What See-Think-Do-Care is
The framework was proposed by Avinash Kaushik in the article listed under sources, as a business framework for content and measurement rather than a channel plan. His complaint was that most marketing, and nearly all measurement, was built around the small group of people ready to buy right now, while the far larger addressable audience was either ignored or served with a message meant for someone else. See-Think-Do-Care answers by defining the audience from the widest plausible pool inward.
The clusters are defined by intent, not by demographics and not by channel. See is the largest addressable qualified audience: everyone with a real connection to what you sell. Think is the portion of that audience with some commercial intent. Do is the portion with strong intent to act. Care is a separate cluster of existing customers who have bought more than once. The word "qualified" carries weight. The See audience for a hiking-boot brand is people who walk in hills, not everyone who owns feet.
Framed this way, the framework makes visible how lopsided a typical plan is. Paid search on product terms, product pages, retargeting and checkout work all sit in Do. Most of the audience sits in See and Think, and most businesses have little to say to them. What the framework hides, by design, is sequence. It does not claim people move tidily from See to Do, and reading it as a funnel misses the point. The marketing funnel framework models that movement; See-Think-Do-Care models the audience as it exists at a moment in time.
The four clusters
| Cluster | Who is in it | What they want from you | What a good outcome looks like for them |
|---|---|---|---|
| See | The largest addressable qualified audience: people who share the interest your product serves | Something worth their attention that is not a pitch | They enjoyed or learned something and remember who provided it |
| Think | The part of that audience with some commercial intent: considering, comparing, not yet decided | Help with the decision: honest comparisons, criteria, demonstrations | They understand their options and your place among them |
| Do | The part with strong commercial intent: ready to act | A frictionless path to buying, booking or signing up | They completed the task with no surprises |
| Care | Existing customers who have bought more than once | Service, recognition, and reasons to stay and recommend | They feel the relationship is worth keeping |
Two features of the table matter more than the labels. First, the clusters nest on the left but not on the right: See contains Think and Do, yet the content each wants is distinct. A comparison guide bores the See audience and under-serves the Do audience. Second, Care is defined by behaviour, meaning repeat purchase, not by a loyalty-programme sign-up or an email subscription. A subscriber who has never bought is still in See or Think, however many newsletters they open.
A practical test when sorting: ask what the person would type into a search box, or what they would say if asked why they were on your site. "I like cycling" is See. "Which gravel bike suits mixed terrain" is Think. "Order this bike in this size" is Do. "My bike is due its first service" is Care. The journey stages guide describes the same people from the sequence side; this framework describes them from the intent side.
Content and channels by cluster
The content question for each cluster is "what would this person thank us for". For See, that is material that serves the interest itself: how-to video, entertaining or useful pieces on the subject, tools, and answers to questions people ask long before they shop. Channels follow the content: video platforms, social feeds, interest-targeted display and the informational end of search. A YouTube channel strategy is often the clearest See investment a business makes, because it earns attention on the subject rather than renting it on the product.
For Think, content helps a decision in progress: comparisons that admit where you lose, buying criteria, calculators, reviews, demonstrations and email sequences that answer the questions people raise while they weigh options. Channels are category and comparison search, review sites, email and the parts of social where people ask for recommendations. For Do, the content is the product page, the landing page, the checkout and the brand search ad; the job is to remove friction, not to persuade. For Care, content is onboarding, service answers, account email, community and the occasional genuine thank-you.
Planning content across all four clusters is the subject of the content strategy hub. The TOFU, MOFU and BOFU framework maps formats to funnel stages in more detail; the difference here is that See-Think-Do-Care adds Care, which funnel-stage models usually leave out, and it asks for a content intent rather than a format.
Metrics by cluster
The measurement rule is the part of the framework most worth keeping: judge each cluster's work by what that cluster was asked to do. See content is asked to earn attention from qualified people, so its metrics are about reach among the right audience and the depth of attention it holds, such as qualified reach, watch time, scroll depthHow far down a page a visitor scrolled, as a share of the page's total length., return visits and growth in branded interest. Judging a See video by purchases is the error the framework exists to prevent.
Think content is asked to move people towards a decision, so its metrics are intermediate actions: comparison pages viewed, calculators used, guides downloaded, email sign-ups, and the share of Think visitors who later appear in Do. Do work is judged by completion: CVRShare of visits or clicks that convert., revenue, cost per acquisition and task-completion rates on the pages that matter. Care work is judged by what happens after the first purchase: repeat-purchase rate, retention over time, service-contact outcomes, recommendation and the resulting LTVTotal value a customer brings over time..
The source article proposes its own metric set per cluster, and the specific choices matter less than the discipline: one or two metrics per cluster that the cluster can honestly move, each written as a KPIA metric tied directly to an objective. with a target and an owner. Where analytics cannot yet separate clusters, use the proxies you have (query type, landing-page group, content format) and say so in the report, rather than reporting everything against sales.
Worked example (illustrative)
A furniture maker rebalances its plan
A mid-sized furniture maker selling online audits a year of spend against the four clusters. Of a £420,000 budget, £361,000 (86%) went to Do activity: product search ads, shopping campaigns and retargeting. Think received £44,000, mostly comparison-term search. See received £15,000 in a single social campaign, and Care received nothing beyond transactional email. Meanwhile, search demand analysis showed that queries for the category ("solid oak dining table", "how to choose a sofa") outnumbered brand and product queries by roughly nine to one. The team rebalanced for the next year: Do spend was held at £340,000, Think rose to £75,000 to fund a buying-guide series and a room-planner tool, See rose to £60,000 for a video series on caring for wooden furniture, and Care received £20,000 for an onboarding sequence and a repeat-customer service line. Each cluster was given its own target: See was judged on qualified video reach and returning visitors, Think on guide completions and planner sessions, Do on conversion rate and cost per order, and Care on twelve-month repeat purchase. Nobody expected the video series to show purchases in the first quarter, and the reporting was built so that it would not be asked to.
Limits
The framework is a snapshot of an audience, not a model of movement. It says nothing about how long people stay in Think, what moves them to Do, or how many never move at all. For sequencing questions the marketing funnel framework is the better tool, and the two are commonly used together: See-Think-Do-Care to decide who to serve and with what, the funnel to measure flow.
"Largest addressable qualified audience" is also easy to inflate. A team that wants a bigger See budget can define qualification loosely; a finance director who does not can define it tightly. The framework offers no rule for drawing the line, so the definition has to be written down and agreed before budgets are set.
Care is the least developed cluster. The framework tells you to serve existing customers differently and to measure them by retention and recommendation, but it does not describe retention mechanics, lifecycle programmes or service design. It is a prompt to take Care seriously, not a method for doing so. Finally, the measurement discipline assumes analytics that can tell clusters apart. Where tracking cannot distinguish a Think visit from a Do visit, the metrics degrade into proxies, and it is honest to say so.
When awareness is the objective, See-Think-Do-Care for awareness campaigns goes deep on sizing the See cluster, protecting its budget and measuring it without last-click credit.
For business buying, See-Think-Do-Care for lead generation shows how clusters are read per account, with Do as a request for a sales conversation and Care as expansion.
Common mistakes
- Reading the clusters as a sequence. People do not file neatly from See to Do; the framework describes intent at a moment, not a path.
- Measuring every cluster by sales. The single error the framework was built to prevent, and still the most common one.
- Defining See as everyone. Without the "qualified" test, See content becomes general entertainment that earns attention from people who will never buy.
- Putting subscribers in Care. Care is repeat customers. An email list is mostly See and Think.
- Spending on Do alone and calling the rest a brand problem. The audience in Think and See is where the next year's Do demand is formed.
- Using Think content to close. A comparison page that hides where competitors win is read as a pitch, and loses the trust it was meant to earn.
