Review and apply price tests
MARPLA PriceLab uses saved rules and shows a proposal for each product; no external AI is needed for this step.
How to use it
- For each row, choose green Approve, red Reject, or leave it Unreviewed. Reject saves the decision without changing the price, while an unreviewed product is untouched.
- The common apply button processes reviewed products only. The new price must differ from the current price; a same-price test is not created.
- The shared date and time may be in the past: submission then happens now with the actual request time, while both chosen and actual times stay in history. During the first three hours after UTC midnight, Apply now is the default; from 03:00 UTC, the next UTC midnight is selected.
- After a batch run, partial failures remain visible beside successful results. Retry sends only unresolved products; scheduled changes run on the server even when the browser is closed.
Advice and practical context
The two charts answer different questions
The gross-revenue chart shows daily Item Monetization Revenue for the exact ProductKey. It answers how much revenue Roblox assigned to that item on a complete day. The per-player chart divides the same amount by official Daily Active Users for that day, helping account for changes in audience size.
Per-player revenue is not profit per buyer: DAU includes all active users, not only purchasers. Roblox Analytics does not provide row-level costs and fees here, so both charts describe gross-revenue impact rather than accounting net profit.
Compare adjacent stable price periods
The percentage compares the current stable price window with the immediately preceding one. Check boundary dates and use completed days. If advertising, a game update or audience composition changed at the same time, record it as another explanation rather than assigning the full movement to price.
Example: gross revenue rises while the per-player value remains nearly unchanged. That is consistent with higher DAU and does not prove improved purchase behavior. If both charts rise, still inspect units, price and other product changes.
Missing data remains missing
Missing overall analytics is never converted to zero. A zero product row is valid only when Roblox returned the overall item metric for that day. A pink connector between known points marks a gap and does not create intermediate sales.
Do not calculate an average from a visually connected gap. Check coverage for both compared price windows first; when a gap is material, keep the result provisional and wait for comparable completed days.
Technical detailsCalculations, permissions and behavior
Price-effect charts
The gross-revenue chart shows daily Item Monetization Revenue for an exact ProductKey. The per-player chart divides it by official Daily Active Users for the same complete day. Missing overall analytics is never turned into zero; a zero item row is allowed only when Roblox returned the day's overall item metric.
Percentages compare the current stable price window with the immediately preceding one. Roblox Analytics does not provide per-row costs and fees, so this measures gross-revenue impact rather than accounting net profit.
Price management
MARPLA stores the base price of every pass and developer product each day and separately records detected changes. It compares consecutive stable price windows against two objectives: average daily gross Robux revenue and gross revenue divided by every active player that day. The first window is one complete UTC day.
If the audience is below 500 players or the daily result is noisy, the minimum window expands to 3 days; below 100 players or with high volatility it expands to 7 days. The observed PriceLab is MARPLA rules, not an external AI: it shows the next-price proposal and reason, while the owner or game administrator decides for each product. For each test, choose green Approve, red Reject, or leave it Unreviewed.
Reject saves the decision without changing the price; Unreviewed leaves the product untouched. The common apply action processes reviewed products only: it submits a price change for approved products and records rejected products without changing their prices. A same-price test is never created.
One shared date and time applies to the selected decisions: a past time is allowed, but submission happens now with the actual request time while history keeps the chosen time and actual submission. During the first three hours after the current UTC midnight, the default is Apply now; from 03:00 UTC onward, the default is the next UTC midnight. Partial failures remain visible, successful products stay saved, and retry sends only unresolved products.
Scheduled changes run on the server, so the browser can be closed. Roblox Price Optimization and immutable products are never edited.