The Update You’ve Been Planning for 3 Months Might Be the Wrong Update
Big updates are chosen by momentum. They should be chosen by evidence.
Long-planned updates carry a weight that has nothing to do with their value. People have designed, built and talked about them, so cancelling feels like waste. Meanwhile the game may have changed since the plan was written. The question is not whether the update is good. It is whether it is the best use of the next three months.
Why the wrong update gets built
- It was a good idea when it was first written and was never re-examined.
- It is what the loudest players ask for, which is not the same as what most players need.
- It is exciting to build, so it beats unglamorous fixes.
- It was announced, so dropping it feels like a public failure.
- Sunk effort: work already done is treated as a reason to continue.
Requests versus behaviour
What players say and what players do disagree often. Requests tell you what they believe will help. Behaviour tells you where they actually struggle or leave. Use requests as leads and behaviour as evidence. See Players Are Complaining About the Wrong Problem.
Score each candidate
For every update idea, answer five questions in a shared sheet.
- Reach
- How many players does the problem affect, and at what stage of their journey?
- Severity
- Does it make players leave, stall or spend less, or does it only irritate them?
- Frequency
- Does it happen every session, or once a month?
- Likely effect
- Which metric should move: first-session retention, D7, playtime, revenue? Be specific and honest about the size you expect.
- Cost
- Build time, testing, balance, ongoing maintenance and what else it delays.
| Candidate | Reach | Severity | Frequency | Effect | Cost |
|---|---|---|---|---|---|
| New zone | Few reach it | Low | Rare | Playtime for veterans | High |
| Fix tutorial dead end | Every new player | High | Every first session | Early retention | Low |
| Rebalance shop prices | Spenders | Medium | Per purchase | Revenue | Low |
Do not turn this into a formula that outputs a winner. Use it to force the argument into the open. A small fix that touches every new player often beats a large feature that touches a few veterans.
Revenue and retention are different goals
An update that raises revenue per player can hurt retention if it pushes prices or friction early. An update that improves retention can take weeks to show up as revenue. Decide which one you need now and write it down.
Avoid the three-month trap
- Slice the big update. Ship the part that addresses the problem first.
- Set a checkpoint every few weeks where you ask, would we start this today?
- Put a small experiment before the large build: a rough version for a subset, or a quick version of the key mechanic.
- Keep a short list of cheap fixes that always gets a slot in every cycle.
Then use the loop in How to Iterate on a Roblox Experience to judge the result against the baseline you wrote first.