Part 4 — Google captures, Meta and LinkedIn create
Demand capture before demand creation, the realistic role of each channel, and how to read cross-channel effects.
No single channel runs a SaaS acquisition system. Search harvests demand that already exists — finite, priced accordingly, and the correct first dollar. Social creates demand that did not exist — slower, compounding, and the only way past search’s ceiling. The failure mode is funding them in the wrong order or measuring them by the same yardstick.
The funding order
- Google category and brand-defense terms first — the highest-intent demand available, and the baseline every other channel gets measured against.
- Microsoft Ads second, not last: 10-30% additional search volume at typically lower CPCs, an older and more enterprise-weighted audience, and near-free account portability from Google. The case is in the Microsoft Ads argument.
- Meta and LinkedIn for demand creation once capture saturates — visible as impression share maxing out while marginal CAC climbs. LinkedIn buys precise B2B targeting at brutal CPCs (works at $15k+ ACV); Meta buys reach and creative volume at costs that work well below that.
- Retargeting last and smallest: with long B2B cycles it is genuinely useful, and also the easiest budget line to over-fund because it looks efficient by construction.
Demand creation runs on creative, not targeting
On Meta especially, targeting has consolidated into the algorithm and creative volume became the new targeting: the ad itself finds the audience. For SaaS this means problem-narrative creative — the pain named plainly, the before/after shown — outperforms feature tours nearly always, and a systematic testing cadence outperforms sporadic brilliance. Budget creative production as a real line item or skip demand creation entirely.
Reading the interplay
Social-created demand frequently converts through branded search days later, which means last-click reporting bills social’s wins to Google. Watch branded search volume against social spend phases, judge demand creation on blended CAC trend over quarters (not in-platform ROAS over weeks), and when the numbers disagree, remember which one is attribution-proof. Part 5 builds this measurement properly.