clients, all offer sharper insights into what’s driving LTV and where churn risk hides. It’s not just about retention. It’s about segmentation, attribution, and strategy.

If you’re not using analytics tools, you can run a cohort analysis with nothing more than a spreadsheet and some SQL. Start by pulling a user-level table with signup dates and a clear measure of engagement or revenue. Group users by the month they joined, then track how many are still active or paying in each period that follows. This gives you a retention curve. Most curves drop off early, but what matters is where they level out. That flatline is your core retained user base. The goal is to raise that floor with every new cohort.

What makes this powerful is how tightly it ties into your operating roadmap. A single retention curve won’t help much unless you know why users are dropping off. That’s where pairing cohorts with behavioral data matters. Are users churning before they hit a key feature? Are those who invite teammates sticking around twice as long? Instead of trying to fix retention broadly, you’re suddenly armed with precision.

Most importantly, this kind of analysis gives you narrative control. Investors don’t just want to see data. They want a story. When you say, “We overhauled onboarding in March. Retention for April cohorts improved by 15 percent, and our CAC payback dropped by three months,” It signals that you know what drives your business and how to move the right levers.

Plugging the Leaky Bucket

While new customer acquisition and top-line revenue growth often take center stage in early-stage, high-growth startups, overlooking churn is a risky bet against long-term viability. The excitement of net-new revenue can easily mask the so-called “leaky bucket” of churn. Addressing retention early is essential. If left unchecked, churn builds over time and can severely limit a company’s ability to grow efficiently at scale. A leaky bucket will always leak, no matter how much water you add, and the larger the bucket, the more it leaks. Understanding retention gets you closer to real durability and to plugging the hole in the proverbial bucket.