Which charts belong on a launch week dashboard?
Put the charts that can change a decision this week on one screen. In launch week there are three decisions: do we patch the build, do we push a channel or a creator harder, and is the launch on plan. So the dashboard has three groups: players and builds, game health, and traffic and links.
The 11 dashboard charts read your own events: the events of the game build, the page views of your website and the clicks on your short links. Steam and itch.io numbers do not come from your events. They come from the stores, and Manalens shows them in two built-in reports, the Steam view and the itch.io view. As of 19 November 2026, you cannot put Steam wishlists or sales in a dashboard chart, so keep the Steam view open in a second tab.
Eight of the 11 charts are in the product templates Game KPIs, Game health and Channels and campaigns. You move them to one dashboard. The other three are SQL charts with a spec that you can copy from this page.
Sources: Manalens dashboards, Steam and itch.io documentation.
Who is the launch week dashboard for, and when do you use it?
It is for a small studio that ships a paid game on Steam, on itch.io or on both, and has one person who looks at numbers. That person can be the developer. You do not need SQL knowledge: you copy the charts.
Use it from release day to day 7. Build it one or two weeks before release, with the demo or a beta build as test data, so that the first real number does not arrive on an empty chart. Set the dashboard date range to the launch week. After day 7, change to a weekly review, because daily numbers of a small game move a lot by chance.
It does not fit a free-to-play game with in-game purchases as well. Such a game also needs revenue per player, and the Game KPIs template has those charts (ARPDAU and ARPPU).
- Steam launch date planner →
- Steam and itch.io launch checklist →
- A weekly metrics review for your game after launch →
Sources: Manalens dashboards documentation.
Which questions does the launch week dashboard answer?
Each question below has one chart or one report in the chart list. A chart that answers none of them does not belong on this dashboard.
- How many new players arrive each day, and how many play?
- Did players move to the new build after a patch?
- Do the players of launch day come back the next day?
- Is a build broken, and which error do we fix first?
- Which channels and campaigns send visitors to our website?
- Which creator links and posts get clicks?
- Which countries do the new players come from?
- How many wishlists and sales does Steam report, in total and from tagged links?
- How many views, downloads and payments does itch.io report?
Sources: Manalens dashboards documentation.
What is the chart list of the launch week dashboard?
The list has 14 rows: 11 dashboard charts and 3 built-in reports. A dashboard chart has a chart spec that you can copy. A built-in report is a page of the product that you open next to the dashboard. The order of the rows is the order on the dashboard, from top to bottom.
Every row is something that you can build or open in Manalens as of 19 November 2026. The sections after the table have one chart each: what it shows, how to read it, what to do, the steps and the spec.
| Chart | Question it answers | Data source | Chart type | Breakdown |
|---|---|---|---|---|
| Daily active players | How many new players arrive each day, and how many play? | Game events: game_open (is_first_open), all events of the build | Line | None (two lines: players, new_players) |
| Players by day and build | Did players move to the new build after a patch? | Game events with context.app.build | Stacked bar | app_build |
| D1 retention | Do the players of launch day come back the next day? | Game events: game_open (is_first_open), session_start | Number | None |
| Install cohorts | Does each day of new players come back, day by day? | Game events: game_open, session_start | Retention | Cohort day |
| Crash-free sessions | Is the game stable for the players of this week? | Game events: app_unclean_exit, error | Number | None |
| Errors per session by build | Which build is broken? | Game events: error, with context.app.build | Bar | app_build |
| Top errors | Which error do we fix first? | Game events: error | Table | fingerprint |
| Sessions by channel | Which channels send visitors to our website? | Sessions (website page views; game sessions count as Direct) | Stacked bar | channel_type (landing channel) |
| Campaigns | Which tagged campaigns send visitors, and do they stay on the site? | Sessions with a UTM source or campaign on the first page view | Table | utm_source, utm_medium, utm_campaign |
| Short link clicks by creator and campaign | Which creator links and posts get clicks? | Short links: link_clicked events | Table | creator, campaign, slug, target_host |
| New players by country | Which countries do the new players come from? | Game events: game_open (is_first_open), country from the request IP | Table | country |
| Steam view: visits, wishlists and purchases by UTM | Which tagged links brought Steam visits, wishlists and purchases? | Steam UTM Analytics CSV, imported by you | Built-in report (not a dashboard chart) | utm_source, utm_medium, utm_campaign, utm_content, utm_term |
| Steam view: all wishlists and sales by day | How many wishlists and sales does Steam report in total? | Early access: Steamworks Financial API key, daily sync | Built-in report (not a dashboard chart) | Day |
| itch.io view: views, downloads and payments by day | How many views, downloads and payments does itch.io report? | itch.io API key, daily sync | Built-in report (not a dashboard chart) | Day; channels and sources of matched payments |
Sources: Manalens dashboards and chart spec documentation.
How do you build and read the "Daily active players" chart?
It shows two lines per day: the players who sent an event from the game build, and the new players. A new player is an install whose first game_open has is_first_open set to true. Days are UTC days.
How to read it: the new players line is the size of the launch, and it usually falls after day 1. The gap between the two lines is the players who came back. This chart counts installs that sent events, not copies sold. A buyer who has not started the game is not in it, and an offline player shows later, when the game sends its queue.
What to do: compare the new players of each day with the net units of the same day in the Steam view. If far fewer players start the game than buy it, check that the release build has the SDK and that the game can reach your collector endpoint.
- For "Daily active players": open Dashboards, click the Game KPIs template and click Create board. Skip this step if you have the Game KPIs board.
- On the Game KPIs board, open the chart menu of "Daily active players" and move the chart to your launch week dashboard.
Chart spec for the create_chart tool of the MCP server (add your boardId)
{
"title": "Daily active players",
"spec": {
"version": 1,
"query": {
"kind": "sql",
"sql": "SELECT toDate(received_at) AS day, uniqExact(anonymous_id) AS players, uniqExactIf(anonymous_id, event = 'game_open' AND JSONExtractBool(properties,'is_first_open')) AS new_players FROM raw_events WHERE type = 'track' AND JSONHas(context,'app') AND channel != 'server' AND anonymous_id != '' AND received_at >= toDateTime({{date_from}}) AND received_at < toDateTime({{date_to}}) + INTERVAL 1 DAY AND JSONExtractString(context,'device') != 'bot' GROUP BY day ORDER BY day WITH FILL FROM (toDate({{date_from}})) TO (toDate({{date_to}}) + 1)"
},
"view": {
"mark": "line",
"encoding": {
"x": {
"field": "day"
},
"y": {
"field": "value"
},
"color": {
"field": "series"
}
},
"transform": [
{
"fold": [
"players",
"new_players"
],
"as": [
"series",
"value"
]
}
]
}
}
}Sources: Manalens app events and dashboards documentation.
How do you build and read the "Players by day and build" chart?
It shows the players of each day, split by the build that they ran. The build is the build value that you give the SDK. A player who ran two builds on one day counts in both, so the bar can be a little higher than the players of the day.
How to read it: after a patch, the color of the new build must take over the bar in the next days. A color that stays for days is a group of players on an old build, for example players who play offline.
What to do: if an old build keeps its players, read the error charts for that build before you conclude that the patch fixed the problem. Give every upload a new build value, or this chart and the two error charts cannot tell the builds apart.
- For "Players by day and build": open your launch week dashboard, click Add chart, then Blank chart. Set Chart title to "Players by day and build" and Chart type to Stacked bar.
- Copy the text of "sql" from the spec of "Players by day and build" below and paste it into SQL query.
- Set X column to day, Y column to players and Breakdown column to build. Click Add chart.
Chart spec for the create_chart tool of the MCP server (add your boardId)
{
"title": "Players by day and build",
"spec": {
"version": 1,
"query": {
"kind": "sql",
"sql": "SELECT toDate(received_at) AS day, if(JSONExtractString(context,'app','build') != '', JSONExtractString(context,'app','build'), '(none)') AS build, uniqExact(anonymous_id) AS players FROM raw_events WHERE type = 'track' AND JSONHas(context,'app') AND channel != 'server' AND anonymous_id != '' AND received_at >= toDateTime({{date_from}}) AND received_at < toDateTime({{date_to}}) + INTERVAL 1 DAY AND JSONExtractString(context,'device') != 'bot' GROUP BY day, build ORDER BY day, build"
},
"view": {
"mark": "bar",
"encoding": {
"x": {
"field": "day"
},
"y": {
"field": "players",
"stack": true
},
"color": {
"field": "build"
}
}
}
}
}Sources: Manalens app events and dashboards documentation.
How do you build and read the "D1 retention" chart?
It shows one number: of the players whose first game_open was one day before a day of the date range, the share who started a session on that day. With the date range on the launch week, it is the D1 retention of the launch cohorts so far.
How to read it: on release day the card shows 0.0%, because no player is one day old yet. The first real value arrives on day 2. A premium game with a short story can have a low value and still please its players, so compare with your demo or your playtest, not with mobile benchmarks.
What to do: if D1 is much lower than in your demo, open the first session. Look at the crash-free sessions card and the top errors first, then at the place where new players stop.
- For "D1 retention": open Dashboards, click the Game KPIs template and click Create board. Skip this step if you have the Game KPIs board.
- On the Game KPIs board, open the chart menu of "D1 retention" and move the chart to your launch week dashboard.
Chart spec for the create_chart tool of the MCP server (add your boardId)
{
"title": "D1 retention",
"spec": {
"version": 1,
"query": {
"kind": "sql",
"sql": "WITH installs AS ( SELECT anonymous_id AS player, min(toDate(received_at)) AS install_day FROM raw_events WHERE type = 'track' AND JSONHas(context,'app') AND channel != 'server' AND event = 'game_open' AND JSONExtractBool(properties,'is_first_open') AND anonymous_id != '' AND received_at >= toDateTime({{date_from}}) - INTERVAL 30 DAY AND received_at < toDateTime({{date_to}}) + INTERVAL 1 DAY AND JSONExtractString(context,'device') != 'bot' GROUP BY player ), played AS ( SELECT anonymous_id AS player, groupUniqArray(toDate(received_at)) AS days FROM raw_events WHERE type = 'track' AND JSONHas(context,'app') AND channel != 'server' AND event = 'session_start' AND anonymous_id != '' AND received_at >= toDateTime({{date_from}}) AND received_at < toDateTime({{date_to}}) + INTERVAL 1 DAY AND JSONExtractString(context,'device') != 'bot' GROUP BY player ) SELECT countIf(install_day BETWEEN toDate({{date_from}}) - 1 AND toDate({{date_to}}) - 1) AS d1_players, if(d1_players = 0, 0, countIf(install_day BETWEEN toDate({{date_from}}) - 1 AND toDate({{date_to}}) - 1 AND has(days, install_day + 1)) / d1_players) AS d1, countIf(install_day BETWEEN toDate({{date_from}}) - 7 AND toDate({{date_to}}) - 7) AS d7_players, if(d7_players = 0, 0, countIf(install_day BETWEEN toDate({{date_from}}) - 7 AND toDate({{date_to}}) - 7 AND has(days, install_day + 7)) / d7_players) AS d7, countIf(install_day BETWEEN toDate({{date_from}}) - 30 AND toDate({{date_to}}) - 30) AS d30_players, if(d30_players = 0, 0, countIf(install_day BETWEEN toDate({{date_from}}) - 30 AND toDate({{date_to}}) - 30 AND has(days, install_day + 30)) / d30_players) AS d30 FROM installs LEFT JOIN played USING player"
},
"view": {
"mark": "text",
"encoding": {
"text": {
"field": "d1",
"format": ".1%"
}
}
}
}
}Sources: Manalens metrics documentation (Game section).
How do you build and read the "Install cohorts" chart?
It shows one row for each day of new players. The columns are the days after the first game_open, and each cell is the share of that cohort that started a session on that day. The retention type is recurring and the period is one day.
How to read it: read down a column, not along a row. The column of day 1 compares the launch day cohort with the cohorts of day 2 and day 3. A period that is not complete shows a dash, not a low number. Small cohorts move a lot: in a cohort of 40 players, one player is 2.5 points (our arithmetic, an example).
What to do: if the cohort after a patch keeps more players on day 1 than the cohort before it, the patch probably helped. If the cohort of a creator video day keeps fewer players, that audience was a weaker fit, and the Campaigns table shows who sent it.
- For "Install cohorts": open Dashboards, click the Game KPIs template and click Create board. Skip this step if you have the Game KPIs board.
- On the Game KPIs board, open the chart menu of "Install cohorts" and move the chart to your launch week dashboard.
Chart spec for the create_chart tool of the MCP server (add your boardId)
{
"title": "Install cohorts",
"spec": {
"version": 1,
"query": {
"kind": "retention",
"settings": {
"granularity": "day",
"retention_type": "recurring",
"entry_event": {
"type": "track",
"value": "game_open"
},
"return_event": {
"type": "track",
"value": "session_start"
}
}
},
"view": {
"mark": "x-retention"
}
}
}Sources: Manalens retention documentation.
How do you build and read the "Crash-free sessions" chart?
It shows 1 minus the share of game sessions that crashed. A session crashed when it has an app_unclean_exit event, or an error event with the severity critical and handled set to false. The Godot and Unity SDKs send app_unclean_exit at the next start of the game, in the session that crashed.
How to read it: a crash shows only after the player starts the game again, so the number of the last hours is too good. A player who never comes back after a crash is never counted. Read it as the best case. Manalens stores no native crash dumps or symbols, so this card counts crashes and does not explain a native crash.
What to do: if the card falls after a new build, open "Errors per session by build" and "Top errors" at once. Decide in the first hour whether to roll back or to patch.
- For "Crash-free sessions": open Dashboards, click the Game health template and click Create board. Skip this step if you have the Game health board.
- On the Game health board, open the chart menu of "Crash-free sessions" and move the chart to your launch week dashboard.
Chart spec for the create_chart tool of the MCP server (add your boardId)
{
"title": "Crash-free sessions",
"spec": {
"version": 1,
"query": {
"kind": "sql",
"sql": "WITH s AS ( SELECT session_id, any(anonymous_id) AS player, countIf(event = 'app_unclean_exit' OR (event = 'error' AND JSONExtractString(properties,'severity') = 'critical' AND NOT JSONExtractBool(properties,'handled'))) > 0 AS crashed, countIf(event = 'error' AND JSONExtractString(properties,'severity') IN ('error','critical')) > 0 AS errored, countIf(event = 'error') AS errors FROM raw_events WHERE type = 'track' AND JSONHas(context,'app') AND channel != 'server' AND session_id != '' AND received_at >= toDateTime({{date_from}}) AND received_at < toDateTime({{date_to}}) + INTERVAL 1 DAY AND JSONExtractString(context,'device') != 'bot' GROUP BY session_id ) SELECT count() AS sessions, if(count() = 0, 0, 1 - countIf(crashed) / count()) AS crash_free_sessions, if(uniqExact(player) = 0, 0, 1 - uniqExactIf(player, crashed) / uniqExact(player)) AS crash_free_players, if(count() = 0, 0, countIf(errored) / count()) AS errored_sessions, if(count() = 0, 0, sum(errors) / count()) AS errors_per_session FROM s"
},
"view": {
"mark": "text",
"encoding": {
"text": {
"field": "crash_free_sessions",
"format": ".1%"
}
}
}
}
}Sources: Manalens errors and crashes documentation.
How do you build and read the "Errors per session by build" chart?
It shows the number of error events per game session for each build, for the 20 builds with the most sessions. Builds with no build value show as (none).
How to read it: compare the bar of the newest build with the bar of the build before it. Example with invented numbers: 0.10 errors per session on build 571 and 0.90 on build 572 means that build 572 added a problem. A build with very few sessions can have a high bar by chance.
What to do: if the newest build is the highest bar, stop the next upload and fix the top error of that build first.
- For "Errors per session by build": open Dashboards, click the Game health template and click Create board. Skip this step if you have the Game health board.
- On the Game health board, open the chart menu of "Errors per session by build" and move the chart to your launch week dashboard.
Chart spec for the create_chart tool of the MCP server (add your boardId)
{
"title": "Errors per session by build",
"spec": {
"version": 1,
"query": {
"kind": "sql",
"sql": "SELECT build, if(sessions = 0, 0, errors / sessions) AS errors_per_session FROM ( SELECT if(JSONExtractString(context,'app','build') != '', JSONExtractString(context,'app','build'), '(none)') AS build, countIf(event = 'error') AS errors, uniqExactIf(session_id, session_id != '') AS sessions FROM raw_events WHERE type = 'track' AND JSONHas(context,'app') AND channel != 'server' AND received_at >= toDateTime({{date_from}}) AND received_at < toDateTime({{date_to}}) + INTERVAL 1 DAY AND JSONExtractString(context,'device') != 'bot' GROUP BY build ) ORDER BY sessions DESC, build LIMIT 20"
},
"view": {
"mark": "bar",
"encoding": {
"x": {
"field": "build"
},
"y": {
"field": "errors_per_session"
}
}
}
}
}Sources: Manalens errors and crashes documentation.
How do you build and read the "Top errors" chart?
It shows one row for each error fingerprint: the latest message, the error type, the count, the players, the first and last time seen in the date range, and the latest build. Errors with the same fingerprint are one row.
How to read it: sort your work by players, not by count. One player in a loop can send the same error many times. The column latest_build tells you whether the newest build still has the error.
What to do: fix the error with the most players first. After the patch, check that last_seen of that row stops moving for the new build.
- For "Top errors": open Dashboards, click the Game health template and click Create board. Skip this step if you have the Game health board.
- On the Game health board, open the chart menu of "Top errors" and move the chart to your launch week dashboard.
Chart spec for the create_chart tool of the MCP server (add your boardId)
{
"title": "Top errors",
"spec": {
"version": 1,
"query": {
"kind": "sql",
"sql": "SELECT argMax(JSONExtractString(properties,'message'), original_timestamp) AS message, any(JSONExtractString(properties,'error_type')) AS error_type, count() AS errors, uniqExact(anonymous_id) AS players, min(original_timestamp) AS first_seen, max(original_timestamp) AS last_seen, argMax(JSONExtractString(context,'app','build'), original_timestamp) AS latest_build, JSONExtractString(properties,'fingerprint') AS fingerprint FROM raw_events WHERE type = 'track' AND event = 'error' AND received_at >= toDateTime({{date_from}}) AND received_at < toDateTime({{date_to}}) + INTERVAL 1 DAY AND JSONExtractString(context,'device') != 'bot' GROUP BY fingerprint ORDER BY errors DESC, fingerprint LIMIT 20"
},
"view": {
"mark": "x-table"
}
}
}Sources: Manalens errors and crashes documentation.
How do you build and read the "Sessions by channel" chart?
It shows the sessions of each day, stacked by the acquisition channel of the first page view of the session. It keeps the eight largest channels and puts the rest in Other. In this product, channel_type is the acquisition channel, for example Organic Search or Referrals. The column named channel is another thing: the origin of an event (web, app or server).
How to read it: a game session has no referrer, so game sessions are in Direct. In launch week this makes Direct very large. Read the other colors: a new color or a taller part on one day is a channel that woke up, for example Organic Video on the day of a creator video.
What to do: when a channel grows, open the Campaigns table to find the link, and give that channel one more post or one more creator while the launch is fresh.
- For "Sessions by channel": open Dashboards, click the Channels and campaigns template and click Create board. Skip this step if you have the Channels and campaigns board.
- On the Channels and campaigns board, open the chart menu of "Sessions by channel" and move the chart to your launch week dashboard.
Chart spec for the create_chart tool of the MCP server (add your boardId)
{
"title": "Sessions by channel",
"spec": {
"version": 1,
"query": {
"kind": "sql",
"sql": "WITH s AS ( SELECT anonymous_id AS visitor, day, if(channel_type != '', channel_type, 'Direct') AS channel, if(utm_source != '', utm_source, '(none)') AS utm_source, if(utm_medium != '', utm_medium, '(none)') AS utm_medium, if(utm_campaign != '', utm_campaign, '(none)') AS utm_campaign, pageviews AS pv, custom_events AS ce, revenue AS session_revenue FROM sessions ), top_channels AS ( SELECT channel FROM s GROUP BY channel ORDER BY count() DESC, channel LIMIT 8 ) SELECT day, channel_group, count() AS sessions FROM ( SELECT day, if(channel IN (SELECT channel FROM top_channels), channel, 'Other') AS channel_group FROM s ) GROUP BY day, channel_group ORDER BY day, channel_group"
},
"view": {
"mark": "bar",
"encoding": {
"x": {
"field": "day"
},
"y": {
"field": "sessions",
"stack": true
},
"color": {
"field": "channel_group"
}
}
}
}
}Sources: Manalens dashboards and sessions documentation.
How do you build and read the "Campaigns" chart?
It shows one row for each combination of utm_source, utm_medium and utm_campaign: visitors, sessions, bounce rate in percent, pages per session and revenue. The revenue is the revenue property of the events in the same session.
How to read it: this table is about your website, not about Steam. A campaign whose link goes straight to the Steam page never shows here. It shows in the Steam view. For a premium Steam game the revenue column is mostly empty, because the sale happens on Steam.
What to do: if a campaign has many sessions and a high bounce rate, the page did not match the promise of the post. Change the landing page or send that link straight to the store page.
- For "Campaigns": open Dashboards, click the Channels and campaigns template and click Create board. Skip this step if you have the Channels and campaigns board.
- On the Channels and campaigns board, open the chart menu of "Campaigns" and move the chart to your launch week dashboard.
Chart spec for the create_chart tool of the MCP server (add your boardId)
{
"title": "Campaigns",
"spec": {
"version": 1,
"query": {
"kind": "sql",
"sql": "WITH s AS ( SELECT anonymous_id AS visitor, day, if(channel_type != '', channel_type, 'Direct') AS channel, if(utm_source != '', utm_source, '(none)') AS utm_source, if(utm_medium != '', utm_medium, '(none)') AS utm_medium, if(utm_campaign != '', utm_campaign, '(none)') AS utm_campaign, pageviews AS pv, custom_events AS ce, revenue AS session_revenue FROM sessions ) SELECT utm_source, utm_medium, utm_campaign, uniqExact(visitor) AS visitors, count() AS sessions, if(countIf(pv > 0) = 0, 0, round(100 * countIf(pv = 1 AND ce = 0) / countIf(pv > 0), 1)) AS bounce_pct, if(countIf(pv > 0) = 0, 0, round(sum(pv) / countIf(pv > 0), 2)) AS pages_per_session, round(sum(session_revenue), 2) AS revenue FROM s WHERE utm_campaign != '(none)' OR utm_source != '(none)' GROUP BY utm_source, utm_medium, utm_campaign ORDER BY sessions DESC, utm_source, utm_medium, utm_campaign LIMIT 50"
},
"view": {
"mark": "x-table"
}
}
}Sources: Manalens dashboards documentation.
How do you build and read the "Short link clicks by creator and campaign" chart?
It shows the clicks on each short link, with the creator tag, the UTM campaign, the slug and the host of the target, for example store.steampowered.com or an itch.io page. Each open of a short link records one link_clicked event before the visitor goes on to the store.
How to read it: a click is not a sale. It is the only number that you get within minutes for a link to a store page, because the store does not run your analytics. Link preview bots that send a HEAD request are not counted.
What to do: find the creators with many clicks here, then check their wishlists and purchases in the Steam view, grouped by campaign and content. Thank the creator who performs and ask for a second video in the same week.
- For "Short link clicks by creator and campaign": open your launch week dashboard, click Add chart, then Blank chart. Set Chart title to "Short link clicks by creator and campaign" and Chart type to Table.
- Copy the text of "sql" from the spec of "Short link clicks by creator and campaign" below and paste it into SQL query.
- A table shows every column of the result, so there is no column to choose. Click Add chart.
Chart spec for the create_chart tool of the MCP server (add your boardId)
{
"title": "Short link clicks by creator and campaign",
"spec": {
"version": 1,
"query": {
"kind": "sql",
"sql": "SELECT if(JSONExtractString(properties,'creator') != '', JSONExtractString(properties,'creator'), '(none)') AS creator, if(JSONExtractString(properties,'campaign') != '', JSONExtractString(properties,'campaign'), '(none)') AS campaign, JSONExtractString(properties,'slug') AS slug, JSONExtractString(properties,'target_host') AS target_host, uniqExact(message_id) AS clicks FROM raw_events WHERE type = 'track' AND event = 'link_clicked' AND received_at >= toDateTime({{date_from}}) AND received_at < toDateTime({{date_to}}) + INTERVAL 1 DAY AND JSONExtractString(context,'device') != 'bot' GROUP BY creator, campaign, slug, target_host ORDER BY clicks DESC, slug LIMIT 50"
},
"view": {
"mark": "x-table"
}
}
}Sources: Manalens short links documentation.
How do you build and read the "New players by country" chart?
It shows the 15 countries with the most new players in the date range. The collector sets the country from the IP address of the request and then drops the address. It never stores the IP address.
How to read it: compare the list with the languages of your store page and your game. A country in the top rows that has no translation is a possible next language. A player on a VPN counts in the country of the VPN.
What to do: if one country is much larger than you planned, check its regional price and its reviews before you plan a translation.
- For "New players by country": open your launch week dashboard, click Add chart, then Blank chart. Set Chart title to "New players by country" and Chart type to Table.
- Copy the text of "sql" from the spec of "New players by country" below and paste it into SQL query.
- Click Add chart. The table lists country and new_players, the largest country first.
Chart spec for the create_chart tool of the MCP server (add your boardId)
{
"title": "New players by country",
"spec": {
"version": 1,
"query": {
"kind": "sql",
"sql": "SELECT if(JSONExtractString(context,'country') != '', JSONExtractString(context,'country'), 'Unknown') AS country, uniqExact(anonymous_id) AS new_players FROM raw_events WHERE type = 'track' AND JSONHas(context,'app') AND channel != 'server' AND event = 'game_open' AND JSONExtractBool(properties,'is_first_open') AND anonymous_id != '' AND received_at >= toDateTime({{date_from}}) AND received_at < toDateTime({{date_to}}) + INTERVAL 1 DAY AND JSONExtractString(context,'device') != 'bot' GROUP BY country ORDER BY new_players DESC, country LIMIT 15"
},
"view": {
"mark": "x-table"
}
}
}Sources: Manalens app events and dashboards documentation.
How do you read "Steam view: visits, wishlists and purchases by UTM"?
It shows the visits, tracked visits, wishlists and purchases that Steam counted for your tagged links, by day and by one UTM field, with the wishlist rate and the purchase rate. The data is the UTM Analytics report of Steamworks. You download it as a CSV file and import it.
How to read it: Steam counts a wishlist or a purchase only within 72 hours of a visit by a person who is logged in to Steam in that browser. So these numbers are a part of all wishlists and sales, often a small part. Compare links with each other. Do not read the numbers as totals.
What to do: move effort to the sources with the best purchase rate, not the most visits. Steam finalizes conversions 4 days after the visit, so import the launch week again after day 7 before you judge a creator.
- In Steamworks, open your game, then Marketing & Visibility, then the UTM Analytics tab. Choose the dates of the launch week and download the .csv file.
- In Manalens, open Settings, then Integrations. Under Steam, choose the CSV file, click Preview, check the column mapping and click Import.
- Open Overview, open the source picker and choose Steam. Set Group by to source, campaign or campaign and content.
Sources: Manalens Steam documentation; Steamworks documentation: UTM Analytics.
How do you read "Steam view: all wishlists and sales by day"?
It shows all wishlist adds, deletes, the net change and wishlist purchases per day, and the net units and net sales in USD per day. The numbers come from the Financial Web API of Steamworks, so they include the people who never clicked a tagged link.
How to read it: the automatic Steam sync is in early access, so check the first totals against Steamworks. It runs once a day, so this is the number of yesterday, not of this hour. Wishlist days are GMT days and sales days are Pacific Time days, as Steam reports them. Net sales are sales less returns and taxes, before the Steam revenue share. They are not your payout. Returns are inside the net numbers: the view has no separate refund line.
What to do: compare net units per day with your launch goal on day 3 and on day 7. For the hour by hour picture of release day, and for refunds as their own number, use the sales reports of Steamworks.
- In Steamworks, open Users & Permissions, then Manage Groups, and create a Financial API Group. Copy the Web API key from the group page.
- In Manalens, open Settings, Integrations, Steam. Paste the key and your app IDs and click Connect Steam. This needs the owner or admin role and an MFA session.
- Click Sync now. The first sync reads the last 90 days. Then open Overview and choose Steam in the source picker.
Sources: Manalens Steam and revenue documentation; Steamworks documentation: IPartnerFinancialsService.
How do you read "itch.io view: views, downloads and payments by day"?
It shows the views, downloads, payments and gross revenue of your itch.io game for each day, with marks on the days that a new build went up. It also shows the game events from builds with the storefront itch.
How to read it: itch.io gives only totals since the game was made, so Manalens keeps the total of each day and shows the difference. Daily numbers start on the day after you connect. Connect before release day, or the launch days show 0. The revenue is gross: the share of itch.io and payment fees are not subtracted.
What to do: compare downloads per view on the days around a post or a bundle. The channel of a payment shows only for buyers whose email your server sent, so treat that part as a sample.
- In your itch.io account settings, create an API key under API keys.
- In Manalens, open Settings, then the Integrations tab. In the itch.io section, paste the key, click Connect itch.io and choose your game.
- Open Overview, open the source picker and choose itch.io.
Sources: Manalens itch.io integration documentation.
What data does the launch week dashboard need first?
Set up four things before release day: the SDK in the release build, tagged links, the Steam data and, if you sell there, the itch.io connection. Each chart in the list names the data that it reads, so a missing step shows as an empty chart.
The steps below are in the order that we would do them. Do them with a test build one or two weeks before release, and check each chart with your own test sessions.
- Add the Godot or Unity SDK to the release build. It sends game_open, session_start, session_end and app_unclean_exit for you. For another engine, send the same events to the events endpoint over HTTP.
- Give the SDK the app_version, the build and the storefront (steam or itch) of each upload. The build charts and the itch.io view read these values.
- Send game errors as error events. The Godot SDK sends engine and script errors on Godot 4.5 and later, and the Unity SDK sends unhandled exceptions. Use track_error for your own cases, such as a failed save.
- Turn the SDK off in the editor, so that your own test runs do not count as players or crashes in launch week.
- Put the web SDK on your game website, so that Sessions by channel and Campaigns have page views to read.
- Tag every Steam link with utm_source or utm_campaign. Steam records only links to your app page or a sale page that have one of the two.
- Create one short link for each creator and each post under Links, and fill in the Creator field. The short link adds the UTM fields and records the click.
- Connect Steam: import the UTM Analytics CSV, and add a Steamworks Financial API key (the automatic Steam sync is in early access) if you want all wishlists and sales by day.
- If you sell on itch.io, connect the itch.io API key before release day. If you also sell from your own website with Stripe or Paddle, connect it under Integrations: each payment becomes a purchase event and each refund a refund event.
- Game tracking plan generator →
- Steam UTM link builder →
- Steam and itch.io launch checklist →
- Stripe and Paddle revenue setup →
Sources: Manalens app events, errors and crashes, short links, Steam, itch.io and revenue documentation.
What is the daily routine with the launch week dashboard?
Look at the dashboard at fixed times, and decide only on the days in the table. A launch has more numbers than decisions. The routine keeps the two apart.
The times are for a team in one time zone. Dashboard days are UTC days, Steam wishlist days are GMT days and Steam sales days are Pacific Time days, so do not compare one day across them to the hour.
| When | What to check | What to do |
|---|---|---|
| Release day, first 2 hours | Daily active players, Crash-free sessions, Top errors | Confirm that new players arrive and that events come in. Decide on a hotfix if one error hits many players. |
| Every morning | Daily active players, Players by day and build, Errors per session by build | Write down new players and the newest build. Open a bug for each new top error. |
| Every afternoon | Sessions by channel, Campaigns, Short link clicks by creator and campaign | Find the channel or creator that grew. Send one follow-up post or message. |
| After each patch | Players by day and build, Errors per session by build, Crash-free sessions | Check that players moved to the new build and that its error bar is lower. |
| Day 2 | D1 retention, Install cohorts | Compare D1 of the launch day cohort with your demo. If it is much lower, study the first session. |
| Every day, after the Steam sync | Steam view: all wishlists and sales by day (Early access); itch.io view | Compare net units with your goal. Compare new players with units sold. |
| Day 4 and day 7 | Steam view: visits, wishlists and purchases by UTM | Import the UTM CSV again. Steam finalizes conversions 4 days after the visit. |
| Day 7 | All charts, with the date range on the launch week | Export the CSV of each chart, write the launch review and change to a weekly review. |
Sources: Manalens dashboards and Steam documentation.
What are the common mistakes with a launch week dashboard?
These are the mistakes that we think cost the most time. This list is our advice, not a result from a study.
- To build the dashboard on release day. Build it with a test build before release, so that an empty chart means a real problem.
- To ship every upload with the same build value. Then the build charts show one color, and you cannot see whether a patch helped.
- To leave the SDK on in the editor. Your own test runs then count as players, sessions and crashes.
- To read new players as copies sold. The chart counts installs that sent events. Sales are in the Steam view and in Steamworks.
- To read Steam UTM wishlists as all wishlists. Steam attributes only a part, so compare links with each other.
- To judge a creator on day 1. Steam finalizes UTM conversions 4 days after the visit, and you must import the CSV again.
- To trust Crash-free sessions in the first hour after a patch. A crash is counted at the next start of the game.
- To decide on the D1 value of one small cohort. Wait for the cohorts of day 2 and day 3 before you change the game.
- To add every chart you can think of. A dashboard holds up to 24 charts, but each extra chart makes the morning check longer.
Sources: Manalens dashboards, errors and crashes and Steam documentation.
What can the launch week dashboard not show?
Some launch numbers are not in a dashboard chart, and some are not in Manalens at all. A SQL chart reads only your events and sessions (the tables raw_events, sessions and session_events). Store data has its own reports. The table is the state as of 19 November 2026.
Yes means that the data is in the product, but not in a dashboard chart. Not yet means that the product has the data and no chart or line for it. No means that the product does not have the data.
| Data | In Manalens | Where to get it |
|---|---|---|
| Steam wishlists and sales as charts on this dashboard | Not yet. They are in the Steam view, and Steam net sales are in the Revenue metric of Overview. | The Steam view, next to the dashboard |
| Steam sales of the current hour | No. The automatic Steam sync (in early access) runs once a day. | Steamworks sales reports |
| Steam refunds as their own number | Not yet. The automatic Steam sync (in early access) reads returns, and the Steam view shows net units and net sales only. | Steamworks sales reports |
| Wishlists by source for every wishlist | No. Steam attributes only wishlists within 72 hours of a tagged, logged-in visit. | No tool has this. Compare tagged links with each other. |
| Steam store impressions and traffic by Steam page | No. The product does not import the Steam traffic report. | Steamworks Store and Platform Traffic Reporting |
| Steam reviews and concurrent players | No. | Our free Steam launch week tracker, or your store page |
| Who bought on Steam | No. Steam gives totals with no buyer, so Steam sales are not in user profiles. | Not available from Steam |
| itch.io refunds, referrers and impressions | No. They are not in the itch.io API that the sync reads. | Your itch.io dashboard |
| Stripe and Paddle refunds | Yes. Each refund is a refund event, and the Revenue report shows gross, refunds and net. | The Revenue report |
| Installs by channel, campaign and creator | Yes. The Installs tab of the Attribution page matches a first game_open to a website visit. | The Attribution page |
| Native crash dumps and symbols | No. The product counts crashes and stores script errors. | A crash reporter such as Sentry or Backtrace |
Sources: Manalens Steam, revenue, itch.io, install attribution and errors documentation; Steamworks documentation: UTM Analytics and Store and Platform Traffic Reporting.
Which dashboard templates go with the launch week dashboard?
Each dashboard template has one job. The published templates are linked below.
The templates in the list below are planned. They have no page yet, so they have no link.
- Retention cohorts dashboard (planned)
- Creator campaign dashboard template for Steam games →
- Game analytics dashboards: which charts matter →
Sources: Manalens content plan.
How do I start my launch week dashboard?
Manalens is the analytics and growth engine built for games. Open Dashboards, click Blank dashboard and name it Launch week. Create the Game KPIs, Game health and Channels and campaigns boards from their templates and move the eight charts over. Then add the three SQL charts with the specs on this page, or give the specs to your AI tool through the MCP server.
Sources: Manalens product documentation.
Frequently asked questions
Which charts do I need in the launch week of a Steam game?
You need 11 charts in three groups. Players: daily active and new players, players by build, D1 retention and install cohorts. Health: crash-free sessions, errors per session by build and top errors. Traffic: sessions by channel, campaigns, short link clicks and new players by country. Steam wishlists and sales are in the Steam view.
Can I put Steam wishlists and sales on a Manalens dashboard?
Not yet, as of 19 November 2026. A dashboard chart reads your own events and sessions. Steam wishlists and sales come from the Steam UTM report and the Steamworks Financial API (the automatic Steam sync is in early access), and they show in the Steam view of Overview. Steam net sales are also part of the Revenue metric of Overview.
Do I need SQL to build the launch week dashboard?
No. Eight of the 11 charts come from the Game KPIs, Game health and Channels and campaigns templates, and you move them to one dashboard. For the other three, you paste the SQL from this page into a blank chart. An AI tool can also save each spec through the MCP server.
How often should I check the dashboard in launch week?
Check it two times a day and after each patch. Look at players, builds and errors in the morning, and at channels, campaigns and short links in the afternoon. Read the Steam view after the automatic Steam sync (in early access). Decide on retention from day 2 and on creators after day 4, when Steam finalizes conversions.
Does the dashboard work for an itch.io launch?
Yes. The 11 charts read the events of your build and your website, so they work for any store. Set the storefront of the build to itch. The itch.io view of Overview adds views, downloads, payments and gross revenue by day, from the day after you connect the itch.io API key.