Week of March 18

AI Is Changing What Counts as a Customer— and What Counts as Performance

Insight #1

When AI Becomes the Shopper, Who Verifies the Buyer?

Agentic commerce introduces a new layer of risk: performance that looks real—but isn’t

What's the News:

World (Sam Altman’s identity startup) launched a tool to verify that AI shopping agents are acting on behalf of real humans—anticipating a surge in AI-driven browsing and purchasing behavior.

As AI agents begin to research products, compare options, and even complete transactions, this tool aims to distinguish real human intent from automated activity.

Why It Matters:

Bots have inflated ad engagement for years—but agentic commerce takes this further.

AI agents can now mimic high-intent behaviors across the full journey, from site visits to conversions. That means marketers may see strong performance across traffic and conversion metrics that isn’t tied to real customers.

More activity ≠ more demand.

This creates a new measurement challenge: performance signals that look legitimate—but don’t translate to real revenue or incremental growth.

Silverback's POV

Marketers Should:

Prioritize incrementality testing to separate real demand from artificial activity

Anchor reporting in CRM and revenue data—not platform-reported metrics

Invest in outcome-level validation (server-side tracking, first-party data)

Pressure-test performance spikes to confirm they reflect real customer behavior

Avoid optimizing toward engagement or conversion signals without validating business impact

Insight #2

Google Personalizes Search—At the Individual Level

AI Mode and Gemini push search deeper into first-party data and real-time context

What's the News:

Google is expanding Personal Intelligence across AI Mode, Gemini, and Chrome—leveraging first-party data (like Gmail and Photos) to deliver highly personalized search experiences.

This includes recommendations based on past purchases, personalized itineraries, and suggestions inferred from user interests and behavior.

Why It Matters:

Search results are becoming increasingly individualized—and harder to replicate.

As Google leans further into personal context, two users searching the same query may see entirely different results. This raises the bar for relevance while making traditional rank tracking and performance measurement less reliable.

At the same time, it reinforces the importance of deeply understanding your audience—because Google is doing exactly that.

Silverback's POV

Marketers Should:

Invest in deeper audience research to understand behaviors, preferences, and decision drivers

Build content that reflects real user context—not just generic keyword targets

Diversify visibility strategies beyond rankings (e.g., presence across formats, platforms, and query types)

Modernize measurement frameworks to account for personalization and reduced visibility into SERPs

Identify gaps in visibility across audience segments, not just keywords

Insight #3

LinkedIn’s Algorithm Now Shows Content Based on What You Care About, Not Who You’re Connected To.

AI shifts the feed from “who you follow” to “what you engage with”

What's the News:

LinkedIn rebuilt big parts of its feed algorithm using LLMs and transformer models to retrieve and rank content more like “topics you care about”—not just “who you follow.”

Two core changes:

  • A single AI system now understands what posts are about and connects related topics, even when different language is used
  • A ranking model looks at how you engage over time (what you read, like, comment on) to adjust what shows up in your feed

LinkedIn is also reducing the reach of engagement pods, automation, and low-value “comment bait” posts.

Why It Matters:

Your network matters less than your content.

LinkedIn can now understand what your posts mean—and who should see them. That means strong, relevant content can reach beyond your followers, while generic or low-value posts are less likely to spread.

It also means engagement quality matters more. It’s not just about getting likes—it’s about whether people actually read, engage, and continue interacting.

Silverback's POV

Marketers Should:

Focus on a few core topics and consistently create content around them

Expand into related topics to increase reach through how LinkedIn connects ideas

Write posts people actually want to read—clear, structured, and useful

Aim for real engagement (comments, discussion, time spent), not just surface-level metrics

Avoid engagement pods, automation, and “comment bait” tactics—they’re actively being deprioritized

We’re reinventing performance marketing because everything changed while you were reading this.

We help you navigate this new landscape with data-driven insights, platform expertise, and creative that connects, so your brand not only keeps up, but gets ahead.

Work with Us