Insights to build for impact

Spec-driven AI has built what I need since September 2025.

Every range in the growth benchmark tool traces to a named publisher, and the rest are labelled estimates.

K-factor is your growth multiplier: how many new users each user brings in.

A product person's Claude Code setup, explained plainly: one chat window, a folder of facts, tickets, AI reviewer panels, and one human gate that ships work.

A consultancy raised money to become a product company, then kept closing bespoke deals.

Product means two things: what users interact with, and the responsibility a product team takes on.

We trained about 40 product managers at Avanza and watched how fast outcome talk turns into feature talk.

A product manager's first job is testing the business assumption under an idea before the team builds it.

Making "Book a Demo" the only way into your product loses more deals than it protects.

If you don't measure your product team's win rate, you ship on faith.

What is PLG?

A growth loop you can't measure won't grow anything.

A good product still loses deals when prospects wait too long to feel its value.

Datadrivet with Tom Airaksinen (PE Accounting) on scrappy mixed methods: cheap prototypes, Hotjar surveys, learning SQL, and a survey question that backfired.

Datadrivet talks to David Jurelius about tracking, cookie deprecation, ITP, and the groundwork that has to exist before any A/B testing can be trusted.

Datadrivet with Storytel's Seif Fendulky on conversion: the six questions a user needs answered before they buy, plus a classic A/B testing mistake to avoid.

Datadrivet talks to David Jurelius about the five roles a streaming analytics team needs to keep data clean across many platforms before any test runs.

Datadrivet on NPS: how the 0 to 10 question splits detractors, neutrals and promoters, what scores count as good, and an honest twist for measuring it.

Datadrivet on churn: the two reasons to lower it, three real cancellation stories (Hello Fresh, Netflix, Headspace), and how to find which leak is yours.

Datadrivet runs through nine practical Chrome extensions for product managers, from full-page screenshots and tech-stack detection to load-time checks and Loom.

Datadrivet talks to Kathleen Asjes of fintech Dreams about mixing quantitative and qualitative research to know if you're working on the right things.

Datadrivet talks to Gardenize CEO Jenny Rydebrink about app store optimization, cleaning up event tracking, and an onboarding change that doubled upgrades.

Datadrivet with Theresia Silander, founder of Eatit, on the mistakes the healthtech startup made and the fixes that came from starting with the user's problem.

Datadrivet talks to Loop54's Adam Hjort about a five-question survey he sends users every six months, plus why churned customers are worth a phone call.

Datadrivet with Kathleen Asjes, Head of Research and Insights at Dreams, on a shared customer journey and weekly sessions that get insights to every team.

Andreas Quensel on how Heja grew past 130,000 teams with 13 people: one clear activation metric, an end-to-end growth model, and data as the team's language.

Datadrivet on getting started with experiments: the three test types, real win rates from Google, Facebook and Microsoft, and a weekly loop to start testing.

Datadrivet with Avinode's Alexandra MacRae on the Customer Health Score: NPS, CES, CSAT and activity in one red-yellow-green matrix that drives action.

Datadrivet with Isa Cederberg (Birds Relations) on no-code product discovery: testing a prototype, Typeform and Bubble.io, a testable build in 3 to 4 weeks.

Datadrivet on the Next Feature Fallacy: why shipping the next feature rarely fixes growth, and why optimizing for learning beats optimizing for speed.

Datadrivet with Polestar's Fredrik Sterner Cederlof on mapping the full customer journey, NPS and CES across 45 digital touchpoints, and feedback in Slack.

Datadrivet on how a team gathers test ideas, clusters them into tracks, votes, and ranks the hypotheses with a modified ICE score so the best bets rise.

Datadrivet on the churn flow: the cancellation form with smart logic that tries to keep customers, keeps it pleasant, and shows you why people are leaving.

Datadrivet with Mike Rooseboom of Epidemic Sound on growth inside product, fixing licensing confusion, Core Web Vitals, and an experiment that hurt conversion.

Datadrivet walks through a hierarchy of evidence: from expert opinion and user research up through analytics, A/B testing, and A/B testing the A/B test.

Datadrivet with Johan Johansson (Carat) on a six-step framework for turning data into action: goals, observations, insights, impact, ownership, follow-up.

Datadrivet on the competencies that let a team experiment at high speed: a four-step setup and the cross-functional roles that make fast testing actually work.

Datadrivet with Jimmy Hagelfors (Brick) on App Store Optimization: what to optimize, when to ask for reviews, A/B testing in the stores, plus five tips.

Datadrivet on growing in a downturn: a three-month payback window, shifting to organic channels, fixing activation, and why measurable ROI beats high burn.

Datadrivet with Carl Lager (ArK Kapital) on the four stages of a startup, what to measure at each, and why qualitative insight matters before you have data.

Datadrivet on the digital analyst role: reading user behavior from numbers, decoding what it means, and turning patterns into ideas the whole team acts on.

Datadrivet with Valtech's Zarko Lindqvist on why being data-driven is a whole-org mindset, with lessons from Kivra's two-tests-a-week and Cancerfonden.

Datadrivet talks to Petter Flordal of Madden Analytics about asking customers for feedback continuously and building inventory forecasting around real needs.

Datadrivet goes behind the scenes on growth-team hiring: why a standard job ad boxes in the role, and what the Growth Hackers Sweden community suggested.

Datadrivet talks to Johan Johansson of Carat about senior versus junior analysts and the three conditions an organization needs to keep good analysts.

Datadrivet defines Growth with a simple 3-2-1 model: three areas of work, two working methods, and the one team setup that makes the whole system run.

Datadrivet on the two kinds of churn: the customers who decide to leave versus the ones who leave by accident, and why each group needs its own kind of fix.

Datadrivet returns after a year away.

Datadrivet talks to Readly's Head of Product, Emelie Ardby, about why working data-driven and prioritizing well are the most important parts of the job.

Datadrivet on why experiment throughput stalls: a workshop with a team that scaled from zero to five experiments a month and wants ten, and what held them back.

Datadrivet on working with analysts: why dashboards only show the past, why analysts are best placed to ask the business questions, and how to free their time.

Datadrivet on product-led growth: how it differs from marketing-led and sales-led growth, why it weighs retention, plus growth loops, experiments, and examples.

Datadrivet talks with Robert Ingemarsson of TimeWave on deleting unused features, why customers never complained, and the loss aversion behind dead code.

Datadrivet talks with Björn Idrén of CDON on daily data insights, a high-converting Voi download location, Klarna's coffee corner, and Jasmin's insight days.

Datadrivet on how to start a growth team: why the name can work against you, where the team belongs in the org, and where to point the first experiments.

Datadrivet breaks down seven things product-led companies share, from free trials and onboarding to transparent pricing, virality and self-serve buying.

Datadrivet with Johan Baltzar, co-founder of Steep, on democratizing analytics, defining metrics first, mobile-first design, and testing visions with users.

Datadrivet talks with Erwan Derlyn of Odepar on why startups fail when they build a key first and then hunt for a lock, and how to find demand before you build.

Datadrivet talks to Amanda AI CTO Torkel Ohman about automating ad creation for Google, Meta, and Bing, and growing from a drop-shipping test to a team of 30.

Datadrivet on 7-day retention: why a company losing 80% of new users was wasting acquisition spend, and three practical steps to make users come back.

Datadrivet with Hemnet CPO Francesca Cortesi on continuous discovery, a strong A/B-testing culture, and smoke-testing a service against global benchmarks.

Datadrivet explains product-led growth with two examples, Slack and Neo4j, where bottom-up adoption spread inside companies until buying became inevitable.

Datadrivet on Joni's WOW moment from data: 83% of a landing page's visitors were existing customers, and what that one finding changed about budget and teams.

Datadrivet on experiment volume: why teams run from a handful a month to a thousand a year, the 10 to 20 percent hit rate, and the roles you need to do it.

Datadrivet on the Product Death Cycle: why shipping the features users request rarely lifts adoption, and why the users who already left hold the real learning.

Datadrivet talks to Isa Cederberg of Birds Relations about the difference between a Product Owner and a Product Manager, from scrum backlog to product strategy.

A Datadrivet mini-episode on what counts as an experiment.

Datadrivet talks to Maria Petrova, VP of Product at Supermetrics, about a regular process for voice of the customer and reviewing user feedback every Friday.

The B2B SaaS growth benchmarks that actually matter in 2026 (retention, LTV:CAC, CAC payback, K-factor, activation) with sourced ranges and how to read them.

A North Star Metric is the single number that captures the value your product delivers.

A practical, no-data-team guide to benchmarking startup growth: which six metrics matter, where the ranges sit for B2B vs consumer, and how to read them.

Product-market fit goes beyond the Sean Ellis 40% test.

Retention dropping after Day 1 is almost always an onboarding or value-clarity problem, not product quality.

Benchmark vs target: a benchmark shows where you stand, a target is where you commit to land.

How real product growth works: a stable base of long-tenured users plus new cohorts retained and amplified by K-factor.

You don't need a referral program to raise K-factor.

Good Day-1 retention is 20% to 30% for consumer apps and 5% to 25% of signups for B2B SaaS.

What's a good month-over-month MAU growth rate?

Learn to read a retention curve: what the plateau, the slope, and the early cliff each tell you about product-market fit, with B2B and Consumer benchmarks.

Is your churn normal or fatal?

Why the same growth metric (retention, K-factor, LTV:CAC, payback) means something different for B2B SaaS vs. consumer apps.

The 2026 benchmark numbers for consumer apps (D1/D7/D30 retention, K-factor, CAC payback and activation), with sources, and why consumer differs from B2B.

A 6:1 or 8:1 LTV:CAC ratio usually means you're starving growth, not winning at it.

A good 90-day retention rate is 2.5% to 15.6% of signups for B2B SaaS (Amplitude) and 1% to 4% for consumer apps.

Product metrics move together.

A good CAC payback period is 6 to 12 months for B2B SaaS, 1 to 6 months for consumer apps.

Free growth benchmark calculator.

A long CAC payback period usually points to a retention problem, not a CAC problem.

A healthy LTV:CAC ratio is 3:1 to 5:1 for B2B SaaS and around 3:1 for consumer apps.

A good activation rate is roughly 25% to 40% for B2B SaaS and 15% to 30% for consumer apps, but only once you've defined the right "aha" moment.

July's SaaS benchmark reports show CAC paybacks still compressing across three new studies, plus a widely cited B2B retention figure worth reading twice.