Reiser's Pieces: When Influence Forms Upstream - The New Fight for Consideration

Last issue, we talked about agentic commerce and why it’s shrinking the shelf. When AI becomes the front door to shopping, it doesn’t just recommend. It narrows brand options, builds carts, and can complete checkout. Fewer options means the basics matter more than ever, because “not recommended” becomes a sales problem.

That was the downstream change. 

Today, we’re focusing upstream. Before an agent can recommend anything, influence has to occur. And the consumer influence system that brands have depended on for years has been rewired. There’s a new consumer influence system. Individuals create the proof, platforms distribute it, and LLMs like ChatGPT, Gemini, and Google’s AI Overview turn it into a shortlist of what gets trusted, considered, and bought. 

Let’s start with the proof layer: consumer and creator content. Reviews. TikTok demos. YouTube explainers. Reddit threads. Group chat screenshots. All of it is raw material shoppers use to decide. 

Then there’s distribution. Commerce moved into feeds and creator ecosystems. Influence became infrastructure.

The global creator economy is projected to grow from $191B in 2025 to $528.39B by 2030.

TikTok Shop had about 2.3M active affiliates in the U.S. as of February 2026.

On Black Friday, 10 million affiliate videos were posted.

That’s a machine at scale. There are 246 million U.S. social media users, more than half of whom have made influencer-inspired purchases. 

But now, add the filter layer. This is where the game changes. 

LLMs don’t just surface information. They compress it. We’re moving from keywords to conversation, from lists of links to a shortlist and a recommendation. “Best gluten-free chicken recipe for guests” becomes “my in-laws are here, one is gluten-free, and I need to impress them.” The system reads across sources and gives a recommendation. Sometimes it drives a click; sometimes not. Either way, the influence still happens. 

That’s why measurement starts to break down. Not so long ago, influence drove a click, and the click was your proof. In this world, intent can be satisfied before the click, which means decisions can get shaped without a visit to your page. And if you’re not present in the inputs the LLM pulls from, you don’t just lose share of voice. You lose share of consideration. 

So how do you activate this system without chasing every shiny object? 

First, acknowledge that consumer engagement is becoming human as LLMs become a dominant form of influence. Then determine how your brand can create authentic conversations organically. Not slogans. Truth. Use cases. Tradeoffs. Who it’s for, and who it’s not for. If you don’t define that in the places where people discuss the category, the category will define it for you. 

Second, earn presence where the category gets decided. Reviews need a strategy. Creator partnerships need to be utility-driven, not just reach-driven. Community participation needs a point of view and expertise, not brand speak. 

Third, modernize measurement. If you can’t see how often you show up in LLM recommendation sets for the category questions you should own, you’re flying blind. The metric isn’t just clicks. It’s share of shortlist. 

We’re already working with brands on exactly this: Mapping the category questions that matter, determining how to improve share of shortlist, and building a repeatable cadence to keep it moving. 

I leave you with this. When a shopper asks the question you should own, are you influencing what the market (and the model) says back, or are you hoping your ad shows up after the decision is already made? 

Time to build the influence engine that earns visibility before the click, in the shortlist, and at the point of purchase. 

---

Jason Reiser is Chief Executive Officer, Market Performance Group