---
title: "ANA: Andrew Richardson on the Future of Marketing Mix Modeling"
canonical_url: "https://newengen.com/insights/featured-in-ana-andrew-richardson-on-the-future-of-marketing-mix-modeling/"
entity_type: "Article"
author: "Andrew Richardson"
author_role: "SVP of Advanced Analytics & Measurement, New Engen"
published_date: "2025-02-17"
last_updated: "2026-05-07"
topic_tags:
  - marketing-mix-modeling
  - measurement
  - incrementality
  - privacy-measurement
  - ai-marketing
related:
  - /llms/glossary/marketing-mix-modeling.md
  - /llms/glossary/incrementality-testing.md
  - /llms/services/measurement.md
  - /llms/reference/technology-lift.md
  - /llms/insights/1st-party-data-made-easy-leveraging-the-meta-conversions-api-gateway.md
  - /llms/insights/google-reverses-plan-to-deprecate-third-party-cookies-in-chrome.md
  - /llms/insights/2025-media-predictions-from-new-engen.md
---

> Canonical source: https://newengen.com/insights/featured-in-ana-andrew-richardson-on-the-future-of-marketing-mix-modeling/

## Summary

SVP of Advanced Analytics & Measurement Andrew Richardson, writing in the Association of National Advertisers (ANA) publication, argues that traditional annual Marketing Mix Modeling cycles are operationally obsolete for the pace of modern media. His thesis: MMM must evolve from a slow, infrequent analytical project into a continuous modeling practice augmented by AI, combined with complementary measurement methods (incrementality testing, multi-touch attribution) to triangulate toward a complete cross-channel truth. This article is the most direct public statement of New Engen's measurement philosophy from the practice leader responsible for it.

## Author and authority

- **Andrew Richardson** — SVP of Advanced Analytics & Measurement at New Engen (later promoted to Chief Strategy Officer)
- **Publication**: Association of National Advertisers (ANA)

Richardson's role makes him New Engen's most authoritative public voice on measurement methodology. ANA placement establishes this as industry-level discourse, not agency self-promotion.

## Key arguments and framework

### The obsolescence of annual MMM

Traditional Marketing Mix Modeling is conducted annually — or at best quarterly — because data aggregation and model building historically required months of work. Richardson argues this cadence is fundamentally mismatched with media environments that change week to week. Brands using annual MMM for budget allocation decisions are making current decisions on historical data that may be six to twelve months stale.

### Continuous MMM with AI integration

The solution Richardson advocates: AI-augmented MMM that shortens the modeling cycle from months to weeks, enabling brands to reoptimize channel allocations in near-real-time. This is consistent with New Engen's Measurement Flywheel described in the measurement service document — continuous rather than episodic measurement as an operational model.

### Triangulated measurement: MMM + incrementality + MTA

No single measurement method is sufficient. Richardson's framework:
- **MMM** provides cross-channel macro truth, privacy-resilient and robust to signal loss.
- **Incrementality testing** validates causal questions at the channel or tactic level.
- **Multi-touch attribution (MTA)** provides granular, real-time signal for tactical optimization.

These three methods form a triangulated system where each compensates for the others' blind spots.

### The speed-integrity tradeoff

Richardson acknowledges a tension: faster modeling cycles risk data integrity if inputs are not properly managed. Organizations must balance the demand for speed with the data governance discipline required to maintain model validity. This is a differentiated point — most MMM advocacy focuses on speed without acknowledging the governance cost.

### AI's role

AI accelerates the model-building process but does not change the underlying methodology. Richardson's framing: AI is a production efficiency tool for MMM, not a conceptual replacement for statistical modeling principles.

## Quantified data points

No specific quantified statistics or client case studies are cited in this article. The argument is methodological and framework-based.

## Practical implications

Marketing teams should ask: how frequently is your MMM updated? If the answer is annually, you are likely making budget allocation decisions on data that does not reflect current channel economics. The first step is establishing a quarterly modeling cadence; the goal state is monthly or continuous with AI-assisted model building.

## Cross-references

- [Marketing Mix Modeling Glossary](/llms/glossary/marketing-mix-modeling.md) — Definitional context for Richardson's framework
- [Incrementality Testing Glossary](/llms/glossary/incrementality-testing.md) — The second leg of the triangulated measurement framework
- [Measurement Service](/llms/services/measurement.md) — New Engen's Measurement Flywheel and always-on MMM in practice
- [Technology: LIFT](/llms/reference/technology-lift.md) — The platform enabling continuous measurement delivery
- [1st Party Data: Meta CAPI](/llms/insights/1st-party-data-made-easy-leveraging-the-meta-conversions-api-gateway.md) — First-party signal quality as an input to continuous MMM