Twelve weeks. We start with foundations (why marketing exists at all), then move into analytics tools, social media data, generative AI, and predictive modelling.
WEEK 1
Marketing Frameworks and Evolution
WEEK 2
Marketing Analytics, Trends and Software
WEEK 3
Social Media Analytics: Methods and Capabilities
WEEK 4
Social Media Content and Engagement Analysis
WEEK 5
Assessment
WEEK 6
Influencer Marketing and Network Analysis
WEEK 7
Generative AI for Marketing Analytics
WEEK 8
Data Visualisation and Storytelling
WEEK 9
Marketing and Social Media Strategy Optimisation
WEEK 10
Customer Lifetime Value and Churn
WEEK 11
Predictive Modelling
WEEK 12
Assessment
0.3
What you'll get out of today
Outcome
By the end of this lesson you should be able to…
LO1
Explain the three core marketing frameworks (3Cs, STP, 4Ps) in plain language.
LO2
Use data about customers, your own company, and competitors to make better marketing decisions.
LO3
Describe how promotional channels have shifted from billboards and TV to mobile-first digital experiences.
LO4
Map the consumer decision journey (ABCDE) onto the company's conversion funnel (AIDA).
Why this matters
Frameworks are the scaffolding. Every data-driven marketing decision later in the course (recommender systems, sentiment analysis, churn prediction) plugs into one of these three frameworks.
Section 1
What is Marketing?
The one-line definition, and why analytics changed the game
1.1
Marketing in one sentence
Definition
Marketing is the work of matching what customers need with what a company can profitably offer — and communicating that match convincingly.
Analogy — the matchmaker
Think of a marketer as a matchmaker. On one side there are people with problems, wants, and budgets. On the other side there are companies with products, services, and expertise. The matchmaker's job is to find pairings that make both sides happy — and to do it more reliably than the matchmaker down the street.
Marketing is not just advertising. Ads are the tip of the iceberg. Underneath is research, segmentation, pricing, distribution, and post-purchase service.
1.2
Where analytics fits in
Historically, marketers made educated guesses. Today, they measure.
Then (pre-2000)
TV ratings from small survey panels
Rough sales figures by region
Quarterly customer surveys
"Half the money I spend on advertising is wasted; the trouble is I don't know which half." — attributed to John Wanamaker
Now (data-driven era)
Every click, scroll, and pause is logged
A/B tests run continuously on live traffic
Real-time dashboards per campaign, per segment
Recommendations personalised per user
Data point
Netflix has publicly stated that around 80% of what viewers watch comes from its recommendation system — not from browsing or search. That is marketing analytics doing the matchmaking, at scale, for over 300 million subscribers.
Section 2
Framework #1 — The 3Cs
Customers, Company, Competitors
2.1
The three lenses
To find where you can win, you need to look at any business problem through three lenses at once.
Figure 2.1: The 3Cs overlap where a winning strategy lives.
Customers — what do they actually need?
Company — can we deliver it better than anyone else?
Competitors — who else is offering something similar, and how are they different?
2.2
Customers — what problem are we solving?
Every product exists to solve some problem or satisfy some want. If you can't finish this sentence in one line, you don't understand your customer yet:
Try this test
"Our customer struggles with ______, and today they solve it by ______, which is bad because ______."
Worked example — Uber (2009)
Customer struggles with flagging a taxi in the rain. Today they solve it by standing on the curb waving. Which is bad because it's slow, unreliable, and they can't see how far away the car is. Uber's whole product design came from that one sentence.
2.3
Company — what can only we do?
The second lens turns inward. What resources, skills, or history does your company have that competitors don't?
Technology (patents, algorithms, infrastructure)
Data (proprietary datasets, customer history)
Brand (trust, recognition, emotional connection)
Distribution (physical stores, logistics network)
People (specialised expertise)
Data point — Amazon's advantage
Amazon operates over 175 fulfilment centres worldwide and holds roughly 200 million Prime members. A new e-commerce entrant can copy the website in a weekend. Copying that logistics network and membership base takes decades.
2.4
Competitors — who else is playing?
The third lens asks: who else is chasing the same customer, and how do we look different?
Direct competitors sell essentially the same thing (Coca-Cola vs Pepsi). Indirect competitors solve the same problem differently — orange juice, water, and energy drinks also compete for "thirsty consumer" spend.
Analogy — the food court
Imagine you own the sushi stand in a food court. Your direct competitor is the other sushi stand. But your indirect competitors are the burger place, the pho stand, and even the person who packed a sandwich from home. Miss the indirect ones and you'll lose share without knowing why.
2.5
Worked example — Sam flies MEL to ADL
Sam needs to get from Melbourne to Adelaide next Friday. Let's run the 3Cs from Qantas's point of view.
Lens
What Qantas asks
Data that helps answer it
Customer
What matters to Sam? Speed, price, luggage, timing, frequent-flyer points?
Booking history, search queries on qantas.com, survey NPS scores
Company
What can Qantas offer that hits those needs? Lounge access, direct flights, mobile check-in.
Route profitability data, on-time performance stats, capacity by day
Competitor
What do Virgin Australia, Jetstar, coach services, and driving offer at the same time?
Competitor fare-scraping data, industry route-share reports
The MEL–ADL route is roughly 725 km. Driving takes about 8 hours; flying takes about 1 hour 20 minutes. That gap is Qantas's core promise on this route — but only for time-sensitive travellers.
2.Q
Knowledge check — Section 2
Q1. A student asks, "Isn't the 3Cs framework just common sense?" What's the best response?
A framework's power is not novelty — it's discipline. Companies routinely build great products (strong Company view) that nobody wants (missed Customer view) or that a competitor has already dominated (missed Competitor view).
Q2. Coca-Cola treats bottled water as which type of competitor?
Bottled water solves the same underlying "quench thirst" need. That's why Coca-Cola actually owns several water brands (Dasani, Smartwater) — competing with itself is safer than losing the customer entirely.
Section 3
Framework #2 — STP
Segmentation, Targeting, Positioning
3.1
What STP does for you
The 3Cs tell you where to look for a winning position. STP tells you how to actually pick and reach the right customers.
Step
Question it answers
Segmentation
What natural groups exist in the market?
Targeting
Which of those groups should we go after?
Positioning
What do we want to be known for in the minds of that group?
Analogy — fishing
Segmentation is figuring out which species live in the lake. Targeting is deciding which species you actually want to catch. Positioning is choosing the right bait so those fish bite you and not your competitor's line.
3.2
Segmentation — divide the market
Definition
Segmentation uses data to divide a market into groups of people who share similar characteristics — and who differ meaningfully from other groups.
A good segment is:
Internally similar — people inside behave alike
Externally different — clearly distinct from other segments
Measurable and reachable — you can find, count, and communicate with them
Big enough to matter — worth the marketing spend
Data point — Spotify Wrapped
Every December, Spotify's "Wrapped" campaign is essentially segmentation made public: it slices ~600 million users into playful groups ("Top 0.1% of Taylor Swift listeners", "Indie folk enthusiast") based on 12 months of behavioural data. Users love sharing which group they landed in — free viral marketing built entirely on segmentation.
Demographics tell you who someone is. Behaviour tells you what they'll actually do. When you have both, behavioural data almost always predicts better — but demographics are easier to collect.
3.4
How data scientists actually segment
Segmentation used to be done with intuition and simple crosstabs. Today, it's done with cluster analysis — unsupervised machine-learning algorithms that group customers automatically based on many variables at once.
Figure 3.4: A 2-D projection of three clusters found by k-means on customer behaviour data.
Common algorithms: k-means (fast, needs number of clusters upfront), hierarchical clustering (produces a dendrogram tree), DBSCAN (finds arbitrarily shaped groups). We'll implement these in later weeks.
3.5
Targeting — which segment(s) do we chase?
Once you have segments, targeting picks the ones worth pursuing. Two questions decide it:
Can we serve this segment better than anyone else? (Company fit)
Is this segment financially worth it? (Size × spend × margin, minus cost to serve)
Worked example — Tesla's early targeting
In 2008, Tesla launched the Roadster at ~USD 109,000. They targeted a tiny segment: wealthy, environmentally conscious early adopters. That segment was small, but the margin per unit was huge and the buzz created demand for later mass-market cars (Model 3, Model Y). This is called a beachhead strategy — win one small segment first, expand later.
3.6
Positioning — own a spot in the mind
Definition
Positioning is the single idea you want your target segment to associate with your brand — the reason they'd pick you over the alternatives.
Analogy — mental shelf space
Every customer has a mental supermarket. Each product category is an aisle. On each shelf, only a few brand names fit. Positioning is a fight for shelf space — not in the actual store, but inside the customer's head. Once a brand owns "safest car" (Volvo) or "cheapest flights" (Jetstar), it's very hard for a competitor to bump them off.
Positioning examples you already know
Volvo — safety
Apple — premium design and simplicity
Aldi — low prices without embarrassment
Rolex — status and heritage
3.Q
Knowledge check — Section 3
Q1. A retailer segments customers using k-means on purchase-frequency and average-basket-size. Which segmentation type is this?
Both frequency and basket size describe what customers do — that's behavioural data. Demographics would be age/income; psychographics would be lifestyle/values.
Q2. Aldi positions itself as low-cost. Why is this a strong positioning?
Great positioning is (i) a single sharp idea, (ii) consistent across all touchpoints, and (iii) backed by real capability. Aldi's private-label focus and small store footprint make the low prices real, not just advertised.
Section 4
Framework #3 — The 4 Ps
The Marketing Mix: Product, Price, Place, Promotion
4.1
What the 4 Ps are for
Once you've picked your target segment and positioning (STP), the 4 Ps are the levers you actually pull to deliver on that positioning.
Analogy — a recipe
Positioning is what dish you promised the customer. The 4 Ps are the ingredients you use to actually cook it. If you promised a fine-dining experience but chose cheap ingredients (Product), a fast-food price (Price), a food-truck location (Place), and billboard ads at a truck stop (Promotion), the customer walks away confused. Every P must support the positioning.
P
Core question
Product
What features and quality does the offering have?
Price
How much do we charge, and how do we discount?
Place
How do we get the product to the customer?
Promotion
How do we tell the customer we exist?
4.2
Product — data guides what to build
Product decisions used to be driven by executive intuition. Today they're driven by usage data, A/B tests, and voice-of-customer analytics.
Data-driven product example — Netflix thumbnails
Netflix personalises the artwork you see for each show. It runs a multi-armed bandit that shows different thumbnail images to different viewers and learns which image drives more clicks for each user segment. A rom-com might appear with a romantic scene for one viewer and a comedic scene for another — same product, different presentation, tested with data.
Data questions marketers ask:
Which features do our top 20% users actually use?
What are customers complaining about in reviews and support tickets?
Which product variant wins in a live A/B test?
4.3
Price — data reveals what people will pay
Price is the fastest lever to move revenue — and the easiest to mess up. Analytics helps in three ways:
Elasticity models — how does demand change if we raise the price 5%?
Willingness-to-pay surveys — Van Westendorp price-sensitivity meters
Dynamic pricing — real-time adjustments based on demand and inventory
Data point — Uber's surge pricing
Uber's surge algorithm adjusts prices in real time when demand outpaces driver supply. When New Year's Eve hits Sydney at midnight, prices can spike 2–3×. That isn't greed — it's an elasticity experiment running every minute of every day, using GPS-tracked supply and demand signals to clear the market.
Data point — The Big Mac Index
The Economist's Big Mac Index shows that a Big Mac costs about USD 5.69 in the US but only USD 3.50 in South Africa (2024 figures). Same product, geographically differentiated pricing — Place and Price interacting.
4.4
Place — where and how the product reaches the buyer
Where should we open the next store? (Geospatial analysis of foot traffic and competitor locations)
Which warehouses should stock which SKUs? (Demand forecasting)
Which delivery partners give us the best cost-per-parcel? (Logistics analytics)
Data point — 7-Eleven Japan
7-Eleven Japan uses point-of-sale data from ~21,000 stores to restock each store three times a day with a mix tailored to that store's local demand — different sandwiches near office districts, different snacks near schools. Place, powered by data.
4.5
Promotion — how the message reaches people
Promotion is the P we'll spend most of this course on, because it's where digital and social media analytics live. It has two halves:
Component
Question
Message
What are we saying about the product?
Channel
Where do we say it — and does our audience actually live there?
Key point
A great message on the wrong channel is invisible. A weak message on the right channel is annoying. You need both.
4.6
The 4 Ps must be consistent
The classic mistake is treating the 4 Ps as independent knobs. They aren't — they must line up with the positioning.
Brand
Positioning
Consistent 4 Ps?
Rolex
Prestige and heritage
Premium product ✓, high price ✓, jewellers only (Place) ✓, magazine + tennis sponsorship (Promotion) ✓
Software-heavy product ✓, transparent price ✓, own showrooms (no dealers) ✓, CEO tweets instead of ads ✓
Each row is internally coherent. Break one P and the story falls apart — imagine Rolex being sold on Amazon at a discount.
4.Q
Knowledge check — Section 4
Q1. A boutique gym sets its monthly fee at AUD 249 and only opens in premium suburbs. This is an example of…
Both the high price and the exclusive locations signal the same message — this gym is for a premium segment. That's the 4 Ps working together.
Q2. Which of these is a Product-level (not Promotion-level) data question?
Churn caused by a feature is telling you something is wrong with the product itself. The other two are about how you promote or acquire — Promotion decisions.
Section 5
The Evolution of Promotional Channels
From billboards to your pocket
5.1
Three eras of promotion
Era
Dominant channels
Data available
Pre-digital (pre-1995)
TV, radio, print, billboards, direct mail
Panel surveys, Nielsen ratings, coupon redemption
Web era (1995–2010)
Websites, banner ads, email, search engines
Page views, click-through rate, session logs, cookies
Mobile / social (2010–present)
Mobile apps, social platforms, influencer content, video
Real-time location, in-app events, social graph, ad IDs
Data point — where ad money goes
Global digital advertising spend surpassed traditional (TV, radio, print combined) around 2019. By 2024, digital accounts for roughly 70% of all global ad spend, with mobile taking the majority of the digital share.
5.2
What changed with digital: measurement
The biggest shift is not where ads run — it's what we can measure. In the pre-digital era, a billboard's impact was inferred. In the digital era, every impression can be tied to a specific user's later behaviour.
Billboard on a highway
Estimated eyeballs from traffic counts
No idea who actually looked
Can't tell if any viewer bought later
Same message for everyone
Instagram ad
Exact impressions, per-user
Knows demographics, location, interests
Click → landing page → purchase all linked
Different creative per audience segment
5.3
Traditional channels going digital
Traditional channels haven't died — they've absorbed digital capabilities. Modern billboards can now change content in real time based on weather, sports scores, or time of day.
Real example — Target Pharmacy pollen billboards
Target ran digital billboards that displayed the day's local pollen count and promoted allergy medication from the pharmacy. High pollen days triggered the ad automatically. The channel is old (billboard); the data trigger is new. This is contextual advertising in physical form.
Even TV has become data-driven: connected TVs let advertisers show different ads to different households watching the same show.
Section 6
The Consumer Journey
ABCDE — how customers decide AIDA — how companies convert
6.1
ABCDE — the customer's five steps
From the customer's point of view, buying anything more than a snack usually follows five stages:
Awareness — customer recognises a need or want
Browsing — searches for options that could meet it
Consideration — compares the shortlisted options
Decision — picks one and buys
Evaluation — judges satisfaction; decides whether to repurchase and recommend
6.2
Real journey — buying a new phone
Stage
What the customer does
Data trail they leave
Awareness
Their old phone battery keeps dying — friend mentions a new Pixel.
WhatsApp mention, social feed exposure
Browsing
Googles "best phone under $1200 2026", reads reviews on JB Hi-Fi, watches YouTube reviews.
Search queries, product page views, video watch time
Consideration
Narrows to iPhone 17 vs Pixel 10 vs Samsung S26. Reads Reddit threads.
Repeat visits, side-by-side comparison clicks
Decision
Buys the Pixel 10 online during a Boxing Day sale.
Purchase event, credit card, shipping address
Evaluation
Loves the camera, tells friends, writes a 5-star review.
Review text, social posts, referral link clicks
Every column-3 signal is a data point marketers can capture and act on.
6.3
AIDA — the company's four steps
From the company's point of view, the same journey looks like a funnel that leaks at every stage.
Attention — get the customer to notice we exist
Interest — hold them long enough to hear the benefits
Desire — make them want this specific product
Action — get them to buy
6.4
ABCDE and AIDA side by side
Consumer view (ABCDE)
Company view (AIDA)
Common label
Awareness
Attention
Consumer
Browsing
Interest
Lead
Consideration
Desire
Prospect
Decision
Action
Customer
Evaluation
(Customer service / retention)
Advocate
Two sides of the same table
Both frameworks describe the same journey. ABCDE looks at it from the buyer's chair; AIDA looks at it from the marketer's chair. Neither is more correct — but knowing both lets you design campaigns that actually match how customers behave.
6.5
The funnel leaks — and that's where analytics helps
Analogy — a leaky bucket
Imagine pouring water into a bucket with holes at every level. Attention brings 100 people. Only 40 become interested. Of those, 15 develop desire. Only 3 take action. Marketing analytics is about finding which hole leaks the most and patching it first.
Typical funnel metrics:
CTR (click-through rate) — top of funnel health
Conversion rate — bottom of funnel; industry average ~2–3% for e-commerce
Bounce rate — leakage at the browsing stage
Cart abandonment — leakage right before Action
Fixing the biggest leak often gives more revenue than doubling the ad budget.
6.Q
Knowledge check — Section 6
Q1. A customer adds an item to their cart but doesn't check out. Which funnel stage has failed?
Adding to cart shows desire is there. The failure is at Action — checkout. That's why abandoned-cart emails are one of the highest-ROI marketing tactics.
Q2. Why does the ABCDE model include an "Evaluation" step even though the sale is already done?
Retention and word-of-mouth are cheaper than acquisition. Companies with strong evaluation-stage performance grow faster because each new customer brings referrals — turning customers into advocates.
Week 2 dives into marketing analytics software and trends — the concrete tools (Python, GA4, Tableau) that turn today's frameworks into real dashboards.
7.2
One-sentence takeaway
If you remember only this
Marketing is a matchmaker between customer problems and company solutions. The 3Cs help you spot the match, STP helps you choose whom to serve, and the 4 Ps let you deliver on it — all measured, tested, and refined with data at every step.