An attribution model in Google Analytics determines which interactions get credit when a user converts after multiple touchpoints. If someone clicks a Google ad, returns via organic search, then converts through a direct visit, the attribution model decides how much credit each channel receives for that conversion. Google Analytics 4 uses data-driven attribution as the default model. This machine-learning approach analyzes converting versus non-converting paths in your actual data to assign fractional credit where it statistically matters. If your paid search clicks show stronger conversion correlation than your display impressions, data-driven attribution weights them accordingly. You need at least 400 conversions per month and 10,000 ad interactions for this model to function properly, otherwise GA4 falls back to last-click. Universal Analytics offered seven models: last click, first click, linear, time decay, position-based, and data-driven. Most businesses defaulted to last-click because it was simple, but it systematically undervalued top-of-funnel efforts like content marketing or initial brand searches. First-click did the opposite, ignoring what closed the deal. Linear split credit equally across all touches, which sounds fair but treats a passing blog visit the same as a demo request. At Ottawa SEO, we typically see 15-35% attribution shifts when clients move from last-click to data-driven models in GA4. SEO and content often gain 8-20% more attributed conversions because data-driven recognizes their role in initiating journeys that paid search later converts. Conversely, direct traffic usually loses 10-25% because it gets inflated credit under last-click when it's often just the final step. The practical issue is that attribution models only track what Google can see. If someone researches on mobile, converts on desktop, or clears cookies, you get fragmented paths. Cross-device reporting helps but isn't perfect. Also, attribution windows matter: GA4's default 90-day window for conversions means older touchpoints can still claim credit, which is realistic for B2B but potentially too long for impulse e-commerce. You can adjust this in your conversion settings based on your actual sales cycle length.