Quick answer: Record a baseline of daily report volume, crash rate, and top categories for two weeks before the sale. Set alerts at 3x baseline daily volume and 20% concentration in any single error category. Schedule rotating team coverage across evenings and weekends. Pause risky feature work but keep shipping critical fixes. The goal: catch new-player-specific issues fast because they disproportionately become Steam reviews.
You wake up on the first day of the Steam Summer Sale and your bug tracker has 400 new reports where it normally has 20. The triage channel is on fire. Some reports are duplicates of known issues. Some are new bugs. A few are terrifying crash spikes from hardware configurations you’ve never seen. Your team is panicking. This is the normal experience of a successful indie game during a Steam sale — and you can prepare for it.
Why Sale Periods Are Different
Sale periods change the composition of your player base in ways that affect bug reporting:
- New players: Sale buyers are disproportionately first-time players who don’t know workarounds. They hit every onboarding bug.
- Hardware diversity: Sale buyers span the full Steam hardware distribution, not just your usual audience. Expect bugs on GPU models you’ve never seen.
- Lower commitment: A player who paid $5 in a sale refunds faster than a player who paid $25 at launch. Bugs that cause refunds cost you more per-bug.
- Concentrated time windows: Most sale purchases play in the first few days after buying. Your bug report spikes follow the same shape.
- Review pressure: Players who encounter bugs during a sale are more likely to leave negative reviews than players at other times, because the sale itself created the expectation of value.
Step 1: Build a Pre-Sale Baseline
Two weeks before any major sale, start recording your baseline metrics. You need numbers to compare against, otherwise you can’t tell what counts as a spike.
Record daily:
- Total bug report count
- Unique reporters (deduped by user ID or email)
- Crash-free session rate
- Top 5 error categories by volume
- Hardware distribution (GPU, OS, RAM brackets)
- Reports per 1,000 active players (normalizes for player count)
The “reports per 1,000 active players” metric is the most useful because it controls for player count. A flat rate means your game got more popular but quality stayed the same. A rising rate means something got worse.
Step 2: Set Up Volume and Category Alerts
Alerts let you sleep at night. Configure them to fire when specific thresholds are crossed so you hear about spikes within minutes instead of days.
// Example alert rules (pseudocode)
// Alert 1: Total daily volume above 3x baseline
{
name: "Daily volume spike",
query: "bug_count_24h > baseline_avg * 3",
cooldown: "6h",
channel: "#bugnet-alerts"
}
// Alert 2: Any single category dominating reports
{
name: "Category concentration",
query: "max(category_count_1h) / total_count_1h > 0.2 AND total_count_1h > 20",
cooldown: "2h",
channel: "#bugnet-alerts"
}
// Alert 3: Crash-free session rate dropped
{
name: "Stability regression",
query: "crash_free_rate_1h < baseline_crash_free_rate - 0.01",
cooldown: "1h",
channel: "#bugnet-oncall"
}
// Alert 4: New hardware category appearing
{
name: "Novel GPU reports",
query: "first_seen_gpus_in_reports_1h > 5",
cooldown: "12h",
channel: "#bugnet-alerts"
}
Tune the cooldowns to avoid alert spam. You want to hear about the initial spike, not a ping every 5 minutes for the next six hours.
Step 3: Schedule Coverage
Sale spikes hit hardest on weekends and evenings — exactly when your team isn’t normally watching. Schedule explicit rotating coverage for sale weeks, with clear handoff times and on-call responsibilities.
A minimal coverage plan for a two-week sale:
- Day watch: One person monitors triage 9am-6pm their local timezone
- Evening watch: A different person covers 6pm-11pm (usually a couple hours of active triage)
- Weekend day watch: Split Saturday and Sunday between two people, paid overtime or comp time
- On-call: One person with phone notifications for the P0-critical alerts
Rotate these roles daily so no one person burns out. Pay for the coverage — this is work, and unpaid sale week duty is how you lose good engineers.
Step 4: Freeze Risky Changes
Before the sale starts, declare a code freeze on feature work. You still ship bug fixes, and you still respond to critical issues. But you don’t ship a new boss encounter or a refactor of the save system during a sale week. The risk of a regression is amplified by the number of players who would encounter it.
Communicate the freeze clearly to your team and stick to it. “We’ll ship that feature right after the sale ends” is a perfectly fine decision and prevents the scenario where you introduce a bug that affects 5,000 new players in the middle of their first session.
What to Actually Do When the Alert Fires
Your volume alert fires. Now what? Here’s a triage flow:
- Is it one bug or many? Check the category distribution. If 90% of reports are the same bug, you have a single problem to fix. If it’s spread across categories, it might just be volume.
- Is it new or known? Cross-reference the top categories against your known-issues list. A new top category is more urgent than a known one reaching higher volume.
- Is it hardware-correlated? Check the hardware distribution of reporters. A bug that affects only AMD GPUs or only Intel integrated graphics is specific enough to fix fast.
- Is it causing refunds? Check your Steamworks refund page. If refund rate is spiking, the bug is affecting purchase decisions directly and needs an emergency hotfix.
- Is it in reviews yet? Check recent Steam reviews. A bug that appears in 3+ negative reviews needs to be the first thing you fix.
Post-Sale Retrospective
After the sale ends, do a retrospective. What bugs spiked? Which ones did you fix in time? Which ones became review fodder? What did you learn about your new-player experience? This retrospective feeds your preparation for the next sale. Over time, your team gets very good at handling these windows with minimal stress because the playbook is clear.
The ultimate goal is that a Steam sale becomes a routine event you handle calmly, not a crisis. Studios that survive and thrive on Steam usually started by having one terrible sale experience and then building the process that prevented it from happening again.
Related Issues
For Steam-specific bug capture, see How to Set Up Bug Reporting for Steam Games. For Steam review handling, check How to Handle Bug Reports From Steam Reviews. For Steam Next Fest prep which shares many characteristics, read Steam Next Fest Bug Preparation Checklist.
A sale is a stress test. Pass it and you earn months of positive reviews.