
For a Steam Next Fest demo, track a short decision funnel: session start, first interaction, tutorial completion, first core-loop completion, meaningful checkpoint, session end and crash-free completion. Pair that in-game telemetry with Steamworks traffic, UTM and daily wishlist reports. The goal is to find where players lose confidence and which acquisition sources produce qualified store visits—not to collect every possible event.
Steam Next Fest: October 2026 runs from October 19 at 10:00 AM PDT to October 26 at 10:00 AM PDT. A demo can attract far more players than an ordinary test, but raw player count does not explain whether the build teaches the game, reaches the core promise or motivates a wishlist. This guide defines a compact analytics plan that an indie team can implement before the event and read while the festival is still active.
Last source check: October 6, 2026. Steam dates and reporting behavior are linked to the official Steamworks documentation in the Sources section.
What should a Steam Next Fest demo measure?
Measure four layers separately. Mixing them into one dashboard hides the reason a number changed.
| Layer | Question | Primary signals | Decision |
|---|---|---|---|
| Acquisition | Did the right audience reach the store page? | Trusted UTM visits, tracked visits, source/medium, daily wishlist additions | Keep, change or stop a campaign/source |
| Activation | Did players start and understand the demo? | Session starts, first interaction, tutorial completion | Fix launch friction or onboarding |
| Core value | Did players reach the experience promised on the store page? | First core-loop completion, meaningful checkpoint, retries, session duration | Change pacing, clarity or demo scope |
| Reliability | Could players complete the intended path? | Crash-free sessions, fatal errors, performance samples, blocked progression | Prioritize technical fixes before more traffic |
A 12-event schema for a Next Fest demo
Use stable event names and attach only the properties needed to answer a product question. Record a build version on every event so hotfixes do not become mixed into one cohort.
| Event | When it fires | Useful properties | Question answered |
|---|---|---|---|
demo_session_start |
The first playable scene is ready | build_version, platform, language | How many playable sessions actually began? |
first_interaction |
The player performs the first meaningful input | seconds_from_start, input_type | Is the opening immediately understandable? |
tutorial_start |
The guided onboarding begins | tutorial_variant | How many starters enter onboarding? |
tutorial_complete |
The required onboarding objective is completed | duration_seconds, retry_count | Where does onboarding lose players? |
core_loop_start |
The first real loop begins | mode, level_id | How many players reach the promised activity? |
core_loop_complete |
The first complete action–feedback–reward loop ends | duration_seconds, outcome | Can players experience the core value? |
checkpoint_reached |
A meaningful content or progression milestone is reached | checkpoint_id, elapsed_seconds | How far does each cohort progress? |
challenge_failed |
A designed failure state occurs | challenge_id, reason | Which obstacle creates avoidable friction? |
challenge_retry |
The player retries after failure | challenge_id, attempt_number | Does failure create motivation or abandonment? |
settings_changed |
A player changes graphics, controls or accessibility | setting_group, old_value, new_value | Which defaults create friction? |
demo_session_end |
The demo closes or reaches the designed ending | exit_reason, duration_seconds, last_checkpoint | Where and why do sessions end? |
technical_failure |
A crash, fatal error or progress-blocking failure occurs | build_version, error_group, last_checkpoint | Which failures cost playable sessions? |
Do not send free-form player names, email addresses, Steam IDs or device fingerprints into a product analytics event stream. Prefer aggregate-compatible properties and document the collection in the privacy notice used by the demo.
Seven KPIs that turn events into decisions
1. First-interaction rate
Formula: unique players with first_interaction ÷ unique players with demo_session_start × 100.
A low rate points to launch delay, unclear controls, focus problems, an opening cutscene that is too long, or players leaving before the build becomes interactive.
2. Tutorial completion rate
Formula: unique tutorial_complete players ÷ unique tutorial_start players × 100.
Break this down by build, language and input type before rewriting the entire tutorial. A problem isolated to one language or controller path needs a narrower fix.
3. Core-loop reach rate
Formula: unique players with core_loop_complete ÷ unique session starters × 100.
This is often the most important demo metric. If the store page promises tactical battles, construction, exploration or deck building, the player should reach one complete version of that promise early enough to judge it.
4. Meaningful checkpoint rate
Formula: unique players reaching the selected checkpoint ÷ unique session starters × 100.
Choose one checkpoint that represents an informed experience, not merely time spent. Examples include completing the first strategic turn, making the first build choice or finishing the first full mission.
5. Crash-free session rate
Formula: 1 − sessions with a crash or fatal error ÷ started sessions.
Track this by build version, operating system and graphics tier. Do not increase acquisition while a large technical segment cannot complete the intended path.
6. Median session duration
Use the median rather than only the average. A few extremely long idle sessions can make the average look healthy. Read duration beside the last checkpoint and exit reason: ten minutes that reach the core loop can be more valuable than forty minutes stuck in setup.
7. Wishlist conversion from tracked Steam visits
Formula: Steamworks-attributed wishlist conversions ÷ tracked UTM visits × 100.
Steamworks counts eligible wishlist, purchase or activation conversions within its documented attribution window. It also states that demo downloads from a UTM link to the base game are not currently reported as conversions. Treat Steamworks UTM results and in-game telemetry as two aggregate views; do not attempt to join individual players across them.
How to read the funnel during Next Fest
| Observed pattern | Likely investigation | First action |
|---|---|---|
| Store traffic rises, tracked wishlists do not | Audience fit, capsule/store promise, traffic source | Compare UTM sources and align the page with the demo’s clearest value |
| Sessions start, first interaction is weak | Boot time, focus, input detection, opening pacing | Shorten the path to control and test the affected platform/input |
| Tutorial starts, completion collapses | Instruction clarity, localization, step order, control friction | Inspect the exact tutorial step and replay affected sessions internally |
| Tutorial completes, core-loop completion is weak | The demo delays its main promise or asks for too much setup | Move a complete core loop earlier and remove nonessential gates |
| Core loop is reached, meaningful checkpoint is weak | Pacing, difficulty spike, unclear goals, insufficient feedback | Segment failure and retry behavior by challenge |
| Progression is healthy, wishlists remain weak | Positioning, audience mismatch, demo ending or CTA timing | Review the promise–experience match and the final call to action |
| One build or hardware group has poor completion | Regression, performance or compatibility | Ship a targeted fix and keep pre/post-build cohorts separate |
A practical reporting cadence for October 2026
- Before press preview: validate event names, build version, exit reasons and crash grouping with internal and external test accounts.
- October 8–18: use the press-preview window to find technical blockers and onboarding failures; avoid drawing market conclusions from a small or specialized sample.
- October 19: record a baseline snapshot before the event starts at 10:00 AM PDT and confirm the live demo is sending the expected schema.
- During October 19–26: review reliability frequently, but batch design decisions into a consistent daily window so the team does not react to hourly noise.
- After October 26: freeze the event cohort, document build changes and compare acquisition, activation, core value and reliability as separate layers.
For the production deadlines around these dates, use the Steam Next Fest October 2026 dates and demo checklist. For cohort calculations after the event, use the D1, D7 and D30 game retention calculator. To interpret store interest beyond a single campaign, read the Steam wishlist measurement guide.
Steam Next Fest demo analytics FAQ
Can Steamworks UTM Analytics show demo downloads?
Not as a conversion when the UTM link points to the base game’s store page. Steamworks currently documents wishlists, purchases and activations as conversion types and explicitly notes that standalone demo or DLC downloads are not included as base-game UTM conversions.
Should a small team track every menu click?
No. Track the smallest event set that distinguishes acquisition, activation, core value and reliability. Extra events increase QA and analysis cost without necessarily improving a decision.
What is a good tutorial or core-loop completion rate?
There is no universal threshold that applies across genres, demo lengths and audiences. Establish an internal baseline, segment by build and platform, then judge whether a change improves the intended step without damaging another layer.
Should the team change the demo during Next Fest?
Fix crashes, blocked progression and severe clarity problems when the change can be tested safely. Avoid stacking multiple design changes without build-level measurement; otherwise you will not know which change affected the funnel.
Sources and methodology
- Steamworks: Steam Next Fest October 2026 — event dates, press preview and eligibility.
- Steamworks: UTM Analytics — visits, tracked conversions, privacy behavior and reporting limits.
- Steamworks: Wishlists — wishlist reporting behavior and demo notification context.
The 12-event schema, KPI formulas and decision table are QUBU’s implementation framework derived from product analytics practice. They are recommendations, not Steam requirements.
Continue with the main guides in the same content cluster.
