Use case #0003

Board marketing packs: the monthly report Marketing Analytics AI assembles automatically

The board marketing report is typically assembled by the marketing team in the final week of each month — pulling numbers from the ad platforms, the CRM, the LOS, and wherever the attribution data lives, formatting them into a presentation, writing the commentary, and producing a document that arrives at the board with one week's worth of effort embedded in it. The Marketing Analytics Agent AI assembles this report on the first working day of every month, automatically, from the same sources, with the commentary generated from the data — so the board receives a current, accurate, decision-ready marketing intelligence pack the morning after the month closes, without anyone spending a week producing it.

What the board needs from a marketing report — and what it usually receives instead

The board needs to make two decisions from the marketing report: whether the marketing budget is being well spent (efficiency question) and whether the marketing programme is building the pipeline the institution needs to meet its disbursement targets (adequacy question). These are strategic questions that require business metrics — cost per funded loan, ROMI, pipeline volume, channel contributions to target. Most marketing reports that reach boards answer neither question cleanly. They contain impression counts, click-through rates, and campaign descriptions that require the board to do the translation to business impact themselves — a translation most board members cannot do because they do not know the intermediate conversion rates that connect impressions to funded loans.

The Marketing Analytics Agent AI writes the board report in the board's language from the start: SGD spent, loans funded, cost per loan, return on investment, and the pipeline position against the disbursement target. The supporting data for each conclusion is present for board members who want to see it, but the headline is always a business outcome, not a channel activity.

"A board that receives a marketing report with impression counts and click-through rates is being asked to trust the marketing team rather than evaluate it. The board pack that leads with cost per funded loan and ROMI asks the board to make a judgement — which is the board's actual job."

The November board marketing pack: assembled December 1, 2025

Marketing Analytics Agent AI · Auto-assembled Dec 1, 2025 · 06:00 AM Monthly Marketing Intelligence Report — November 2025 Singapore finance company · For Board and MD · Confidential · 7 sections · Data through Nov 30, 2025
Section 1 — Executive summary · The one-page board read
November: 4.8× ROMI on SGD58.4L spend · 381 funded loans · CAC SGD15,300 vs SGD18,000 target
Marketing spend in November produced 381 funded loans at an average cost of SGD15,300 per loan — 15% below the Board-approved target CAC of SGD18,000. Total disbursements attributable to marketing channels: SGD134 million. The Net Interest Margin contribution of the funded cohort over their expected tenor is approximately SGD2.8 million, against SGD58.4 hundred thousand spent — a 4.8× marketing ROMI. The portfolio is on track to meet the December disbursement target of SGD160 million given the current pipeline position.
SGD58.4LTotal marketing spend
381Funded loans (attributed)
SGD15,300 vs SGD18K targetBlended CAC
4.8×Marketing ROMI
Board note: The LAP Google Search campaign ran 14 days above CAC target before reallocation on Nov 14. SGD8.4L was spent above efficient level before the rebalance. The rebalance produced an immediate CAC improvement. Board recommendation: approve a standing delegation to the Marketing Analytics AI to reallocate spend above 130% CAC threshold within the existing budget without board pre-approval, with same-day notification.
Section 2 — Channel performance vs CAC and ROMI targets
Lifecycle email delivers 76× ROMI — the institution's highest-return marketing asset by far
November channel performance sorted by ROMI. The standout finding is lifecycle email at 76× ROMI — existing borrower communications that produce pre-approval conversions at SGD840 per funded loan. This channel operates at near-zero marginal cost (email send cost) on a pre-qualified, pre-consented audience. It requires no additional budget to scale — it requires only that the pre-approval eligibility model be expanded to identify more eligible borrowers.
ChannelSpend (SGD)Funded loansCACROMIvs target
Lifecycle emailSGD1,84,000219SGD84076×+95.3%
Google BrandedSGD8,40,000107SGD7,8508.1×+56.4%
referral partner commissionsSGD12,00,000107SGD11,2155.7×+37.7%
Google Non-BrandSGD9,60,00068SGD14,1184.5×+21.6%
Meta RetargetingSGD6,40,00040SGD16,0004.0×+11.1%
Meta ProspectingSGD8,40,00049SGD17,1433.7×+4.8%
WhatsApp broadcastSGD3,60,00016SGD22,5002.8×−25.0%
Google LAP (paused D14)SGD8,20,00023SGD35,6521.8×−98.1%
Section 3 — Pipeline and disbursement target tracking
December target SGD160 Cr: current pipeline (sanctioned, not yet disbursed) covers SGD142 Cr — 11% gap requires action
The December disbursement target is SGD160 million. Current pipeline: SGD142 million sanctioned, of which SGD126 million is expected to disburse in December (based on historical drop-off rates between approval and disbursement). The SGD34 million gap requires either accelerating existing sanctioned loans to disbursement, or generating and sanctioning SGD28–34 million of new leads within the first 10 working days of December. Based on the current CAC of SGD15,300, achieving this gap through paid channels would require an additional SGD40–52 hundred thousand in marketing spend.
Board action required: Approve SGD40L incremental marketing spend allocation for December — targeted to Google Branded and lifecycle email (highest-ROMI channels). Expected incremental funded loans: 26–34. Expected incremental disbursement: SGD9–12 million. This allocation covers the target gap and brings December on track.
Section 4 — Attribution intelligence: what actually drives conversions
Data-driven attribution: referral partner call + lifecycle email together produce 2.8× higher funded rate than either alone
The data-driven attribution model analysed 381 funded borrower journeys in November. The combination of a referral partner personal contact and a lifecycle email within the same 30-day window was present in 68% of all conversions and absent in 91% of all dropped journeys. This combination effect — a human conversation + a digitally delivered pre-qualification — is the institution's single most powerful conversion mechanic. Budget implications: every incremental SGD1 invested in referral partner enablement (training, tools, commission incentive to engage existing borrowers) combined with lifecycle email qualification is currently outperforming the next-best channel pair by 2.8×.
Section 5 — Portfolio quality of November funded cohort
November cohort early DPD rate: 2.1% at 30 days — significantly below historical 4.8% average
The 381 funded loans from November marketing show a 30-day DPD rate of 2.1% — well below the historical portfolio average of 4.8%. This quality improvement reflects the shift to eligibility-gated outreach (only pre-screened borrowers receive offers) and the concentration of funded loans from referral partner and lifecycle email channels, both of which show historically lower early DPD rates than bulk prospecting channels. The WhatsApp bulk broadcast channel, despite above-target CAC, shows the highest early DPD rate in the November cohort (6.8%) — reinforcing the case for reducing or eliminating bulk broadcast in favour of personalised triggers.
Section 6 — Competitive and market context
Competitor rate movements: SBI home loan rate unchanged; two private banks increased rates 25bps in November
Market intelligence from SERP monitoring and competitor bid analysis shows that two private sector banks increased their home loan rates by 25 basis points in November. The institution's current home loan rate of 10.50% is now 25–50bps below these competitors for equivalent borrower profiles. This rate advantage is not being fully reflected in marketing messaging — current ad copy leads with instalment amounts; an explicit rate comparison message ("10.50% vs market 10.75%–11.00%") may improve conversion rates on Google non-brand and Meta campaigns. A/B test recommended for December.
Marketing recommendation: Activate rate-comparison creative variant in December for Google non-brand and Meta prospecting campaigns. Expected CTR improvement: 8–12% based on prior rate-comparison tests. No incremental budget required.
Section 7 — December outlook and budget recommendation
December recommended spend: SGD58.4L base + SGD40L incremental = SGD98.4L total · Target: 520 funded loans
Maintaining the November channel mix with LAP Search excluded, adding the SGD40L incremental allocation to Google Branded and lifecycle email, and activating the rate-comparison creative for non-brand campaigns, the projected December outcome is 510–530 funded loans at an estimated CAC of SGD14,800–15,800 — an improvement on November's SGD15,300 and well within target. Total attributable disbursement projection: SGD178–185 million, against the SGD160 million target.
● Auto-assembled Dec 1, 2025 · 06:00 AM · Data through Nov 30 · All figures from LOS + ad platform integration · No manual data entry ● Next report: Jan 1, 2026 · Auto-delivered to distribution list
Dec 1Report assembled — morning of the first working day after month close · Board has data before the first December meeting
7Sections auto-generated — executive summary, channel performance, pipeline tracking, attribution, portfolio quality, market context, outlook
ZeroManual data assembly required — all sections drawn from LOS + ad platform + CRM integration · No analyst hours consumed
SGD40LBoard action triggered by December pipeline gap analysis — specific, quantified, with projected ROI · Board makes a real decision, not a discussion

A board pack that arrives on December 1 is a board pack that informs December decisions — one that arrives on December 15 is a board pack that documents November history

The traditional marketing report arrives mid-month because assembling it takes a week. By the time the board reads it, the month it describes is half over and the decisions it should inform — budget reallocation, pipeline acceleration, creative testing — have already been deferred by the production delay. The Marketing Analytics Agent AI eliminates the production delay entirely: the report is assembled on the night the month closes, delivered on the morning of the first working day of the new month, and the decisions it recommends can be implemented in the first week of December rather than the third. Board-ready marketing intelligence that arrives when the board can still act on it is the only version of marketing intelligence that is worth producing.

← Back to Marketing Analytics Agent AI