Meta (Facebook + Instagram), Google (Search + YouTube), and TikTok each play a different role in a healthy GTM engine. The mistake is treating them as interchangeable. Same budget, same creative, same KPI. None of them work hard enough that way.
Meta is the cold-traffic discovery engine. Its targeting layer is rich, its placement variety is wide, and its creative units (Reels, Stories, in-feed video) match the format your organic engine is already producing. Meta is where the organic-validated hooks get amplified to a much wider audience. The KPI is cost-per-qualified-lead or cost-per-sale, not impressions. Roughly 50% of a starter $10K paid budget belongs here.
Google Search is the warm-traffic capture engine. Buyers who search "[your category] near me" or "best [your service] for [their industry]" are already in market. They've been pushed to search by your brand-building activity, by a competitor, by an event, by a recommendation. Google Search captures that intent at the bottom of the funnel and converts at 5x to 15x the rate of cold social. Roughly 30% of the starter budget belongs here, weighted toward your highest-intent commercial keywords.
TikTok is the discovery accelerator. Its algorithm is the most aggressive at pushing new creators and new accounts to fresh audiences, which makes it the cheapest place to test hook variations at speed. TikTok works best when the creative is native to the platform (raw, fast, opinionated, vertical) rather than repurposed from elsewhere. Roughly 20% of the starter budget belongs here, allocated to creative volume rather than spend per ad.
Notice what's missing. LinkedIn for B2B over a certain deal size, YouTube long-form ads for high-consideration purchases, and email/SMS retargeting via Klaviyo or Attentive for ecommerce. These layer in once the first three are stable, typically by month two or three of the partnership. The first $10K is about getting the discovery + capture loop working. Everything else stacks on top.