Marketing Mix Modeling for AEO: How to Isolate AI Search Contribution

Every quarter, somewhere inside a B2B company doing meaningful work on answer engine optimization, an AEO lead walks into a quarterly business review with a deck full of screenshots showing ChatGPT and Perplexity citing their brand by name. Forty-five minutes later they walk out with their budget cut. The screenshots were beautiful. The narrative was confident. The CFO killed it anyway.

According to Gartner's 2026 CMO Spend Survey, 47% of B2B marketing leaders reported AEO budget cuts or freezes in Q1 2026 — not because AEO doesn't work, but because the QBR format used to defend it failed under finance scrutiny. The pattern is consistent enough across the practitioners I've talked to in the last two quarters that it is no longer a series of one-off mistakes. It is a structural template problem.

The QBR template most AEO teams inherit is a marketing-mix-modeling deck with the word "AI" pasted into the section headers. It opens with a chart of total citations climbing month-over-month, walks through three case studies of brand mentions in ChatGPT answers, and lands on a flat ask for more headcount and tooling. This format reliably fails because it does not address the three audiences in the room — the CFO who needs to defend the spend to the board, the CEO who needs the strategic narrative to fit the company's positioning, and the CMO who needs the tactical detail to run the team — with the language and evidence each audience requires.

The QBR template that survives in 2026 is structured the way Bain and McKinsey structure client steering committees, the way Gainsight structures customer success plans, and the way HBR's most-cited piece on executive reporting describes the assertion-first pyramid. Twelve slides. Three audiences. One falsifiable ask. This is the walkthrough.

Why The Default AEO QBR Format Fails

The default QBR format AEO teams use was inherited from the SEO playbook, where the audience expected visibility numbers presented as a trend chart and revenue impact presented as a last-touch conversion model. That format works in SEO because the CFO already accepts the underlying attribution model — Google Search Console traffic, GA4 conversion paths, marketing-qualified lead counts — even when finance disputes the magnitude.

In AEO the underlying attribution model is not yet a settled standard. There is no Search Console equivalent. There is no GA4 default report. The CFO walks into the QBR carrying the implicit question — "how do I know any of this is real?" — and the default deck answers with "look at all the citations" rather than "here is the falsifiable revenue link, here is the control comparison, here is the cost per outcome."

Bain's published QBR conventions, codified in their Insights series on results-driven dialogues, insist on what they call "the answer-first slide." Every slide title should be a complete sentence stating the conclusion the data supports. Charts go below the title as evidence, not above it as discovery. The standard AEO QBR violates this convention on almost every slide. A title like "Citation Performance Q1" is not an answer. A title like "AEO-sourced pipeline grew 34% while CAC declined 18%, beating plan by $1.2M" is an answer. The first version invites CFO skepticism by design; the second version puts the burden of refutation on the skeptic, which is the dynamic any defensible QBR needs to establish in the first five seconds.

Gainsight's customer-success QBR playbook — described in their long-running Pulse conference materials and the open-source success-plan templates they publish — uses a parallel structure: every slide ties to a stated business outcome with a numerical health score the customer's executive sponsor has agreed represents success. AEO QBRs that borrow this convention pre-negotiate the success metric with the CFO at the start of the quarter, rather than presenting a metric the CFO has never seen and asking for it to be ratified at the meeting. Pre-negotiation is the single largest determinant of whether a QBR survives.

The McKinsey Quarterly's research on the decline of executive attention found that boards and C-suite reviews are now allocating an average of 4.2 minutes per slide, down from 6.7 minutes five years ago. A deck longer than fifteen slides effectively guarantees that the slides toward the end will receive less than two minutes of attention each. AEO QBRs that bury the ask on slide twenty-two die not from rejection but from inattention.

The Three-Audience Structure: CFO, CEO, CMO

A single deck cannot serve the CFO, CEO, and CMO with the same content because their decision logic is different. The CFO is being asked to allocate cash against a hurdle rate. The CEO is being asked to commit to a strategic posture in front of the board. The CMO is being asked to deploy team capacity against a competitive map. The QBR template that works uses a three-audience structure within twelve slides — not three separate decks, but three threaded narratives layered into one.

The CFO narrative: contribution margin, payback, control comparisons

The CFO's section of the deck — slides three through five in the template below — addresses three questions in order. What did the spend produce in pipeline this quarter? What is the contribution margin on that pipeline relative to other channels? When does the cumulative investment cross the payback line?

The contribution margin slide is the single most important slide in the entire deck. It is also the one most AEO teams skip because the data is hard. The discipline is to take the closed-won opportunities the AEO program is credited with sourcing or influencing, apply a fully-loaded gross margin (not a contribution-to-overhead figure), and compare it against the same calculation for paid search, paid social, content syndication, and outbound SDR. If AEO's contribution margin is competitive, you have a defensible budget. If it is not competitive, no narrative slide elsewhere in the deck saves you. Our AEO ROI payback piece walks through the exact mechanics of the calculation.

The control comparison is the second-most important. CFOs trained in finance treat any reported number without a counterfactual as advocacy rather than evidence. The control comparison can take three forms: a geographic holdout where one region intentionally received no AEO investment, a temporal control comparing the period before AEO investment to the period after, or a synthetic-control construction using Causal Impact-style methodology. Each has trade-offs. The geographic holdout is the strongest but requires sacrificing pipeline in the holdout region; the temporal control is the cheapest but the most contaminated by other concurrent changes; the synthetic control is methodologically defensible but requires a CFO who trusts the statistical method, which most do not without an explainer.

The CEO narrative: strategic posture and competitive defensibility

The CEO's section — slides six through eight — addresses three different questions. What position does our brand occupy in the AI-mediated buyer research surface, and is the position consistent with our broader market positioning? Where are competitors gaining share-of-voice and what is our response? What strategic optionality do we lose if we under-invest for one more quarter?

The strategic posture slide is the trap that catches most AEO leads. The temptation is to use it to celebrate brand mentions; the correct use is to compare the brand's positioning inside LLM answers against the positioning the CEO has communicated to the board. If the CEO has told the board the company is the platform of choice for mid-market manufacturers and ChatGPT consistently positions the company as a tools vendor for hobbyists, that is the strategic posture problem the QBR has to surface — not an AEO failure but a brand-coherence failure that AEO measurement uniquely makes visible.

The defensibility slide draws on the same logic. Competitors gaining share-of-voice on a specific buyer cohort represents a defensibility threat that may compound quarter-over-quarter as the LLM training data the competitor seeds reinforces itself. The CEO needs to see this as a window-of-action argument: under-invest now and the cost-to-recover doubles by year-end, the way McKinsey describes compounding visibility advantages in their digital strategy work.

The CMO narrative: tactical execution and team capacity

The CMO's section — slides nine through eleven — is the deepest in detail but the shortest in stage time. The CMO is in the room as the QBR sponsor and already knows most of the content; what they need is for the deck to surface the operational risks they will be asked about by their CEO peer in the meeting following the QBR.

The three CMO slides cover content velocity (how many of the planned briefs, pages, and assets actually shipped, with reasons for variance), citation coverage by buyer query cohort (which queries are won, which are lost, and which are contested), and incident log (hallucinations, misattributions, model-deprecation surprises that disrupted reporting). The incident log is essential and almost universally omitted. Without it, the CFO eventually discovers an incident through their own channels and uses the omission as evidence that AEO reporting is not credible. With it, the CFO sees the program is operationally mature enough to surface its own failures.

The 12-Slide AEO QBR Walkthrough

The template below is what I use with B2B SaaS clients running AEO programs between $500K and $5M in annual investment. The slide numbers and titles are exact. The expected runtime is forty-five minutes plus fifteen for questions.

1. Executive summary A single answer-first slide stating, in one sentence, the quarter's headline result and the recommendation for next quarter. Example: "AEO sourced $4.2M in pipeline at 19% blended contribution margin, payback achieved in month nine of an 18-month plan; we are asking for the planned Q3 increase to $1.1M to maintain compounding share-of-voice against Competitor X."

2. Methodology disclosure One slide stating the citation sampling method, the LLM coverage (which models, which versions, which date range), the query taxonomy, and the attribution model used to credit pipeline. This slide is the prerequisite to every subsequent finance slide. Without it, the CFO's first question on slide three derails the meeting.

3. Contribution margin by channel A side-by-side bar chart of fully-loaded contribution margin by acquisition channel — AEO, paid search, paid social, content syndication, outbound SDR — with the closed-won pipeline counts annotated. The title states whether AEO's margin is competitive with the next-best channel and by how much.

4. Payback period against quarterly investment A cumulative cash-flow curve showing investment-to-date and AEO-attributed gross profit, with the intercept marked as the payback month. The title states the payback figure relative to the plan committed at the prior QBR.

5. Control comparison Either a geographic holdout result or a temporal control. The slide reports the lift, the confidence interval, and the methodology limitations. The title states the magnitude of the lift in absolute revenue terms.

6. Share-of-voice trajectory A line chart of the brand's share-of-voice across a defined query cohort, indexed against the three top competitors. The title states whether share is expanding, contracting, or stable and the competitor most responsible for any change.

7. Strategic posture audit A two-column visualization: the brand positioning the CEO has communicated to the board, alongside the positioning that ChatGPT, Claude, and Perplexity actually produce when asked about the brand. The title states whether the positions are aligned or diverging.

8. Competitive defensibility map A 2x2 of buyer query cohorts by commercial intent and current win rate, with arrows showing competitor movement. The title states the cohort most at risk of being lost and the investment required to defend it.

9. Content and execution velocity Planned versus shipped count of briefs, pages, and structured-data deployments, with variance explanations. The title states the variance and the most material driver.

10. Citation coverage by query cohort A heatmap of buyer queries by win/contested/lost state, with cohort-level revenue annotated. The title states the cohort with the highest revenue-weighted gap.

11. Incident log and methodology updates A short list of citation incidents (hallucinations, misattributions), model-coverage updates, and methodology changes that affected reporting. The title states whether any incident materially altered the headline numbers.

12. The ask and the stop-doing list The forward commitment: the requested investment, the falsifiable metric that defines success, and the specific tactics or programs the team will discontinue to free capacity. The title states the ask in dollars and the success metric in plain language.

The discipline is to keep the deck to twelve slides exactly. If a topic does not fit, it goes into the appendix and is referenced only in response to a direct question. This convention is borrowed from the Bain steering-committee format and the Amazon six-pager tradition both — abundance of content in the appendix, scarcity of content in the main deck.

Slide-By-Slide Audience Mapping

The mapping is what allows a single twelve-slide deck to thread three narratives without becoming three decks. The CFO's slides ground the deck in finance discipline; the CEO's slides elevate it to strategic posture; the CMO's slides anchor it in operational reality. Each slide has a primary audience whose decision logic the slide is engineered to serve, and a secondary audience whose objection the slide is engineered to pre-empt.

Metrics That Survive Scrutiny Versus Vanity Metrics

The metric layer is where most AEO QBRs collapse. The default impulse is to report what is easy to measure — total citations, brand mentions, share-of-voice across all queries — rather than what is hard to defend. The CFO's instinct is the reverse. Anything reported that is easy to measure but loosely tied to revenue is treated as advocacy; anything reported that is hard to measure but tightly tied to revenue is treated as evidence.

The discipline I run with clients is to maintain two metric layers explicitly. The first layer is the headline scorecard that appears in the QBR — five to seven metrics, each tied to a falsifiable revenue link, each accompanied by methodology disclosure. The second layer is the operational dashboard the team uses internally — thirty to forty metrics covering content production, citation coverage, model coverage, technical health, and competitive surveillance. The internal dashboard never appears in the QBR. Confusing the two is the most common reason QBRs fail.

The headline scorecard typically includes AEO-attributed pipeline, contribution margin, payback period, share-of-voice on a defined commercial-intent query cohort, and cost per qualified citation. The vanity metrics that get cut are total citation count without commercial-intent weighting, brand mention sentiment scores divorced from transaction data, follower or impression-style metrics in social-graph terms, and any composite "AEO score" whose components are not separately auditable. The detailed framework for which seven metrics belong in the executive scorecard is in our CMO AEO dashboard piece.

The HBR essay The Real Value of Middle Managers and HBR's adjacent work on executive reporting both make the same point in different framings: the executive who reads the report is not the one who produced it, and the report's job is to make their decision possible in the time they have. The metric scorecard is the single most important place to apply that principle. If the headline scorecard does not survive a hostile read in four minutes, the QBR does not survive.

The Bain pyramid principle applied to AEO metrics

Barbara Minto's pyramid principle — popularized inside McKinsey and Bain as the structural discipline behind every consulting deliverable — says that conclusions come first, supporting evidence comes second, and detail comes third. Applied to AEO QBR metrics, this means the headline slide does not say "we tracked 47 metrics across our AEO program." It says "AEO is contributing $4.2M in pipeline at 19% margin and is on a path to compound to $16M in pipeline at 23% margin by Q4." The 47 metrics are in the appendix.

The pyramid principle is what allows the deck to survive what consultants call "the executive zigzag" — the pattern where a senior executive jumps from slide three to slide twelve to slide six in the order their attention prompts. A deck built bottom-up — with the conclusion only visible on the final slide — collapses under the zigzag. A deck built pyramid-style allows the executive to land on any slide and immediately understand its conclusion from the title.

How Gainsight's QBR Conventions Map to AEO

Gainsight's customer-success QBR conventions — refined over a decade of running Pulse and codified in their success-plan templates — were originally designed for customer-success managers presenting account health to the customer's executive sponsor. The structural similarity to the AEO QBR is closer than it first appears.

In the Gainsight model, every QBR opens with the executive sponsor's stated business outcomes and walks each metric back to one of those outcomes. The CSM is not allowed to introduce a metric that does not map to an outcome the sponsor has previously ratified. This convention is what prevents the QBR from devolving into a list of activities the CSM has performed; every activity is a vehicle for an outcome the customer has already agreed to.

Applied to AEO, the convention requires the AEO lead to negotiate the success outcomes with the CFO and CEO at the start of each quarter — not at the QBR. The outcomes might be "increase share-of-voice on the mid-market manufacturer buyer cohort by ten points," or "achieve eleven-month payback against the trailing twelve months of investment," or "neutralize Competitor X's citation gain in the procurement-decision query cluster." Each is falsifiable, time-bounded, and ratified in advance. At the QBR, every metric on every slide maps back to one of these pre-ratified outcomes.

This is the single largest behavioral change I ask AEO leads to make. The instinct is to negotiate the success criteria at the QBR by presenting evidence and asking the CFO to ratify the framing. The discipline is to negotiate the success criteria in private a quarter before, present evidence against the agreed criteria at the QBR, and use the meeting to discuss what changes for next quarter rather than to relitigate what success means.

The Cadence Between QBRs

The QBR alone is not sufficient. LLM answer surfaces move too quickly for a quarterly cadence to maintain credibility. The companies that defend AEO budgets successfully run a three-layered reporting cadence: the QBR every twelve weeks, a monthly written operating review, and a weekly Slack-format pulse to the CMO and the marketing leadership team.

The monthly review is two pages. Page one is the same scorecard that anchors the QBR with month-over-month deltas highlighted. Page two is a written narrative — 400 to 600 words — covering material changes. Material changes include model deprecations (GPT-4 to GPT-5 transitions, Claude version bumps, Gemini reranking adjustments), prompt-pattern shifts observed in buyer research behavior, citation displacement events, and methodology updates. The narrative is the operational document; the scorecard is the financial document.

The weekly pulse is five bullets: top-three citation wins, top-three citation losses, one methodology note, one upcoming risk, one ask. The pulse goes into the marketing leadership Slack channel every Monday morning at the same time. The discipline of weekly pulses creates two effects. First, the CMO is never surprised at the QBR by a development they should have known about three weeks earlier. Second, the AEO lead develops the editorial discipline of identifying what actually matters each week, which sharpens the quarterly narrative when the QBR arrives. The full citation-tracking infrastructure that feeds this cadence is described in our citation tracking playbook.

Common QBR Failure Modes and Recoveries

The five most common failure modes I see in AEO QBRs map to specific recoveries.

The methodology ambush. The CFO challenges the citation sampling method on slide three and the meeting never returns to the recommendation slide. Recovery: move methodology disclosure to slide two — before any financial number — and pre-circulate the methodology document a week before the QBR so the CFO's questions are surfaced in private.

The vanity-metric trap. The deck opens with total citation count, the CFO asks "and how does that translate to revenue?" and the AEO lead does not have a clean answer. Recovery: remove total citation count from the QBR entirely; it belongs only in the operational dashboard.

The over-attribution problem. The deck credits AEO with $4.2M in pipeline, the CFO asks how much would have closed anyway, and the AEO lead does not have a control comparison. Recovery: invest in either a geographic holdout or a synthetic control for the next quarter; show the methodology and limitations explicitly on slide five.

The ask without a stop-doing. The deck asks for incremental investment without identifying what the team will discontinue to free capacity. Recovery: every ask slide must include a stop-doing list; the CMO and CFO both look for it.

The surprise incident. The CFO learns about a hallucination or misattribution event from another source — a sales rep, a customer, a competitor's marketing — that the AEO QBR did not surface. Recovery: maintain a formal incident log on slide eleven and brief the CFO in private when material incidents occur, not at the QBR.

Each failure mode is preventable, and each recovery is structural rather than cosmetic. The QBRs that survive are not the ones with prettier charts. They are the ones built on the structural conventions Bain, McKinsey, and Gainsight refined for the audiences they serve. AEO is new; the discipline of presenting evidence to a CFO is not.

Takeaway: The AEO QBR that survives the CFO's scrutiny is not a marketing deck with AI metrics pasted in. It is a Bain-style steering committee document, a Gainsight-style success-plan review, and an HBR-style assertion-first report combined into twelve slides addressing three audiences with three distinct decision logics. Methodology disclosure precedes financial numbers. Contribution margin and control comparisons precede share-of-voice. Strategic posture precedes tactical execution. The ask is paired with a stop-doing list. Vanity metrics live in the operational dashboard, not the QBR. The cadence between quarters is monthly written reviews and weekly pulses, not silence followed by a forty-five-minute presentation. AEO programs are not killed because they fail to produce results; they are killed because their QBRs fail to defend the results they produce. The template above is the defense.