pm benchmarks

Every PM asks “is this number good?” and nobody gives them a straight answer. Advisors say “it depends.” VCs say “what are your comps?” Your manager says “let’s benchmark against last quarter.” None of that helps when you are staring at 8% D30 retention and need to know whether to panic or celebrate.

This page gives you straight answers. Reference numbers by product type, stage, and market. India-adjusted, not Silicon Valley defaults copied from a16z blog posts.

Before you use these numbers

One warning, and I mean it: benchmarks are reference points, not targets.

Your context matters. A 12% D30 retention for a consumer social app is strong. The same 12% for a quick commerce app is a problem. If you blindly chase a benchmark number without understanding why it applies to your product, you will optimise for the wrong thing.

But here is the other truth: having no reference is worse than having an imperfect one. I have watched PMs present metrics to leadership with zero context for whether those numbers are healthy. They get asked “is that good?” and they shrug. That is a career-limiting moment.

Use these tables as a starting diagnosis. Then adjust for your specific product, market, and stage.

Retention benchmarks by product type

Retention is the metric that separates real products from leaky buckets. These numbers represent “good” — meaning top-quartile products in each category, not median. If you are at median, you have work to do.

How to read this table: If your consumer fintech app has 10% D30, you are in the “good” range. If you are at 5%, you have a retention problem. If you are at 15%, you are outperforming most of the market and should focus on growth, not retention optimisation.

The D1-to-D30 ratio tells you where the leak is

D30/D1 ratio matters more than any single number. Here is why:

• Ratio above 0.3: Your activation is working. Users who show up on D1 are finding value. Focus on frequency and engagement depth.
• Ratio between 0.15-0.3: You have a “week one wall.” Users activate but lose interest. Your onboarding gets them in; your core loop does not keep them.
• Ratio below 0.15: Your D1 is inflated by low-intent users (ad installs, referral fraud, COD browsers). Fix your acquisition quality before touching retention.

Activation benchmarks by product type

Activation is the moment a user gets value for the first time. Not “signed up.” Not “opened the app.” The moment they did the thing your product exists for.

India-specific activation killers

Three things destroy activation rates in India that do not show up in global benchmarks:

1. KYC friction. Any fintech product requiring Aadhaar/PAN verification loses 40-60% of users during onboarding. The best Indian fintech products (PhonePe, Razorpay) have invested years in reducing KYC steps. If you require full KYC before first value, your activation ceiling is 35%.

2. Permission overload. Indian users on Android are trained to deny permissions. If your app asks for camera, location, contacts, and storage on first launch, expect 20-30% drop-off before they even see your home screen. Ask for permissions at the point of need, not at install.

3. Language and literacy. English-only onboarding in a Hindi-belt market is a 15-25% activation penalty. Vernacular onboarding is not a nice-to-have — it is a conversion multiplier.

Growth rate expectations by stage

VCs have specific expectations for month-over-month growth at each funding stage. These are India-adjusted — Indian VCs are slightly more forgiving on revenue growth at pre-seed/seed but stricter on unit economics at Series A+.

A common mistake: Showing 30% MoM growth at seed stage and thinking it is impressive. If that growth is 100% paid acquisition with negative unit economics, Indian VCs will see through it. They have been burned by this pattern too many times (remember the 2015-2016 cash-burn era). Sustainable 15% beats unsustainable 30%.

Conversion benchmarks

Experiment velocity benchmarks

How fast your team runs experiments is itself a metric. The best product teams do not just make better decisions — they make more decisions faster.

The real benchmark is not experiments per month — it is learning velocity. A team that runs 4 experiments and changes strategy based on results beats a team that runs 20 experiments and ignores the losers. Count the decisions changed, not the tests launched.

Why Indian teams experiment less

Three structural reasons, none of which are excuses:

1. Smaller engineering teams. You cannot dedicate engineers to experimentation infra when you have 8 engineers total. Solution: use third-party experimentation tools (VWO is Indian — start there) instead of building your own.
2. HiPPO culture. The Highest Paid Person’s Opinion still dominates decision-making in many Indian companies. Experimentation threatens this. You need a founder who is willing to be wrong.
3. Statistical illiteracy. Most Indian PM teams call an experiment “done” after 3 days with 200 users. That is not an experiment — that is a guess with extra steps. Learn sample size calculation. Use it.

Why global benchmarks are 15-30% off for India

If you are reading benchmarks from Lenny’s Newsletter, Reforge, or First Round Capital and applying them directly to your Indian product, you are calibrating with the wrong ruler. Here is why:

Lower smartphone quality drives higher error rates

The median Indian smartphone has 3-4GB RAM, a mid-range processor, and is running Android 11 or older. Your app crashes more. Pages load slower. Animations stutter. This is not a UX problem — it is a physics problem. Error rates on Indian devices are 2-3x higher than on US devices, which directly suppresses activation and retention.

Adjustment: Subtract 5-10% from any global activation benchmark to account for device-driven friction.

Price sensitivity changes every conversion funnel

Indian users will spend 30 minutes comparing prices across 4 apps before buying a Rs 200 item. Free-to-paid conversion in India is structurally lower because willingness-to-pay thresholds are different. A $9.99/month subscription that converts 5% globally will convert 1-2% in India.

Adjustment: Halve any global free-to-paid conversion benchmark for consumer products. B2B enterprise is closer to global because the buyer is the company, not the individual.

Festival seasonality skews monthly metrics

Diwali, IPL season, Holi, end-of-financial-year — Indian products have 4-5 major seasonal spikes per year that do not exist in Western markets. If you measure monthly without adjusting for festivals, your October numbers look like a growth miracle and your November numbers look like a crisis. Neither is real.

Adjustment: Always compare same-period YoY, not sequential MoM, during festival months. Build a seasonality index for your category.

WhatsApp as acquisition channel breaks attribution

In India, a significant chunk of acquisition happens through WhatsApp forwards — shared links, group recommendations, screenshot virality. This traffic shows up as “direct” in your analytics. Your paid attribution model underestimates organic and overestimates paid contribution.

Adjustment: If your “direct” traffic is above 40%, investigate WhatsApp-driven sharing. Build UTM-tagged share links inside your product.

COD inflates top-of-funnel, deflates bottom

Cash on delivery accounts for 50-60% of e-commerce orders in India. COD orders have 2-3x the return rate of prepaid orders. They inflate your D1 retention (user placed an order) but deflate your D30 (user returned the product and never came back).

Adjustment: Track retention separately for COD and prepaid cohorts. Your “real” retention is the prepaid number.

How to use benchmarks without becoming a benchmark zombie

The three-step benchmark protocol

Step 1: Locate yourself. Find your product type and stage in the tables above. Identify which range you fall into. This takes 30 seconds and gives you directional awareness.

Step 2: Adjust for context. Apply the India-specific adjustments. Apply stage-specific adjustments. Apply your unique factors (e.g., if you are in a regulated industry, subtract another 5-10% from activation benchmarks for compliance friction).

Step 3: Trend over threshold. Once you know your adjusted benchmark, stop looking at it. Track your own trend line. A product improving from 5% to 8% D30 over three months is healthier than a product sitting flat at 10%. Direction matters more than position.

The benchmarks nobody talks about

Beyond the standard retention/activation/growth metrics, here are benchmarks that separate competent PMs from great ones:

Time to first “aha” moment: For consumer apps, this should be under 30 seconds. For B2B, under one session. If your time-to-aha is longer than your competitors’, your activation will always trail theirs regardless of how good your onboarding flow looks.

Support ticket rate per 1,000 MAU: Good is under 20 for consumer, under 50 for B2B. If you are above 50 for consumer, your product has UX debt that no amount of support scaling will fix.

NPS by cohort tenure: New users should have NPS 30-40. Users at 6+ months should have NPS 50+. If NPS declines with tenure, your product has a depth problem — it impresses on first use but disappoints over time.

Feature adoption rate for new launches: A new feature should reach 20-30% of eligible users within 4 weeks. If it does not, either the feature is not valuable, not discoverable, or not well-communicated. Below 10% after 4 weeks means the feature is dead on arrival.

Where to go next

• Build the metric system that tracks these benchmarks: Metrics and KPIs
• Diagnose why your numbers are off: Diagnosing Metric Drops
• Understand the growth loops that drive these numbers: Growth Analytics
• Improve the retention numbers you just benchmarked: Retention Loops
• Fix the activation rates that are dragging everything down: Activation Optimization