Ad Manager

• You see the supply. You are selecting specific publishers and placements, not an interest proxy applied inside a network you cannot inspect. The publisher, the site, the format and the audience are all on the listing before you commit.
• The rate card is visible up front. Cost per thousand impressions, minimum spend, volume discount tiers, impressions available per day, and how much of the placement is already booked — all readable before you add anything to a cart.
• Supply is first-hand. Publishers are on En Garde directly: they run their own space, place their own code on their site, and see their own audience cohorts and revenue. The audience signal behind a listing comes from that relationship rather than from a third party modelling it at arm's length.
• Selection criteria are cultural, not just contextual. Cultural relevance is a sortable, filterable field sitting next to CPM, reach, click-through and return, so you can hold a cultural bar and a performance bar at the same time.
• It is a direct buy, not an auction. You purchase a flight at a stated rate for stated dates. That means no bid strategy to manage — and equally, no in-flight bid or pacing levers in the interface. Plan the flight accordingly.

• An En Garde Score out of 100 — cultural relevance of the placement, broken into its dimensions rather than left as one opaque number.
• Audience — monthly reach, age bands, gender split, and the cultural affinities the audience actually skews toward.
• Content categories — the standard category set, so brand safety exclusions are straightforward.
• Pricing — CPM, minimum spend, volume discounts, and an estimated monthly cost on the detail view.
• Delivered history — that placement's own average click-through, conversions and return on ad spend. It is history for the placement, not a forecast for your client.
• Availability — daily impressions open and the share already booked, so a large reach figure does not get mistaken for buyable volume this month.

Packages bundle placements into a single line item with a combined score, combined reach, a package rate and the saving against buying the components separately — useful when you are briefing a multicultural awareness or launch push rather than hand-picking sites.

1. Browse inventory. Sort by highest score, lowest or highest CPM, largest reach, best click-through, best return, or newest. Search by publisher or site, and filter on score range, CPM range, minimum reach, minimum click-through, minimum return, and on individual score dimensions when a client cares about one in particular.
2. Open a placement. The detail view carries the full score breakdown, affinity labels, audience and delivered performance, and the pricing including minimum spend. Add to cart from here.
3. Or work from packages. Filter by use case, minimum score, minimum reach, maximum budget, or featured only, and add a bundle in one action.
4. Review the cart. The cart slides out from the header on any screen with line items, removals and a running total. Placement detail views and the package gallery are the two routes that feed it today.
5. Check out in three steps. Review items, then set campaign name, flight start and end dates and any notes, then pay by card. The order summary stays alongside the whole way.
6. Track it in Orders. Order number, campaign, date, order status, payment status and total, filterable by status. Orders are scoped to the Brand Space you bought them in, so client work stays separated.

The point of buying culturally adjacent inventory is reaching people your client's list does not already contain. Two mechanisms support that, and they are at different stages.

Look-alike modelling, built but not self-serve. The engine takes a seed audience — a saved segment or customer list from Audience — and reduces each member to a signal profile spanning demographics, behaviour such as recency, frequency and value, engagement pattern, and cultural signals. It builds the centre of gravity of that seed, scores every other known individual against it, and returns ranked candidates each with its own similarity score and predicted value. You set the minimum similarity to accept and a ceiling on audience size, which is the trade-off between precision and scale you would expect. Today this runs as work we do with you: there is no button on a segment card and no automatic hand-off into placement selection. The expansion opportunities panel inside audience analytics is illustrative sample content, not a model output — do not plan spend from it.

Cultural adjacency on inventory, available now. Placements carry explicit cultural affinity labels and dimension-level scores. Once you know which cultural communities a seed segment leans into, you can filter directly to the placements that reach them, including audiences sitting alongside your client's existing base rather than inside it. This is the practical route to extending reach today, and it does not depend on the modelling above.

Be clear-eyed about this part: it is a managed hand-off, not instant self-serve trafficking. When payment clears the order moves into processing and the En Garde team sets the campaign up across the placements you bought. The confirmation screen states the same expectation everyone gets — setup typically 12 to 24 hours, serving typically within 24 to 48 hours, with email at each stage.

The order record holds the commercial facts: items, totals, flight dates, order status and payment status. It does not carry live delivery counts. Once a campaign is running, performance reports in Analytics alongside the client's other channels, including the cultural view of performance, so a direct buy here and a platform campaign are read side by side rather than in separate tools. Changes and cancellations go through the team rather than a control on the screen.

• Payment is by card at checkout. There is no invoicing, purchase order or credit line in the flow today, which matters if your finance process needs one.
• Minimum spend is set per placement. Thin budgets spread across many placements will fail on minimums before they fail on reach.
• The Orders list is the order of record. A per-order detail screen and an in-app cancel action are not built yet.
• A placement's score judges the placement. Your creative is scored separately, so a strong buy paired with weak creative shows up as two distinct problems rather than one.
• Everything is scoped to the active Brand Space, so inventory, carts and orders follow whichever client brand you are working in.