Software as a business is not a deck or a dream board. It is a bunch of messy pull requests, support tickets at odd hours, and a balance between product gut and numbers that refuse to be ignored. Right now the conversation in dev land is heated about React licensing, Kubernetes keeps winning mindshare, and every week a new chat bot pitch lands in the inbox. Through all the noise, the trench work looks the same. Make something people pay for, keep them around, and do it without burning the team or the bankroll. Here are notes from the field that keep paying rent.
Definitions
MRR: monthly recurring revenue. The heartbeat. If it is lumpy, you are in project mode, not product mode.
Churn: the rate at which customers or revenue leave. Logo churn counts accounts. Revenue churn counts money. Track both.
LTV: lifetime value. The gross margin you earn over the life of a customer. Pair it with CAC, the cost to acquire a customer. If LTV over CAC is under three, alarms should ring.
Activation: when a new user reaches first value. The moment they go from trying to winning. Time to first value under fifteen minutes is a solid target.
ARPA: average revenue per account. Helps decide if you can afford sales help or if you must stay self serve.
Examples
- Self serve SaaS: A tiny team ships a reporting tool for Shopify store owners. Sign up is ungated, trial is fourteen days. They killed a third of churn by fixing empty states and adding a tight onboarding checklist inside the app. Live chat during trial doubled conversions. They stopped writing thought pieces and wrote better help docs. Revenue grew because users reached value faster.
- Agency to product: A services shop had a repeat client need, so they spun it into a product with tiered plans. They offered annual with two months free and kept monthly for cash shy customers. A founder handled sales calls for deals above two hundred a month. The calls surfaced must have features that were not on the roadmap. They said no to custom work that did not help the core. Growth was slower than a flashy launch, steadier than their project pipeline.
- Mobile with subscriptions: A solo dev launched a writing app with a free tier and a seven day trial on premium. Paid search looked cheap until refunds wiped the margin. What worked was a simple email course that taught a writing habit and a clean upsell in the app once users finished lesson three. The hook was not features. The hook was progress.
Counterexamples
- Chasing one giant customer and turning the codebase into their private playground. You get a check and a product no one else wants.
- Launching on a platform you do not control and building your only growth channel on it. Policy change, game over.
- Buying ads before retention clears sixty days. You rent traffic, they bounce, and the math never recovers.
- Pricing for what you think you will be next year. Price for the value you deliver today, then earn the right to raise it.
Decision rubric
- Trial or free tier: Pick one. If time to first value is under ten minutes, a time boxed trial is your friend. If learning takes longer, keep a free tier with limits that match the real value line.
- Tiers: Three plans, clear ladders, one plan marked as the favorite. Add annual with two months free. Remove prices that end in odd cents.
- Sales or no sales: If ARPA is under one hundred a month, keep it self serve. If it is above two hundred a month and the cycle is under thirty days, light sales can work. Record calls, tag themes, feed the roadmap.
- What to build next: Does it improve activation, retention, or expansion in the next sixty days. If yes, ship it. If it only helps new signups and your churn is above four percent, wait.
- Metrics to watch weekly: New trials, activation rate, MRR growth, net revenue churn, payback time on CAC. Keep a simple dashboard the whole team can recite.
Lesson learned
The winner is the team that learns faster than it builds. Talk to customers every week. Reduce the time to first value. Price with courage. Keep the surface area small so support does not drown you. Tool picks matter less than the habits around them, though a stack with Stripe, a decent chat tool, a clean deploy path, and a way to plug in other apps through Zapier will save days. When the feed is full of hot takes about licenses, coins, and the next silver bullet, do the boring work. Retention is the quiet growth loop. Support is marketing. A clear product is sales. Profit is oxygen. Keep going.