Launch day is the easy part. The month after it is where a SaaS product is actually decided: whether a new account reaches the thing it signed up for, whether the second invoice gets paid, whether anyone can tell you why not.
What ships in the first release
Not the feature list. The shortest path a real customer can walk end to end — sign up, do the job once, see the result, invite a colleague. Everything else waits until that path holds without a human helping.
That ordering is unpopular in a roadmap review and it is the reason the first retention numbers mean anything at all.
Billing is product work
Plans, seats, trials, proration, dunning, tax. Teams treat this as a checkout widget and then spend a year unpicking it. We model the pricing before we wire the provider, because a pricing change you cannot ship is a strategy you do not have.
Related build detail lives in SaaS and product development.
Instrumentation goes in with the feature
Events are defined in the same ticket as the screen they describe, with a name that will still make sense in a year. A funnel bolted on afterwards measures whatever was easy to measure.
What we want visible from week one:
- Signup to activation, split by the channel the account came from
- Time to the first completed job, not time on site
- Expansion and contraction inside existing accounts
- The support tickets that repeat, counted like a metric
Then analytics and CRO work has something to read.
Growth for a product that sells itself, slowly
B2B SaaS demand is search-heavy and referral-heavy, and the buying committee is usually three people with different questions. Documentation and comparison content carry more weight than campaigns; paid is best spent on the terms that already convert, not on awareness for a product nobody can describe yet.
Pricing pages and onboarding emails are growth surfaces, not decoration — the email and CRM track owns the second one.
Taking over an existing product
We audit the code, the data model and the funnel before proposing anything. A rewrite is a last resort. Most stalled SaaS products we see do not need new architecture; they need the activation path fixed and the billing untangled.
More sectors on the industries page.