Click Tools & Settings in your Google Ads account, navigate to Measurement, then Attributions. From there, select the conversion action you want to modify, click Edit Settings, and choose your attribution model from the dropdown. Google offers six models: Last Click (default for many older accounts), First Click, Linear, Time Decay, Position-Based, and Data-Driven. Data-Driven attribution became the automatic default in 2022 for accounts with sufficient conversion data (typically 3,000+ ad interactions and 300+ conversions in 30 days). Your choice fundamentally changes how credit gets distributed across touchpoints. Last Click gives 100% credit to the final ad interaction before conversion—simple but ignores earlier research clicks. Data-Driven uses machine learning to assign fractional credit based on actual contribution, usually the smartest option if your account qualifies. Linear splits credit evenly across all clicks, which sounds fair but often overvalues incidental touches. Position-Based gives 40% each to first and last clicks, 20% to middle touches—decent for longer consideration cycles. Time Decay favours recent interactions, useful for time-sensitive offers. Critical point: changing your attribution model doesn't change what actually happened, it changes how you measure what happened. If you switch from Last Click to Data-Driven, campaigns that drove early awareness suddenly show more conversions, while bottom-funnel remarketing might show fewer. Your total conversions reported will stay the same, but distribution across campaigns shifts. This creates an apples-to-oranges problem when comparing historical performance. At Ottawa SEO, we typically run Data-Driven for clients with enough volume and review the Attribution Reports monthly to see how top-funnel keywords perform under different models. For lower-volume accounts (under 300 monthly conversions), we stick with Last Click or Position-Based since Data-Driven won't have enough signal. One practical trap: if you're optimizing manual bids or reporting to clients who expect consistent numbers month-over-month, document the change clearly. A sudden 40% jump in branded campaign conversions after switching models looks like you gamed the numbers unless you explain the methodology shift upfront.