We
Build
Certainty.

At Taro Labs, we believe learning should never be left to chance. We imagine a future where AI joins humans as a cognitive co-pilot, turning fleeting information into permanent, perfectly structured knowledge. While Tarot speaks of luck and fortune, we build certainty through technology.

Go to Our Product

What We Build

Our product, nocose, listens with you. It records lectures as they happen, turns speech into faithful transcripts, and distills hours of talk into structured study notes — so nothing you hear is left to memory alone.

Recording
00:00.00
LIVE TRANSCRIPT

Today we’ll pull together all the frameworks and analysis tools from the past seven weeks in one go.

Let’s start by revisiting the FIRST, RACE, and DMS frameworks from weeks one and two.

In Agentic AI Commerce the buying journey itself changes, so completing the strategy sheet step by step is the key.

Goal setting follows the SMART principle: specific, measurable, achievable, relevant, and time-bound.

For the situation analysis we use Five Forces, personas, and RFM to segment the customers.

AEO structures the data so the AI reads attributes, features, and social proof accurately.

Big Data Marketing — Framework Review 2026. 04. 20 15:11 | 23 min

Big Data Marketing — Framework Review

This lecture connects the FIRST, RACE, DMS, SMART, and RFM frameworks into one exam-ready study path.

📎 Administrative notices and roll call were omitted so the note stays focused on lecture content.
Weeks 1–2 Review: Frameworks and Agentic AI Commerce
Digital marketing frameworks such as FIRST, RACE, and DMS become useful only when they are applied to a concrete Agentic Commerce case.
Core frameworks
  • FIRST framework: an analysis framework built from an acronym
  • RACE model: Reach → Act → Convert → Engage stages
  • DMS: a digital marketing strategy framework (also an acronym)
Goal setting: the SMART principle
ElementMeaningKey point
SpecificSpecificityThe goal must be clearly defined
MeasurableMeasurabilityCan it be measured numerically?
AchievableAttainabilityIs it realistically possible?
RelevantRelevanceDoes it tie to a business goal?
Time-boundDeadlineIs there a due date?
📌 Professor’s emphasis: it’s not enough to know the frameworks — you must be able to apply them to real Agentic Commerce cases. “Expansive thinking” is required.
Week 3 Review: RFM, Customer Journey Map, Data Strategy
RFM scores shift by customer segment, so targeting strategy should change as recency, frequency, and monetary value move.
RFM segmentation in depth
  • R, F, and M are each scored on a 1–5 scale
  • Champion segment: high on all of RFM → top tier
  • At-risk segment: low on all, or high on only some
  • As the scores shift, segment rankings change, and so does the targeting strategy
Three strategies for feeding data to AI
  • Reliability & accessibility: give AI structured data it can trust
  • Context optimization: capture user intent and context, not just product specs — and exclude noise (one-off events)
  • Social-proof integration: feed popularity and reviews, not product attributes alone
💡 Why RFM values move differently per segment: customer behavior patterns vary — e.g. someone who buys often but spends little.
Week 4 Review: Visualization and Problem Definition
Good analysis starts with a precise problem, then moves through planning, collection, preprocessing, visualization, and interpretation.
Chart types
ChartMain use
Line graphTrends, daily patterns, goal tracking
HistogramDistribution shape, outlier detection
Area chartTime-window filtering, change over time
Box plotQuartile distribution, outlier range
Scatter plotCorrelation between variables
Five-step analysis process
  • Problem definition — state the problem clearly
  • Analysis planning — decide which data and tools to use
  • Data collection
  • Preprocessing — handle missing values and outliers, encode (categorical → numeric), scale to 0–1
  • Visualization & interpretation
📌 For the exam: compare a good goal with a vague one, and explain why measurability changes the next action.
Key terms
TermDefinitionNote
Agentic CommerceCommerce driven by AI agentsDifferent buying journey
AEOStructuring data so AI reads product info accuratelyWeeks 1–2
SMARTSpecific, Measurable, Achievable, Relevant, Time-boundWeeks 1–2
RFMRecency, Frequency, Monetary segmentationWeek 3
RACEReach → Act → Convert → EngageWeek 3
Exam focus
  • Meaning of each framework acronym: FIRST, RACE, DMS, SMART, RFM
  • The SMART principle: each element’s definition and good vs. bad goal examples
  • Interpreting RFM values: why they move differently per segment and how strategy shifts
  • The five-step analysis process and each step’s core activity
Self-test
Q1: How do the RFM value patterns of the champion and at-risk segments differ?

The champion segment is high on R, F, and M. The at-risk segment is low on all, or high on only some. The key is that the three factors don’t necessarily move in the same direction.

Q2: Why is a measurable SMART goal easier to act on than a vague one?

A measurable goal gives the team a number to compare against, so the next action can be chosen from evidence instead of preference.

Q3: What should be filtered out before feeding lecture data into an AI summary?

One-off noise, roll call, seating notices, and other non-lecture spans should be removed so the model focuses on reusable concepts.

Q4: Why does AEO need structured data instead of only plain product text?

Structured attributes, context, and social proof make it easier for an AI agent to read, compare, and rank information accurately.

Today we’ll pull together all the frameworks and analysis tools from the past seven weeks in one go.

Let’s start by revisiting the FIRST, RACE, and DMS frameworks from weeks one and two.

In Agentic AI Commerce the buying journey itself changes, so completing the strategy sheet step by step is the key.

Goal setting follows the SMART principle: specific, measurable, achievable, relevant, and time-bound.

For each step, identify where customers drop off and what change you’re facing at that point.

The situation analysis uses Five Forces, personas, RFM, and channel performance together.

RFM stands for Recency, Frequency, and Monetary value — it’s the backbone of segmentation.

Put a good goal and a bad goal side by side and the difference becomes obvious.

A measurable, time-bound goal is far easier to act on than a vague aspiration.

Now the STP flow: segmentation, then targeting, then positioning, in that order.

You don’t just differentiate one product — you slice the whole set of segments layer by layer.

That layered segmentation is what lets you position each offer for a specific group.

AEO, Answer Engine Optimization, structures your data so AI reads it accurately.

Provide attributes, content, features, and social proof — not just a bare product spec.

Structured versus unstructured data is the key distinction the AI cares about.

Let’s move on to week three and go deeper on RFM.

RFM scores three axes — recency, frequency, monetary value — each on a scale of 1 to 5.

A champion segment is high on all three; an at-risk segment is low, or high on only some.

Key Features

Built so that nothing between your schedule and your notes is left to chance.

Interpretation in 50+ languages

Lectures are transcribed and simultaneously interpreted across more than fifty languages — follow a talk in one language, study it in another.

Formulas, tables, and code

Study notes keep equations, tables, and code blocks intact and readable, instead of flattening them into plain text.

Scheduled from your calendar

Connect your calendar and nocose pre-arranges a recording session for every lecture on it — ready before the class begins.

A call so you never forget

Right before a scheduled session, a call agent rings your phone. Answer the call or press a single button, and recording starts on the spot.

The Team

A small team of builders and researchers who believe knowledge should compound, not fade.

Tao

Founder & CEO

Based in Seoul / Osaka

Ryan

Co-founder

Based in Seoul / SF