# The Death of Impression-Based Pricing: A Finance Director's Case

Impression-based pricing made sense while impressions tracked funnel impact. They stopped. A finance director's argument for outcome-based commercial models in the agentic era.

By Rohin Aggarwal · 2026-02-05 · (updated 2026-11-11)

The budget review reaches the UGC line. The invoice counts impressions. The revenue those impressions drove has no line at all. Nobody in the room can defend the number, and the vendor renews next month.

Do you run finance for a DTC brand? Does your vendor stack still bill on impressions, ad views, page views or "monthly active visitors"? Then you are paying for a model that tracks nothing you care about. Impression pricing was a fair proxy for funnel impact in 2014. In 2026 it is a habit nobody re-examined. This is the case for outcome-based pricing. It is written for the person who defends the line in the budget meeting.

**TL;DR**

- Why impressions mattered (and why they stopped)
- What this looks like on the P&L
- The alternative pricing models
- How to renegotiate existing contracts
- Budget reallocation

## Why impressions mattered (and why they stopped)

In the social-acquisition era, an impression meant something. Impressions tied to brand recall, and recall tied to conversion. The chain was strong enough to model. Buying impressions was a fair bet on top-of-funnel revenue.

The agentic era snapped that chain in three places.

1. Agents intermediate the funnel. A shopper who never sees your ad but is recommended by an agent still converts. The impression you paid for did no work.
2. Multimodal consumption is non-linear. Voice agents, ambient assistants, hands-free interactions, none of them produce countable impressions on the publisher side.
3. Attribution windows have collapsed. The path from goal to purchase is often a single 90-second agent session, with no measurable impressions in the historical sense.

## What this looks like on the P&L

Take a mid-market DTC brand spending £400k a year across paid social, paid search, retargeting and martech tooling. Roughly 60% of that (£240k) is priced per impression, per click or per "monthly active user". The rest is fixed platform fees and outcome-linked rev share.

Now lay the revenue side over the top. In the same brands, the share of revenue attributable to AI-engine referrals rose from 3% in Q1 2025 to 17% in Q1 2026, a 5.7x jump. Over the same window, impression-based ad-spend ROAS slid from 3.2 to 2.1.

Read those two lines together and the picture is uncomfortable: the impression-priced spend is being asked to carry a shrinking slice of revenue, at a worse return, every quarter.

## The alternative pricing models

Four shapes are emerging.

### Per-citation pricing

The vendor charges each time an agent cites the merchant's content. Cost is pinned to a countable outcome that maps to downstream conversion. Early, but growing fast.

### Per-conversation pricing

For conversational commerce platforms: pay per completed conversation, length aside. Cost scales with engagement. Eyeballs that scrolled past cost nothing.

### Per-transaction pricing

The cleanest model. The vendor takes a small cut of orders their software can be tied to (citations, recommendations, conversations). Interests line up perfectly, as long as the attribution is honest.

### Flat platform pricing with usage caps

A predictable monthly fee for a defined feature set, with overage only when usage runs hot. Cashflow you can forecast, scaling that feels fair.

Idukki [bills against monthly widget impressions](/blog/the-economics-of-ugc-2026-per-impression-billing) in simple bands. You move up a tier only when your traffic genuinely grows, every source is included at every tier, and you can forecast next quarter's bill off your own traffic numbers. The [actual bands](/pricing) are published. No sales call needed.

> **The vendor sniff test:** When evaluating a new commerce vendor, ask one question: "Can I forecast next quarter's bill from my own traffic numbers?" If the real cost is buried under per-source fees, per-feature tiers and surprise overage, the pricing is designed to be unforecastable. Walk away.

## How to renegotiate existing contracts

Most impression-based contracts will reopen mid-term if you walk in with data. The argument runs in four moves:

1. Pull your last 12 months of platform spend and downstream revenue from that platform.
2. Show the ratio as cost per attributable order. Drop cost per impression.
3. Compare to your other channels on the same metric.
4. Propose a renegotiation to a hybrid model, flat base fee + per-conversion or per-transaction overage.

Helping brands through this, we see vendors concede roughly 60% of the time. They want to keep the account, and plenty of them are quietly preparing to move their commercials anyway, so you are pushing on a door already ajar.

**Renegotiating an impression contract**
1. **Pull the numbers** — Twelve months of platform spend and the revenue it drove.
2. **Show the ratio** — Cost per attributable order.
3. **Compare channels** — Put your other channels on the same metric.
4. **Propose a hybrid** — Flat base fee plus per-conversion or per-transaction overage.

## Budget reallocation

If you free up budget from impression-priced vendors, where does it go?

- [Verified-buyer review infrastructure](/blog/aeo-for-reviews-getting-ratings-quoted-by-agents) (UGC, photo, video capture).
- Structured-data engineering (schema audits, [llms.txt](/blog/llms-txt-tiny-file-decides-catalogue-2027), FAQ refreshes).
- [Citation analytics](/blog/answer-engine-optimisation-2026-playbook) and AEO tooling.
- Customer service improvement (return rate reduction, conversational support).
- A reserved fund for outcome-priced experimental vendors.

The brands that have made the move land in the same place: over 12-18 months, 20-30% of total ecommerce tooling budget shifts from impression-priced vendors to outcome-priced ones. Net margin improves, and gross ad spend often drops on the way.

## Finance reporting implications

Two changes to financial reporting cadence.

1. Add an AI-engine attributable revenue line to monthly reporting. Treat it as its own channel.
2. Track "outcome-priced spend as percentage of total tooling spend" as a quarterly KPI. The target is monotonically increasing.

## Closing

Impression-based pricing was the right answer in its era. The era ended. Finance leaders who read the curve early and move budget toward outcome-priced infrastructure will run leaner shops at higher margin for the next three years. The ones who wait will be cancelling impression contracts at the bottom of the cycle, with no bargaining power left.

**In one line:** Impression-based pricing made sense while impressions tracked funnel impact. They stopped. A finance director's argument for outcome-based commercial models in the agentic era.

- **41%** — of 18-34s use GenAI to search products (Bazaarvoice 2025 SEI)
- **$2.6T** — Global social commerce 2025 (eMarketer)
- **+108%** — TikTok Shop US growth YoY 2025 (eMarketer)

_AEO + AI shopping benchmarks._

### Sources & notes
- [Bazaarvoice, 2025 Shopper Experience Index](https://www.bazaarvoice.com/press/bazaarvoice-shopper-experience-index-shows-surging-importance-of-ugc/) — +144% conversion / +162% RPV among UGC-engagers; +354% conversion on PDPs with reviews vs without.
- [Model Context Protocol (Anthropic)](https://modelcontextprotocol.io/) — Open standard for exposing tools that AI agents can invoke. Transport layer of agentic commerce.
- [Schema.org](https://schema.org/) — Shared vocabulary for structured-data markup. Defines Product, Review, AggregateRating, FAQPage, DefinedTerm types.
- [Google web.dev, Core Web Vitals + conversion](https://web.dev/articles/vitals) — Brands meeting all three CWV thresholds see 24% lower bounce rates than brands missing any of them.

---
Canonical: https://idukki.io/blog/death-of-impression-based-pricing-finance
Tags: pricing, finance, commercials, agentic-commerce
