Pre-order conversions jumped 40% on Friday but we didn't change any campaign settings. What actually drove it?
I analyzed all variables across your iPhone 16 campaigns from Thursday to Saturday. Here's the breakdown:
Primary cause: Samsung pulled Galaxy S25 launch ads (58% of the lift)
Samsung paused their Galaxy S25 Ultra comparison campaigns across Google Search Friday morning — likely a budget reset ahead of their fiscal quarter close. Their impression share on "best smartphone 2025" and related terms dropped from 38% to 4%. Your iPhone ads captured that demand directly. I confirmed this by comparing your impression-to-conversion rate on competitive keywords against historical periods when Samsung's share was similarly low.
Secondary cause: Carrier financing promo timing (31% of the lift)
T-Mobile and Verizon both launched new iPhone 16 trade-in promos Thursday evening. Searches for "iPhone 16 trade-in deal" spiked 3.2x, and your Shopping ads on those terms converted at 8.7% vs. your normal 4.1%. This was external demand you happened to catch, not something your campaigns generated.
Not a factor:
• Your Wednesday creative refresh on Meta: Isolated its effect using diff-in-diff. The new "Shot on iPhone" video improved CTR by 0.4pp but had no measurable impact on conversion rate for pre-orders
• iOS 18 media coverage: Organic traffic was flat, suggesting the earned media didn't spill into paid channels
• Weather/seasonality: No meaningful correlation for consumer electronics at this time of year
Implication: The lift is largely non-repeatable — it depended on Samsung going dark and carrier promo timing. I'd recommend against scaling budgets based on Friday's numbers. However, the carrier promo pattern is worth building a trigger for — next time a major carrier launches an iPhone financing deal, we should automatically increase bids on trade-in terms within the hour.
10:04 AM
How confident are you it was Samsung and not just general Friday demand for electronics?
Confidence level: 84%. Here's how I validated the Samsung explanation over the general-Friday hypothesis:
1. Category-level control — I checked your Mac and AirPods campaigns, which don't compete with Samsung. Their Friday conversions were flat (+2%, within normal variance). If this were a general electronics demand spike, those products would have lifted too.
2. Keyword-level decomposition — 73% of the iPhone conversion increase came from keywords where Samsung had >20% impression share the prior week. Keywords where Samsung was never present showed only a 4% lift (noise-level).
3. Historical pattern — Samsung has paused campaigns 4 other times in the trailing 12 months. Each time, your iPhone conversion rate on overlapping keywords rose 25–50%. That consistency across repeated natural experiments is strong causal evidence.
4. What would weaken it — If Apple.com direct traffic also spiked 40% Friday, that would suggest broader iPhone demand unrelated to Samsung. But direct traffic was only up 6%, consistent with normal daily variance.
Samsung re-entry monitoring: I'm tracking their impression share hourly. Based on past patterns, their pauses last 3–5 business days. Expect them back by Wednesday. I'll push an alert to your Slack when their share crosses back above 15%.
10:07 AM
Results are based on causal inference models applied to your campaign data. Always verify critical decisions with your team.