Completions

For CMOs + CFOs + FP&A + marketing-ops directors

The CMO dashboard was the wrong deliverable. The monthly executive summary was.

Most CMOs spend three weeks of every month drafting last month executive summary. Domo, Tableau, Power BI, Looker, ThoughtSpot render the metrics. None of them draft the narrative. The auto-drafted exec summary delivered on close+1 is the deliverable the CEO and CFO actually read.

By Jay Christopher11 min read

What this gets you

  • Day-one delivery of the monthly executive summary — auto-drafted on close+1 from rollup data + attribution rollup + forward-look + LLM narrative commentary. Three weeks of drafting compresses to one day of CMO review + CFO sign-off.
  • 6-axis cadence-stratified reporting pipeline — Rollup + Recommend + Attribute + LP-Letter + Board-Deck + Exec-Summary. Same underlying rollup substrate, different cadence and different per-stakeholder framing.
  • Per-banner + per-vertical context auto-generated — multi-banner operators get per-banner sections plus the cross-banner parent rollup. Multi-vertical operators add per-vertical narrative tone tuned to each banner type.
  • Variance commentary + forward-look + cross-functional reconciliation — auto-explain every variance above the materiality threshold, pull the forward-look section from the recommendation engine, reconcile cross-functional input from Finance and Ops before draft assembly.
  • Approval workflow + versioning + distribution analytics — CMO → CFO → CEO sign-off chain, every version audit-trailed, distribution open-rate and read-time tracked per stakeholder.

By the time the exec summary lands, the data is four weeks old

A CMO at a 200-location multi-banner operator runs the monthly executive reporting cycle on the same calendar every month. The finance team closes the books during the first week. The channel reconciliation runs during the same week. The attribution rollup waits on both. The CMO opens the Domo dashboard on Tuesday of week two and reviews the previous month numbers for the first time.

Week two: variance commentary drafting. Every metric that moved more than five percent needs a written explanation. Pipeline down 12 percent — was that the brand-spend cut, the channel reallocation, the franchisee soft-month in the Southeast, or the calendar effect from one fewer Saturday? The CMO writes paragraphs, pings the channel leads for context, redrafts.

Week three: cross-functional reconciliation and approval cycles. The Finance team has its own monthly. The Ops team has its own monthly. The numbers do not perfectly agree. The CMO meets with the CFO to reconcile. The CFO routes to the CEO. The CEO sends back two questions and a re-rewrite request. The CMO redrafts again.

Week four: distribution. The final monthly executive summary lands in the CEO inbox on Friday of week four. The data inside is now four weeks old. The next month already started 28 days ago. The CMO begins drafting next month executive summary the following Monday.

The auto-drafted exec summary compresses the three-week draft-cycle to one day. The data lands at close, the LLM drafts the narrative at close+0 hours, the CMO reviews close+1, the CFO signs off close+1, the CEO reads close+2. The dashboard renders the metrics; the auto-draft layer renders the narrative.

What is in market — and what each category leaves to you

The dashboard layer is mature. The auto-draft layer is operator-side wiring.

Enterprise BI — Domo, Tableau, Microsoft Power BI, Looker Studio, ThoughtSpot, Sisense, Qlik, Klipfolio

Excellent at rendering the dashboard. Every metric the CMO needs to see is one drill-down away. The auto-drafted narrative commentary, the variance explanations, the forward-look section, the per-banner context, the executive-tone rewrite — jobs that today are done by a human CMO over three weeks — sit downstream of the dashboard and are operator-side work.

Marketing-specific BI — Funnel.io, Improvado, Adverity, Supermetrics, NinjaCat, Whatagraph

Strong at marketing-channel data consolidation and template-based monthly reports. The auto-narrative LLM layer that drafts variance commentary, reconciles cross-functional input, and tunes per-banner tone is not shipped as a product feature; the templates render data, they do not auto-write the executive narrative.

Financial-planning platforms — Causal, Cube, Mosaic.tech, Workiva

Strong at financial close, FP&A modeling, and finance-narrative reporting. Marketing-specific variance-commentary auto-draft against rollup-reporting substrate plus cross-functional reconciliation with Ops plus per-banner narrative tuning is marketing-domain wiring the FP&A platforms do not orchestrate.

Open-source / SMB — Apache Superset, Metabase, Grafana, Google Looker Studio (free tier)

Capable of rendering CMO dashboards for operators below the enterprise tier. Same gap as the enterprise tools at the auto-narrative layer.

The Google Doc the CMO drafts every month

The status quo at most multi-location operators. The CMO opens a fresh Google Doc on the first Tuesday of every month, pastes screenshots from Domo, hand-types variance commentary, pings channel leads for context, reconciles with Finance, routes through approval, and emails the final to the executive distribution list 21 days later. Repeats next month.

The pipeline, end to end

  1. 6-axis cadence-stratified topology. Rollup (per-location and per-channel KPI roll-up, daily and weekly cadence), Recommend (forward-looking spend recommendations, weekly cadence), Attribute (cross-location attribution rollup, weekly cadence), Exec-Summary (monthly cadence), Board-Deck (quarterly cadence), LP-Letter (quarterly cadence with PE-sponsor framing). Same underlying rollup substrate, stratified by stakeholder cadence and stakeholder format.
  2. Pre-close data freshness gate. Finance close target date defines the trigger. The pipeline waits for finance-closed-flag plus channel-reconciliation-complete-flag plus attribution-rollup-complete-flag. When all three fire, the auto-draft job triggers within the hour.
  3. Variance detection. Every metric in the exec-summary template compared against the prior-period baseline, the prior-year same-period baseline, and the forecast. Variances above the materiality threshold (default five percent on revenue, three percent on margin, ten percent on channel ROAS) flag for narrative commentary.
  4. LLM narrative commentary per variance. Each flagged variance routes to an LLM with the metric, the period comparison, the relevant context (channel attribution shift, per-location performance, calendar effect, channel reallocation events from the forward-look engine), and a tone-tuned prompt. The output is a one-to-three sentence variance explanation in the executive register.
  5. Forward-look section from the recommendation engine. The forward-looking recommendation axis (per-market spend recommendations, channel reallocation proposals, capacity-utilization alerts) supplies the exec-summary forward-look section directly. The CMO does not draft the forward-look from scratch; the CMO edits the recommendation-engine output.
  6. Cross-functional input reconciliation. Finance monthly and Ops monthly land in parallel. The reconciliation step diffs the marketing-attributed revenue against finance-recognized revenue against ops-fulfilled revenue and surfaces gaps before draft assembly. Gaps that exceed the threshold escalate to the CMO before the LLM draft fires.
  7. Per-banner + per-vertical narrative split. Multi-banner operators run the variance-commentary pipeline per banner. Per-banner sections assemble into a parent exec-summary with cross-banner rollup at the top. Per-vertical operators tune the LLM prompt for vertical- specific tone (gym banner uses retention language, spa banner uses booking language, restaurant banner uses cover- count language).
  8. Draft assembly. Cover page, executive summary, per-banner sections, channel rollup, attribution rollup, variance commentary, forward-look, cross-functional reconciliation footnote, appendix with raw rollup tables. The draft is a structured document, not a rendered PDF.
  9. CMO review and edit. The CMO opens the draft on close+1, reviews each section, edits where the LLM tone needs adjustment, adds insider context the LLM could not generate (channel-lead conversations, board conversations, competitive intelligence). Edit time target: 60-90 minutes.
  10. CFO sign-off. The CFO receives the CMO-edited draft, reviews the variance commentary and the cross-functional reconciliation footnote, approves or requests revisions. Approval triggers downstream distribution.
  11. Distribution and analytics. The final exec summary distributes to the executive list (CEO, Board chair, Finance chair, operations VPs) via email plus the executive portal. Open-rate, read-time, and section-by-section attention tracked per stakeholder. Signal feeds future-month tone and section tuning.
  12. Roll-up into board-deck and LP-letter axes. Three months of monthly exec summaries roll up into the quarterly board deck (cross-link to the dedicated board-deck axis). The PE-sponsor LP letter reframes the same substrate with sponsor-specific narrative. The substrate is computed once and consumed by all three cadence-stratified reports.
  13. Production-efficiency measurement. Days-from-close-to-CMO-edit, edit-time-per-section, CFO approval cycle time, distribution time, executive read rate. The signal feeds prompt-tuning, variance-threshold tuning, and per-banner template tuning per cycle.

Frequently asked

What is a CMO dashboard?

A CMO dashboard is the executive view of the marketing operation — pipeline by channel, CAC and LTV trends, campaign ROAS, attribution waterfall, marketing-influenced revenue, brand-health metrics, and team-capacity utilization. The category leaders — Domo, Tableau, Power BI, Looker, ThoughtSpot, Sisense, Qlik — render the metrics beautifully. The narrative that the CEO and CFO actually read at the start of every month — the monthly executive summary — is downstream of the dashboard and is the actual deliverable.

Why does the monthly executive summary take three weeks to draft?

The data lands during week one (finance close, channel reconciliation, attribution rollup). The CMO reviews the dashboard during week one, writes variance commentary during week two, reconciles cross-functional input from Finance and Ops during week two, routes through approval cycles with the CFO and the CEO during week three, and distributes the final summary to the executive distribution list at the start of week four. By the time the executive summary lands, the data is four weeks old and the next month already started.

How do you automate monthly executive summary drafting?

The architecture is a 6-axis cadence-stratified reporting pipeline. Rollup (per-location and per-channel KPI roll-up, daily and weekly), Recommend (forward-looking spend recommendations), Attribute (cross-location attribution roll-up), LP-Letter (PE sponsor and LP investor reporting), Board-Deck (quarterly board governance reporting), and Exec-Summary (monthly CMO and CEO and CFO executive summary). The exec-summary axis reads from the rollup data, the attribution rollup, and the forward-looking recommendation engine. An LLM drafts the narrative commentary on the variances, the cross-functional input from Finance and Ops, and the forward-look section. The CMO reviews the draft, the CFO signs off, the CEO reads it. The full cycle compresses from three weeks to one day.

How is this different from Domo, Tableau, Power BI, Looker, ThoughtSpot, or Sisense?

Those platforms render dashboards. They show the metrics. They do not auto-draft the narrative — the variance commentary, the forward-look, the cross-functional reconciliation, the per-banner context, the executive-tone rewrite. Those are jobs that sit downstream of the dashboard and that today are done by a human CMO over three weeks. The 6-axis reporting pipeline puts the auto-draft layer alongside the dashboard layer.

How does monthly executive summary tie to the quarterly board deck and the LP letter?

Same substrate, different cadence and different format. The monthly executive summary is the substrate the quarterly board deck assembles from (three months of monthly exec summaries roll up into the board deck, with the board-governance reframing). The LP letter is the same substrate further reframed for the PE sponsor or LP audience. The 6-axis cadence-stratified topology means the underlying rollup data is computed once and consumed by multiple per-stakeholder per-cadence reports. The narrative-commentary LLM is per-stakeholder tuned.

How does this work for multi-banner or multi-vertical operators?

Per-banner and per-vertical context is encoded in the rollup data and propagates through the auto-drafted narrative. The monthly executive summary for a multi-banner operator includes per-banner sections (each banner gets its own variance commentary, its own forward-look, its own cross-functional reconciliation) plus the cross-banner rollup at the parent level. Multi-vertical operators add per-vertical context (the gym banner reports differently than the spa banner differently than the restaurant banner under the same parent). The narrative is generated per-banner-per-section and assembled into a single document with the parent-level summary on top.

Hire the agent that drafts the exec summary

The per-location-rollup-reporting agent owns the 6-axis cadence-stratified reporting pipeline — Rollup + Recommend + Attribute + LP-Letter + Board-Deck + Exec-Summary — sitting on top of whichever BI platform you license downstream. Monthly exec summary auto-drafts on close+1, board deck assembles from three monthly summaries, LP letter reframes for the sponsor audience. Same substrate, three per-stakeholder reports.

Hire the rollup-reporting agent

We scope on the call and send a private checkout link after.

Related reading: Quarterly board deck · Cohort-framed KPI rollup