A 20% price rise on a business with 5% churn adds more profit in a quarter than most growth initiatives manage in a year. It is also the change most likely to blow up in your face if you do it in the wrong order.
First, find out whether you can
Before touching anything, answer three questions from data you already have:
Has price ever changed before, and what happened? If a previous rise saw no churn spike, you have your answer.
What do competitors charge? If you're at half the market rate, the gap is your headroom.
Do customers cite price when they cancel? Read the actual cancellation reasons. If price appears rarely, it isn't the binding constraint.
The safe sequence
New customers only, first. Raise the public price. Existing customers see nothing. Watch conversion for a full cycle — 60 days minimum.
If conversion holds, consider existing customers. If it dropped meaningfully, you've learned something cheaply and can stop here.
Grandfather the loyal base, or move them slowly. Long-tenured customers at old prices are usually your advocates. Keeping them on legacy pricing costs less than it looks and buys enormous goodwill.
Give notice — 60 days, minimum. Announced in advance, from a person, with a reason.
How to announce it
The email that works is short, direct and specific about the date and the amount. It does not apologise, does not over-explain, and mentions what's improved since the last price change if anything genuinely has. Offer a way to lock in the old rate by moving to annual — a good number of people will take it, and you've converted a price objection into a year of cash.
What to expect
On a well-run rise for new customers only: essentially no churn, and a conversion dip of a few points that usually recovers as you adjust the page around the new number. On existing customers with proper notice: a small spike of cancellations concentrated in the accounts that were least engaged anyway.
The customers you lose to a 20% rise are, almost by definition, the ones generating the most support load per dollar. That's not a consolation — it's often the actual benefit.