Cases: how quests increase session time

Quest is a controlled scenario of attention. When the path is divided into clear steps and equipped with "sparks" (small rewards and visual peaks), the player spends more time in the product without a sense of routine. Below are six proven patterns, numbers and patterns that can be applied out of the box.

1) Focus metric: what exactly we grow and how to measure

Basic:

Avg Session Time
Median Session Time (resistance without "whales")
Sessions per DAU
Active Minutes in Mission
Pacing: time to T1/T2/T3, time between sparks

Related Value:

ΔDAU/WAU (stickiness), ΔParticipation_net/Completion, ΔARPPU (net of bonus), Prize & Bonus Cost per Active
Count by cohorts and in control (holdout) to cut off noise.

Events (minimum):

`session_start/end`, `mission_view/join/progress/complete`, `step_reached {t=1..n}`, `micro_reward`, `path_select`, `cool_off_start/end`, `game_switch`.

2) Case # 1. "Ladder" T1-T3 with micronoms every 5-8 minutes

Idea: a chain of 3-4 steps, where every ~ 5-8 minutes the player receives a small "spark": freespins, bonus dust, badge.

Result (synthetic, close to reality):

Avg Session Time: +14–19%
Median Session Time: +11%
Completion: + 9 pp
Prize & Bonus/Active: + €0.45 (budgeted)

Design:

T1 light (2-4 min), T2 "flow" (6-10 min), T3 final "jerk" (3-5 min)
Progress bar with divisions, micro-loot on the T1/T2
Tempo tips: "Up to T2 ~ 6 rounds of €0.5"

Rule Template (YAML):

yaml mission: ladder_v1 steps:
- id: T1; goal: points>=100; reward: {type: "fs", value: 5}
- id: T2; goal: points>=300; reward: {type: "bonus_cash", value: 1}
- id: T3; goal: points>=600; reward: {type: "badge", rarity: "rare"}
caps: {per_bet: 80, per_minute: 220, per_day: 1500}
pacing_hint: true

3) Case # 2. OR goals ("choose a path") instead of a linear grind

The idea: the player can close the step in different styles - turnover, multiplier or number of rounds.

Result:

Avg Session Time: +12–16%
"Switch-rate" paths: 18-27% (signal that the choice is alive)
Complaints/1k: − 22% (less than "can't catch up")

UX:

Three buttons on the mission screen: "Fast," "Competition," "With a plot"
Hint: "120 points left or 8 rounds or × 20 once"

SQL sketch (share of time in mission):

sql
SELECT AVG(EXTRACT(EPOCH FROM (mission_focus_end - mission_focus_start))/60) AS minutes
FROM mission_focus
WHERE mission_id =:m AND date = CURRENT_DATE;

4) Case # 3. Timebox 20 minutes + "breathing window"

Idea: limit the active phase of the step to 20 minutes, then give the "cool-off" 30-60 minutes without losing progress/streak.

Result:

Avg Session Time: + 9-13% (due to returns)
Sessions per DAU: +0,18–0,25
Fatigue (early exits) − 17%

Communication:

"Game window 20 min. After - pause without penalty, you will continue from the stop."

5) Case # 4. Story thread with "rhythm switches"

Idea: we alternate between "fast" and "interesting" tasks (mini-riddle, new game/mechanics). No longer, otherwise.

Result:

Median Session Time: +15%
CTR on "continue the plot": + 26%
ΔARPPU (net): +€1,1 при Prize&Bonus/Active +€0,5

UX patterns:

Comic screen between T2 and T3 (≤10 sec)
"Try a new game - progress will continue"

6) Case # 5. "Menu of the day" (3 quests to choose from) + weekly rotations

Idea: daily menu: easy/competitive/story quest. Rotation of games once every 24-48 hours.

Result:

Avg Session Time: +8–12%
Content diversity ratio: + 34%
Same complaints − 31%

Economics:

Mass quests - cheap rewards (FS/bonus dust); finishers are part of a 10-20% pool no-vager.

7) Case # 6. Pity timer and guaranteed sparks in cold areas

The idea: if there is no progress for 8-10 minutes, the chance of a micro-reward increases or a one-time boost is given × 2 to points (within budget).

Result:

Session interruptions − 14%
Time to rage-quit + 11%
Fraud flags - no growth (due to anti-farm restrictions)

Protection:

Pity is activated only with real attempts, cap in frequency and cost, excluding "ideal" micropatterns.

8) Telemetry: how to catch "time" and "rhythm"

Events:

`focus_start/end`, `step_reached`, `pity_triggered`, `cool_off_start/end`, `path_select`, `micro_reward`.

Dashboard metrics:

Avg/Median Session Time
Time-to-T1/T2/T3, spark interval
Switch-rate paths, Game-diversity index
Early-exit rate (in the first 5 minutes)
Δ ARPPU (net), Prize & Bonus/Active, complaints/1k

9) A/B design and robustness of results

Unit: user, sticky-assignment, stratification (payer/geo/platform)

Holdout 10–20% или randomized invitation

Guardrails: RG-actuations, fraud flags, SRM

CUPED: pre-session-time and pre-ARPPU as covariates

Window: minimum 2 weeks (to capture weekend), D0-D2, D3-D7, D8 + phase analysis

10) Awards Economy: To Keep Time Growth From Eating Up Margins

Small "sparks" every 5-8 minutes: €0.02- €0.08 equivalent on average

Finishers: 10-30% Unlisted Cache from Pool + Rest in FS/Bonus Cache

Control KPI: Prize & Bonus Cost per Active, Δ ARPPU (net), Net Uplift

[
\ text {Net Uplift} =\text {Incremental Revenue} - (\text {Prize} +\text {Bonus} +\text {Ops} +\text {Fraud})
]

11) UX patterns that increase "pain-free time"

"120 points left ≈ 8 rounds of €0.5" - specifics reduce anxiety

One goal per screen, progress-stairs, micro-animations on "sparks"

Soft anti-repeat: "Change the game? Progress will continue"
Quiet hours and frequency notification limits (do not burn out attention)

12) Antifraud and fair play

Cap points per bet/min/hour/day, minimum bet variance

Detection of headless/proxy/duplicate device-fp
Hold-and-review top prizes + KYC L2
Suppression of "glasses farmers" in CRM personalization

13) RG/compliance: "time" without overheating

Time and deposit limits, pause-timer, self-exclusion

Communications without deposit pressure; tone "invite, not coerce"
Visible RG links on mission and reward screens

14) Checklist for launching quests "for time"

Stairs T1-T3 with micronoms every 5-8 min
OR goals and "menu of the day," rotations 24-48 hours
Timebox 20 min + "breathing window" without penalty
Pity timer with cost and frequency caps
Transparent tempo and "how much to target" prompts
Time & Rhythm Dashboard + A/B with CUPED
Economy: cheap "sparks" + part of a no-vager at the finish line
Antifrod: traps, dispersion, KYC, suppression
RG tools and quiet hours

15) Consolidated Mini Case (Synthetic)

Context: 2 brands, 3 geo, 8 weeks, holdout 15%. Launched: "Ladder," OR-goals, timebox + cool-off, menu of the day, pity-timer.

Results vs control:

Avg Session Time +15. 7%, Median +13. 2%
Sessions per DAU +0. 22
Participation_net +7. 1 pp, Completion + 11. 6 p.p.
ΔARPPU (net) +€2. 0, Prize&Bonus/Active +€0. 7 (in plan caps)
Complaints/1k − 29%, fraud flags <1% of PF

Conclusion: we scale OR quests and "ladders" for seasonal campaigns, leaving pity limited, and "sparks" super cheap.

Quests increase the time in the session not by magic, but by rhythm and clarity: short stages, frequent "sparks," choosing a path, pauses without a penalty and an easy plot. Add strict mouthguards, honest economics and RG - and get an Avg/Median Session Time boost along with net increment rather than burnout and complaints.