Does Your Curve Flatten? Cohort Retention and Net Revenue Retention for Qatar Startups

Signing up a thousand customers says little on its own. The question investors ask next is how many of them are still there six months later. Retention is the clearest test of whether a product delivers lasting value, and in Qatar it carries extra weight. In a market of this size, the pool of potential customers in any segment is finite: every customer who leaves is one you may not be able to replace with another just like them. This guide shows founders how to build a cohort table, read the shape of their retention curve, and calculate net revenue retention, so they know what to fix before spending more on growth.

Start with the right definition and time window

Retention means different things in different businesses. Sequoia Capital's data science team notes that many consumer businesses define retention as the percentage of users who take a given action in a period, while subscription and SaaS companies usually measure it in dollars spent within a period after a customer's first spend. Pick the action that represents real value: placing an order, sending an invoice, completing a booking. Simply logging in rarely qualifies.

Then match the time window to how often a customer naturally needs your product. Sequoia warns that measuring monthly retention for a product people naturally use weekly gives a distorted picture. A grocery delivery app might be measured weekly, a payroll tool monthly, and an annual-contract B2B product by renewal.

Build a cohort table from a spreadsheet

You do not need analytics software to begin. A cohort table, which Sequoia calls a triangle retention chart, groups customers by the period in which they started, then tracks what share of each group is still active in each later period.

• Export a list of customers with their start date and every date they performed the key action (or paid).
• Assign each customer to a cohort by start month (or week).
• For each cohort, count how many were active in period 0, 1, 2 and so on, and divide by the cohort's starting size.
• Lay the cohorts out as rows and the periods as columns. Newer cohorts have fewer columns, which creates the triangle.

Read the table in three directions. Reading across a row shows how one cohort behaves over time. Reading down a column shows whether newer cohorts retain better than older ones at the same age, which tells you if product changes are working. Diagonal patterns usually point to an event affecting all cohorts at once, such as a price change, an outage or a seasonal effect. In Qatar, Ramadan, the summer travel months and the winter tourism season can all create visible diagonals, so annotate them before drawing conclusions.

Read the shape of the curve

Plot each cohort's retention over time, and the curve will usually take one of three shapes that Sequoia describes:

• Declining. The curve keeps falling towards zero. Sequoia's view is that this happens when a product has not achieved product-market fit, and that pouring growth spending into the top of the funnel without fixing weak retention leads to a "leaky bucket" outcome.
• Flattening. The curve drops early, then levels off. This is the sign that a core group of users finds lasting value, and the higher the level at which it flattens, the healthier the product.
• Smiling. The curve rises again as users return or expand their usage. This is rare, and Sequoia associates it with truly exceptional products.

The practical rule follows from the shapes. If your curves decline, Sequoia's advice is to flatten each cohort's curve first, focusing on changing the product to find a value proposition for a core set of users. Growth spending comes later. This connects directly to the 40% product-market fit test: a strong survey result should show up as a flattening curve.

Find your magic moment

Once the curve flattens, compare the customers who stayed with those who left. What did retained customers do in their first days or weeks that churned customers did not? It might be connecting a second team member, placing a second order within a fortnight, or integrating a payment method. That behaviour is often called the "magic moment" or activation point. This is general product practice rather than a fixed formula: test the link by steering new users towards that action and checking whether later cohorts retain better.

Net revenue retention: the investor's version

For businesses with recurring revenue, user counts are only half the story. Net revenue retention (NRR) compares the revenue a group of customers generates today with what the same group generated a year ago. Lenny Rachitsky defines it as the current month's MRR from a group of customers divided by that same group's MRR one year earlier. Expansion (upgrades, more seats, more orders) pushes NRR up; churn and downgrades pull it down. Above 100%, your existing customers grow revenue even before you add new ones.

What counts as good? The most cited benchmarks come from a 2020 survey by Rachitsky and Casey Winters of 20 experienced growth practitioners. Treat them as rough guides from 2020, not Qatar-specific standards:

• Six-month user retention: around 25% (good) to 45% (great) for consumer social; around 30% to 50% for consumer transactional; around 40% to 70% for consumer SaaS; around 60% to 80% for SMB and mid-market SaaS.
• 12-month net revenue retention: around 100% (good) to 120% (great) for bottom-up SaaS, and around 90% to 110% for land-and-expand SMB and mid-market SaaS.

The same article notes that low retention can be acceptable in some cases, such as very early-stage products or businesses with very low acquisition costs. Context matters, and so does the trend across cohorts.

Why this matters for Qatar's ecosystem

Startup Qatar and the Third National Development Strategy aim to grow startups that attract private capital and scale from Doha into the region. Investors backing those companies will look for evidence that customers stay and spend more. A flattening curve and healthy NRR are that evidence. Retention also feeds directly into unit economics: lifetime value depends on how long customers stay, so it shapes the CAC payback and LTV:CAC ratios investors ask for, and your runway maths. In a small home market, strong retention also gives you a stable base while you expand across the GCC.

What to do next

• Define your key action and the natural usage frequency of your product this week.
• Build a monthly (or weekly) cohort table from your customer data, and annotate seasonal events.
• Plot the curves and decide honestly: declining, flattening or smiling.
• If declining, pause scale-up spending and work on the product for your best-retained segment.
• Find your magic moment and redesign onboarding to reach it faster.
• Calculate NRR if you have recurring revenue, and include cohort charts in your investor updates.

Sources

• Retention – Sequoia Capital, Data Science Team
• What is good retention – Lenny Rachitsky, Lenny's Newsletter (June 2020)

Photo: ThisIsEngineering / Pexels