How should I think about omnichannel advertising strategy?

Most organisations still plan advertising by channel — separate budgets, separate teams, separate measurement for paid search, social, and programmatic. That model is structurally broken, not because the channels don't work, but because the customer was never organised that way to begin with. Omnichannel strategy starts from a different premise: the audience moves through a connected ecosystem, and the job is to follow the logic of that ecosystem, not the org chart.

Five fixed truths I build strategy around

1. The channel is not the unit of strategy. The ecosystem is.
TV is no longer TV. Search is no longer just Google. Retail is now a media platform. The strategic question is no longer which channels, but how they connect.

2. Privacy loss is permanent — and it's a forcing function for better strategy.
Signal loss, cookie deprecation, consent frameworks, and platform-level data limits permanently changed what visibility is available. The right response is first-party data infrastructure and measurement that doesn't depend on individual-level tracking.

3. AI is the new orchestration layer — not just an optimisation tool.
Media buying is moving from manual optimisation toward machine-assisted orchestration. The more important question is how humans govern that layer — what inputs it receives, what constraints it operates within.

4. Suppression is as important as targeting.
Intentional audience exclusion — suppressing already-converted customers, low-quality audiences, or out-of-market users — is one of the most consistently undervalued efficiency levers in paid media.

5. Signal quality matters more than scale.
Broad reach with degraded signal produces noise. Smaller, higher-quality audiences with cleaner data and tighter intent consistently outperform on business metrics.

​

 

What the best operators are doing right now

Amazon has effectively dissolved the line between commerce and advertising — simultaneously a retailer, media company, streaming platform, and commerce engine. Spotify rebuilt the remnants of radio around data, with dynamic ad insertion and mood/context targeting now table stakes. L'Oréal runs creator integration, retail media, and AI personalisation as one connected system rather than separate initiatives — the operating model most organisations are still trying to reach.

​

Practical questions I use to assess omnichannel maturity

 

• Are channels measured independently, or against shared business outcomes?

• Is there a first-party data strategy that doesn't depend on third-party signal?

• Is suppression logic built into audience architecture, or an afterthought?

• Is AI being used as a tool (reactive) or an orchestration layer (proactive)?

• Is retail media integrated into the broader media plan, or activated separately?

• Is measurement designed around the customer journey, or around platform-reported metrics?

​

​