Why a balanced four-cluster plan underweights awareness
The framework's usual use is diagnostic: map spend across the four clusters, notice that Do takes most of it, and rebalance. That exercise treats See as one of four. When awareness is the objective, See is the plan, and the questions shift from balance to depth: exactly who is in the cluster, what earns their time, how much money they should command, and how to stop the rest of the marketing machine from treating them as prospects. Each of those is a decision the general framework leaves open, and each is easy to get wrong in an organisation whose reporting was built around sales. The brand versus performance guide explains why that reporting pulls budgets towards Do, and the choices below are the counterweight.
Decisions that change for awareness
| Cluster question | In a balanced plan | When awareness is the objective | Decision it forces |
|---|---|---|---|
| Who is in See | A qualified audience, briefly defined | The whole category audience in the trading area, including distant future buyers | Where the edge of the cluster sits, written down |
| See content | Useful or entertaining pieces, one stream among four | The main output, judged on attention and recall | Which formats and topics earn time from non-shoppers |
| Budget | Shifted from Do towards See | Set from reach goals first, then defended against Do reporting | How much See spend is fixed before results arrive |
| Measurement | One or two metrics per cluster | Reach, brand search and lift for See; last-click excluded | Which numbers the board sees for See |
| Audience hygiene | Not usually addressed | Rules that stop See viewers being retargeted as Do | What follow-up a See viewer receives |
Define See as the full category audience
The edge of See should follow the category, not the funnel. For a bakery, that is everyone in the trading area who buys bread, cakes or coffee anywhere, including households that have never heard of the brand and students who will not buy a celebration cake for years. Two pressures push the line inwards: platforms reward narrow, high-intent targeting, and finance prefers audiences that look close to purchase. Resist both. Size the cluster from category data such as population, household counts or search volume for generic category terms, write the definition down, and keep it fixed across the year. The test of a good definition is that it includes people who would find the brand useful later but excludes people with no plausible connection to the category at all.
Choose See content for attention and recall
See content succeeds when someone with no plan to buy is glad to have seen it and remembers whose it was. That favours formats people choose to consume, such as short video, recipes, local stories and behind-the-scenes pieces, over formats that only make sense to shoppers. The social content formats guide helps match format to that goal, and earned coverage built on a genuine story, as covered in digital PR campaign ideas and data stories, reaches See audiences that paid media struggles to hold. The trade-off is commercial pressure: product-led pieces are easier to justify internally but earn less attention from people who are not shopping.
Set the See budget before Do reports can shrink it
Do metrics arrive daily and look decisive; See results take weeks and look soft. Left alone, budgets drift to whatever reports fastest. The defence is to set See spend from a reach goal, the share of the category audience to reach at a planned FrequencyAverage times a person saw an ad., and to agree in advance that it will not be cut on the basis of last-click results. The guide to media objectives and reach goals shows how to turn a business goal into a reach target. Choose a protected See budget when the brand is small in its category or entering a new area; a well-known brand in a stable market can afford to flex it more.
Measure See with reach, brand search and lift
Last-click attribution gives See content almost no credit, because its audience rarely buys in the same session. Replace it with three measures: ReachUnique people who saw an ad. within the defined category audience, branded search volume against a baseline, and periodic lift in awareness or recall. The brand awareness and brand lift guide covers surveys, lift studies and branded search in detail. Report these to the same meeting that sees sales, so that See is not judged in a separate, less important forum.
Keep See audiences out of Do retargeting
Once a See video runs, its viewers appear in remarketing lists, and the temptation is to follow them with offers. That treats curiosity as intent. Write a rule that See viewers receive more See or Think content, not Do messages, unless they show a stronger signal such as visiting a product or menu page. The YouTube audience targeting guide explains how viewer lists are built, which makes it straightforward to exclude them from purchase-focused campaigns.
Decision points in a worked See-Think-Do-Care for awareness
An invented regional bakery chain, Crumb and Kiln, sizes its See cluster
Crumb and Kiln runs 14 bakeries across one English county. Last year it spent £180,000 on marketing: £151,000 on Do placements (local search ads, delivery-app promotions and retargeting), £22,000 on Think, and £7,000 on See. Growth stalled in towns where it already had a shop, and new shops opened slowly.
| Decision | Before | After |
|---|---|---|
| See definition | "People who have visited our site" (about 40,000) | Adults within 20 minutes of a shop who buy bakery goods anywhere (about 410,000) |
| See content | Occasional product photos | Weekly short videos of bakers at dawn, a seasonal bread series, a local-history PR story on the oldest shop |
| Budget | £7,000 See | £55,000 See, fixed for the year; Do cut to £105,000; Think £20,000 |
| Measurement | Delivery-app orders by last click | Reach of 45% of the See audience at 3 or more exposures a month; branded search; twice-yearly awareness survey |
| Audience rule | Video viewers retargeted with 10% off codes | Viewers receive the next video; only menu-page visitors enter Do retargeting |
After two quarters, prompted awareness in the survey rose from 31% to 39% of the defined audience, branded search rose 28% against the same months a year earlier, and the newest shop reached its sales plan two months sooner than the previous opening. Delivery-app orders by last click fell 6%, which the team had agreed in advance not to treat as a failure.
The decisions protected each other. Without the wider definition the See budget would have targeted people who already knew the brand. Without a fixed budget, the early fall in last-click orders would have reversed the shift within a month. Without the retargeting rule, the new viewers would have been met with discount codes that taught them to wait for offers.
Measuring the See cluster: KPIs and cadence
Read reach and frequency against the category audience weekly during active flights, since they show whether the plan is delivering. Branded search is best read monthly against the same period a year earlier to remove seasonality. Awareness and recall surveys, or platform lift studies, belong at the end of each flight or twice a year. Share of VoiceYour share of market visibility. against local competitors adds context to all three. The reach and frequency calculator and impression calculator help translate the reach goal into the impressions and spend the plan needs. Last-click sales can stay in the report for completeness but should carry no weight in judging See.
Where the See cluster misleads
A wide definition can become an excuse: if everyone is in See, any reach counts as progress. Keep the category test strict enough that reach to irrelevant audiences is visible as waste. Branded search can rise for reasons unrelated to the campaign, such as a news story or a new shop opening, so read it alongside lift data. And because See works slowly, a poor plan and a good one look alike for the first few weeks.
Common mistakes
- Sizing See from existing visitors. It shrinks the cluster to people the brand already reaches.
- Product adverts labelled as See content. Non-shoppers give them no attention.
- Cutting See on last-click results. The measure was never designed to see it.
- Separate meetings for brand results. See loses every argument held in a room it is absent from.
- Discount codes for See viewers. Curiosity is treated as intent, and goodwill is spent.
