Keyword and interest targeting
Query-level targeting built first, with interest and audience layers added for scale rather than used as the whole plan.

Visual search buying
People arrive with a query and a plan, not with a scroll habit. That single difference changes targeting, creative, and how far ahead a campaign has to be scheduled.
The most useful reframe for Pinterest is to stop calling it social. Users open it to look for something, type a query, and save results for later. Behaviorally that is much closer to search than to a feed where content is pushed at a passive audience.
Because saving is the primary action, content has an unusually long shelf life. A pin that performs can keep generating impressions and clicks months after the campaign that introduced it, which is not how paid social normally behaves. That changes the calculus on creative production, since good assets keep working.
The second consequence is planning lead time. Users plan ahead here, often by weeks or seasons. A campaign launched in the week of an event has already missed the research window. We build seasonal calendars that start well before the moment being sold into.
None of that makes Pinterest right for everyone. It earns budget in categories where the purchase is visual and planned. In categories that are neither, the same money usually works harder elsewhere, and we say so.
What it is
Keyword thinking, catalog structure, and a calendar that starts early.
Targeting starts with keywords, not just interests. Because users type queries, keyword targeting captures stated intent in a way interest targeting does not. Interest and audience layers still matter for scale, but a program built only on interests leaves the most qualified traffic on the table.
Creative is vertical, text-legible, and built to be understood without sound or motion. The strongest performers usually communicate the outcome immediately: the finished room, the completed project, the styled product in context. Assets that require a caption to make sense underperform, because the pin is often seen at thumbnail size inside a dense grid.
Shopping and catalog feeds turn a product range into ongoing inventory rather than a set of handmade campaigns. A clean feed with accurate titles, pricing, and availability lets the platform surface individual products against relevant queries, which is where most of the durable performance in retail accounts comes from.
Seasonal planning is a discipline rather than a preference. Research behavior runs ahead of purchase behavior, so campaigns are scheduled backward from the moment of demand. That also means measurement windows need to be longer than they would be on an impulse-driven platform, or a campaign gets judged before its saved pins have paid out.
Fit
We would rather say no early than sell a program that cannot work.
Deliverables
Search discipline applied to a visual platform.
Query-level targeting built first, with interest and audience layers added for scale rather than used as the whole plan.
Assets built to read at thumbnail size and show a finished outcome, since the pin has to make sense before anyone reads a word.
Product feed structure, titles, pricing, and availability maintained so individual products can surface against real queries.
Campaigns scheduled backward from demand, because research here runs weeks ahead of purchase and late launches miss the window.
Ongoing performance of saved assets monitored, so evergreen winners get supported instead of retired with their campaign.
Measurement windows matched to a longer consideration path, with a clear read on whether the category justifies continued spend.
How we run it
Qualify the category, then build for shelf life.
We assess query volume and visual fit honestly, and recommend against the channel when the category does not support it.
Query research plus catalog cleanup happen together, since both determine what can actually be surfaced.
A first set of vertical assets built around outcomes, with enough variation to identify what the audience responds to.
Campaigns go live ahead of the demand window, with budget weighted toward the research period rather than the event itself.
Assets that keep earning saves get maintained and re-promoted, since shelf life here is an asset most platforms do not offer.
Functionally it behaves like a visual search engine. Users arrive with a query, type it, and save results for later, which makes keyword targeting more important than the interest-based approach used on feed-driven social platforms.
Saving also gives content a long shelf life, so a strong pin can keep producing traffic months after its campaign ends.
Ahead of the demand window, usually by weeks and sometimes by a full season. Users research and plan before they buy, so a campaign that launches in the week of an event has already missed the audience it wanted.
Where this connects
Visual search rewards the same thinking as text search.
Query targeting here draws on the same demand research as keyword research since what people type to find a product does not change much between platforms.
Product feed accuracy is the same discipline that supports Google Ads where shopping performance also lives or dies on catalog hygiene.
The organic side, where indexed pins keep earning clicks for years, is covered under Pinterest management and organic saves are the clearest indicator of which creative deserves budget.
Assets and visual standards across channels come out of graphic design so a pin looks like the brand and not like a one-off export.
Whether a long-window channel deserves budget is decided in campaign management where measurement windows are set before results get judged.
Questions
Query targeting, clean feeds, and a calendar that starts before the season does.