Journal · November 2025
Reading retention after you raised the price
The blended curve will look brave for a while. That bravery is mostly people who never saw the new price.
Price increases are catnip for blended LTV. Revenue per surviving user jumps because legacy annual plans keep billing at the old rate while new users pay more — or churn trying. If you plot one retention curve across the change date, you will almost always see a flattering wobble. Boards remember the wobble.
Split immediately: legacy price versus new price, and if you can, grandfathered annual versus new monthly. In our Retention Narrative for Boards desk we force the one-pager to show both curves on the same axes, with n labelled. If new-price n is still tiny, say so in the title, not in a footnote.
What “stabilisation” actually looks like
New-price cohorts often dip in weeks 3–6 as trial-to-paid and first-renewal take the hit. That dip is information. If it recovers and the net contribution still clears your hold-rule, keeping the price can be correct. If the dip is a cliff and refunds spike, you do not have a “positioning” problem; you have a price the market rejected.
Do not compare new-price day-90 LTV to old-price day-90 LTV until both cohorts have lived ninety days under their own price. Comparing a mature legacy cohort to a baby new-price cohort is how CFOs get told a story that unravels in the next quarter.
VAT and store display
If the store listing shows a VAT-inclusive price in Thailand and your ledger recognises ex-VAT, write that on the slide. A 15 percent list-price rise is not a 15 percent net rise. Vault Nodehub worksheets include a one-line gross-to-net walk specifically so this does not become an argument in the room.