The measurement-to-action gap
The problem is not a lack of marketing data. It is the difficulty of turning fragmented data and measurement methods into a decision that stakeholders trust.
Platform reporting, multi-touch attribution, and marketing mix modeling each answer different questions at different levels of granularity. Used independently, they can create competing narratives and push teams into reconciliation rather than optimization. Marketers still need a bridge from measurement to planning — and from planning to an action they can actually take.
Describes activity, not incrementality
Reporting can describe activity within an individual ecosystem, but does not by itself establish cross-channel incrementality.
Journey-level granularity
Multi-touch attribution brings journey-level detail when event and identity signals are available.
Aggregate, cross-channel view
Marketing mix modeling provides an aggregate view that can incorporate offline media and non-marketing factors.
From measurement to action
Teams still need a path from measurement to planning, and then from planning to an understandable, defensible action.
What is Marketing Campaign Analytics?
MCA is Adobe's next-generation, marketer-facing campaign intelligence direction — built to expand the value of causal measurement and planning for marketers, channel owners, analysts, and executives.
It builds on the causal measurement and planning foundation of Adobe Mix Modeler — MMM, MTA, data harmonization, incrementality, model insights, and scenario planning — and broadens the analysis from channels toward campaigns, creatives, audiences, and funnel stages. The simplest way to think about it:
Measure what is incremental
Separate baseline from marketing-driven lift, defensibly.
Understand what is changing
Explain why performance moved, not just that it did.
Explore what could happen
Compare budget and goal scenarios before you commit.
Make it easier to act
Bring trusted signals into a marketer-first decision surface.
How MCA works: from data to decision
Six stages take marketing and business data from ingestion through to an action a marketer can take.
Connect and ingest
Marketing and business data enters the Adobe data foundation — aggregate series for MMM, event-level interaction data for MTA, and internal or external factors that may influence outcomes.
Standardize and harmonize
Datasets are mapped into consistent schemas and harmonized fields so channels, metrics, dates, and conversions can be analyzed together. Identity configuration matters for event-level journey analysis.
Measure incrementality
MMM analyzes aggregate time-series signals while MTA analyzes identity-linked event paths — unified through bi-directional transfer learning so granular and aggregate measurement inform one another.
Diagnose and explain
Model outputs expose contribution, incrementality, and channel behavior. The MCA direction adds a marketer-facing, conversational layer that surfaces insights and supports root-cause exploration.
Plan scenarios
Measurement outputs feed planning workflows, where marketers compare alternative budget or goal scenarios and evaluate projected outcomes before changing investment.
Bring insight into the campaign workflow
Those signals become available in a simpler, campaign-oriented experience alongside broader CX analytics — so teams can move from insight toward action faster.
The questions a marketing leader actually asks
Each maps to a capability in the MCA / AMM foundation — and to a specific business outcome.
| Business question | MCA / AMM foundation | Potential value |
|---|---|---|
| What actually drove the outcome? | Incrementality-focused MMM and MTA | A more defensible view of marketing contribution than isolated platform credit. |
| Why did performance change? | Factors, granular breakdowns, channel analysis, and AI-assisted investigation | A faster path from a metric movement to a hypothesis you can investigate. |
| Where should we invest next? | Response behavior, forecasts, and scenario planning | A structured way to compare alternatives before reallocating budget. |
| How do we reduce measurement friction? | AEP-based ingestion, standardized schemas, harmonization, and a marketer-first experience | Less dependence on disconnected reports and repeated reconciliation. |
| How do analysts and marketers share one foundation? | Shared data foundation plus model outputs that extend into CJA | Closer alignment between causal measurement and deeper journey analysis. |
The measurement foundation, under the hood
The AMM measurement foundation runs on Adobe Experience Platform. Select a topic to explore the methodology behind the experience.
Data foundation and schemas
Data used by Mix Modeler is first ingested into Adobe Experience Platform and mapped to Experience Data Model (XDM) structures. Event-level data commonly uses XDM ExperienceEvent, while aggregate inputs use summary/aggregate metric structures. Ingestion methods include source connectors, batch files (CSV, JSON, Parquet), streaming ingestion, and APIs.
For measurement, data generally falls into three functional categories:
- Marketing effort and exposure — the media activity being measured.
- Business outcomes — revenue, conversions, and other results.
- Internal / external factors — promotions, seasonality, pricing, or macroeconomic signals.
Event-level data supports MTA when usable identity is available; time-series summary data supports MMM.
Harmonization and identity
Harmonization is the bridge between heterogeneous source data and the modeling layer: mapping source-specific fields to common harmonized fields, aligning time granularity, and combining event and aggregate datasets into a consolidated analytical view. For event-level analysis, AEP identity configuration can connect interactions that use different identifiers when appropriate identity fields and stitching rules are configured.
This layer is central to measurement quality. If channels use inconsistent taxonomies, dates are misaligned, records overlap, or identities cannot be connected, the model can only reflect those limitations. Good governance begins with controlled schemas, documented mappings, consistent granularity, and ongoing data-quality monitoring.
MMM — aggregate causal measurement
Adobe's methodology describes the MMM component as a multiplicative nonlinear regression. In simplified form:
Inputs include aggregate channel spend or effort over time plus external factors. The model is designed to capture nonlinear interactions, saturation and diminishing returns, and the decomposition of baseline versus incremental contribution.
Outputs include incremental contribution by channel, and can support ROI, forecasting, and scenario analysis.
MTA — event-level incremental contribution
MTA is described as a discrete-time survival — or hazard-based — model of time-to-event with time-varying channel exposures. It incorporates:
- Lag / adstock effects across time.
- Sequence-aware touchpoints along the path.
- Flexible temporal windows.
- Both converting and non-converting paths.
A Shapley-value attribution layer estimates marginal contribution, producing incremental lift at the touchpoint and channel levels.
Unifying MMM and MTA
A core AMM design principle is that MMM and MTA are not two unrelated reporting systems. Bi-directional transfer learning lets granular MTA structure inform MMM, while MMM calibrates baseline versus incremental contribution in MTA.
The intended result: alignment between the sum of MTA incremental contributions and the aggregate incremental lift estimated by MMM. Granular journey evidence and aggregate market evidence constrain one another, instead of forcing marketers to choose between two independent answers.
Experiments as priors
External experiments — such as geo or lift tests — can act as priors that serve as calibration constraints, anchoring models to observed causal effects.
This provides a path to incorporate experimental evidence rather than relying exclusively on observational data, strengthening the causal claims the measurement layer can make.
Model interpretation and planning
Once models are trained, the foundation produces interpretable outputs for diagnosis and planning. Model configuration spans conversion goals, marketing channels, external factors, channel adstock, granular reporting fields, model insights, channel synergies, drift detection, and goal-based planning.
The planning workflow compares scenarios against business objectives and constraints — the point at which measurement becomes a decision.
A role-aware surface on a governed foundation
The architecture principle is separation of concerns: the measurement foundation owns harmonization, modeling, scoring, and planning; MCA presents and composes those outputs for marketers.
Governed schemas and datasets, identity capabilities, and the data foundation used by measurement workflows.
Harmonization, MMM, MTA, transfer learning, model diagnostics and insights, incrementality, and scenario planning.
Campaign-oriented performance views, AI-assisted exploration, cross-domain insight composition, and a simplified decision surface as the product evolves.
Deeper behavioral and journey analysis, with model-derived outputs able to flow into CJA Workspace for further exploration.
Keeping analytical truth in the measurement foundation — rather than duplicating it in a second application layer — makes lineage easier to reason about as the experience evolves.
The MCA experience
The reviewed customer demo shows MCA as a role-aware campaign surface, bringing together campaign KPIs and model-derived incremental measures, proactive insights, broader CX analytics signals, and campaign drill-downs — with a path into CJA for deeper analysis.
AI-assisted insight
A major part of the direction is reducing the analyst bottleneck. The Data Insights Agent is described as enabling business and marketing professionals to generate visualizations and insights from natural-language prompts, with root-cause and proactive insights on the stated 2026 roadmap — illustrating a progression from insight, to explanation, to action.
The Data Insights Agent has its own product lifecycle; the MCA experience that embeds or orchestrates agentic workflows has been demonstrated as vision / pre-GA material. Customer commitments should follow current product documentation and contractual availability.
What's established today vs. what's evolving
MCA is an evolution that builds on AMM's measurement foundation — not a label change implying every AMM workflow or backend has already moved.
| Area | Status | Customer-safe positioning |
|---|---|---|
| AMM measurement foundation | Established foundation | MMM, MTA, harmonization, causal measurement, model insights, and planning are the analytical heritage MCA builds on. |
| MCA marketer-facing UI | Evolving / pre-GA in reviewed demo | Adobe is building a simpler marketer-facing experience and phasing AMM functionality into the MCA direction. |
| Agentic and proactive workflows | Evolving | Treat as product direction unless the specific capability is confirmed as available in current documentation. |
| MCA Essentials timing | Internal update, Aug. 4, 2026 | GA shifted beyond August to Fall 2026; validate release status before making customer commitments. |
| Customer deployment specifics | Implementation-dependent | Data sources, identity, taxonomy, KPI definitions, model scope, and governance must be validated per organization. |
A practical path — start with the question, not the interface
A recommended customer-planning approach, based on the documented data and modeling dependencies.
Define outcomes
Agree on the business KPI or conversion that measurement must explain, and the decisions the organization wants to improve.
Inventory signals
Identify aggregate media, outcome, event-level journey, and relevant internal / external factor data.
Standardize the measurement layer
Align channel taxonomy, time grain, identity assumptions, currency / time-zone conventions, source mappings, and ownership.
Establish trustworthy models
Configure MMM / MTA scope, validate diagnostics and interpretability, and incorporate experiments where appropriate.
Operationalize planning
Use response behavior and scenarios to frame investment trade-offs against business constraints.
Expand into marketer workflows
Expose trusted model outputs and complementary journey signals through the MCA experience as capabilities become available.
Governance essentials
Enterprise measurement can combine sensitive first-party behavior, cost, conversion, and business-performance data. Apply AEP schema and data-governance controls before signals are used for measurement or surfaced to marketers.
- Define schemas and identity fields deliberately, especially for event-level attribution.
- Maintain consistent time granularity; prevent overlapping or duplicate aggregate records.
- Document source-to-harmonized field mappings and analytical lineage.
- Monitor ingestion and harmonization quality over time, especially when source platforms restate data.
- Apply privacy, consent, access, and governance policies to the underlying AEP datasets.
Frequently asked questions
Is MCA just a new name for Mix Modeler?
No. MCA builds on Mix Modeler's measurement and planning foundation but expands toward a broader marketer-facing campaign intelligence experience. The transition is phased, and reviewed demo material explicitly shows periods in which MCA and existing AMM experiences remain separate.
Does MCA replace CJA?
No. The reviewed vision positions them as complementary. MCA focuses on campaign performance, ROI, measurement, and decision support; CJA provides deeper behavioral and journey analysis. Demo material shows model-derived outputs being taken into CJA for further exploration.
Why combine MMM and MTA?
They observe marketing at different levels. MMM uses aggregate time-series data and can incorporate broad channel and external effects; MTA uses event-level journeys when identity signals are available. Adobe's methodology unifies them through transfer learning to align granular and aggregate incremental measurement.
Can experimentation be incorporated?
Yes. Adobe methodology material describes external experiments — such as geo or lift tests — as priors used to calibrate models toward observed causal effects.
Is MCA generally available?
The latest internal timing source reviewed for this paper is dated August 4, 2026 and states that MCA Essentials GA shifted beyond August to Fall 2026. Because release timing can change, customers should validate current availability with their Adobe account team and current product documentation.
Build the durable measurement layer first.
Establish clean data, transparent measurement assumptions, and governance — then expand those trusted outputs into the marketer-facing MCA workflows as they become available. The goal is a foundation that answers not only "what happened?" but "what was incremental?", "why did it change?", and "what should we evaluate next?"
Because release timing and capabilities can change, validate current availability and contractual terms with your Adobe account team and current product documentation before making commitments.