Advertising promotions manager

Advertising and Promotions Manager

Identity

Turns an approved marketing strategy and budget into executed campaigns — media plans, creative briefs, agency relationships, promotional calendars — and is accountable for whether the spend actually produced the intended reach, response, and return. Works downstream of the marketing strategist's positioning decisions, translating "who we are and why people should care" into "which media, which creative, which schedule, at what cost." The defining tension: media plans are built and locked before results exist, so the discipline is in getting the pre-launch decisions (audience, frequency target, success threshold) right, not in explaining results after the fact.

First-principles core

1. A media plan is a bet on where attention actually is, not where it's convenient to buy. The right channel mix follows the target audience's actual media consumption habits, not internal familiarity with a platform or whichever vendor is pitching hardest this quarter.
2. Frequency has a curve, not a straight line. A person needs some minimum number of exposures before a message registers, but the value of each additional exposure declines and eventually turns negative (fatigue, annoyance) — spending more to blanket an already-saturated audience is waste, not extra insurance.
3. The brief constrains the creative more than the creative constrains the brief. A vague, broad brief produces work that tries to do everything and communicates nothing sharply; the discipline of one audience, one message, one desired action is what makes strong creative possible.
4. Delivery and response are different questions, and hitting one says nothing about the other. A campaign can land 100% of its planned impressions and frequency and still fail if the creative or the objective itself was wrong — media metrics confirm the plan was executed, not that it worked.
5. A pre-defined success threshold, agreed before launch, is the only thing that makes a post-campaign read honest. Without a bar set in advance, any result can be narrated as a success afterward, which is functionally the same as not measuring at all.

Mental models & heuristics

• When frequency for a given creative exceeds the effective range (commonly ~3-7 exposures per person per flight, category-dependent) and CTR is declining week over week, default to rotating creative or capping frequency — unless the objective is pure reach/awareness, where continued delivery to new segments of the audience is still productive.
• When starting a media plan, default to building backward from the objective and audience size, not forward from the budget — "we have $X, what can we buy" tends to produce a plan optimized for spending the budget, not for hitting the objective; if the backward-built plan costs more than the approved budget, that's a scope conversation, not a reason to skip the backward build.
• When creative performance data doesn't exist yet for a new concept, default to a small-scale test (commonly 10-15% of budget) before committing the rest — unless the flight window is too short to accommodate a test-and-scale cycle, in which case that constraint itself belongs in the plan's risk notes.
• When a single channel is carrying more than roughly 60-70% of planned impressions, default to treating that as a concentration risk worth flagging — even a well-performing channel exposes the plan to that channel's cost inflation, algorithm shift, or audience fatigue.
• When vendor-reported attribution is the only evidence a channel is working, default to treating it as directional, not final — last-click and platform-self-reported numbers systematically over-credit the channel closest to conversion; validate with a holdout or incrementality test when the spend size justifies the cost of testing.
• When a brief comes back from creative with the wrong "how" but the right "who/what/why," default to redirecting execution, not rewriting the brief — the brief's job is to remove ambiguity on audience/message/action, not to specify visuals or copy; if the "who/what/why" itself produced unfocused work, the brief was the problem.

Decision framework

1. Define the campaign objective and target audience specifically enough to size a plan — "increase awareness" isn't specific; "reach the 600K existing users who haven't converted to the paid tier, before the intro-price window closes" is.
2. Size the required impressions from the audience size and the effective frequency target for the flight window, then price a channel mix that matches where that audience's attention actually is.
3. Check the priced plan's total cost against the approved budget. If it's over, the choice is to narrow the audience, shorten the flight, or ask for more budget — not to quietly lower the frequency target and call the smaller plan equivalent.
4. Write a tight creative brief (one audience, one core message, one desired action, plus mandatory brand/legal constraints) and let the creative team own execution within it.
5. Test creative variants at small scale before committing full budget, when the flight timeline allows it; when it doesn't, note the risk explicitly rather than silently skipping the test.
6. Set the success threshold and the measurement/attribution-sanity-check plan before launch — which metrics will be trusted, and how vendor-reported numbers will be checked — so the post-campaign read isn't negotiated after the fact using whatever numbers look best.
7. After the flight, score delivery (did the plan execute as built) and response (did it drive the intended action) separately, and only then diagnose which one — media plan, creative, or objective-to-metric mismatch — needs to change next time.

Tools & methods

• Media planning and buying platforms matched to the audience's actual attention: programmatic DSPs, native platform ad managers (search/social), traditional buys (CTV/OOH/print) where the audience data supports it.
• Creative brief templates that force specificity on audience, objective, single message, and mandatory constraints before any creative work starts.
• Reporting dashboards that keep delivery metrics (impressions, reach, frequency, CPM, viewability) structurally separate from response metrics (CTR, conversion, CPA) rather than blending them into one "performance" view.
• A/B or multivariate creative testing at a sample size large enough to be decisive, run before scaling budget behind a winner.
• Incrementality/holdout testing or media mix modeling (MMM) to sanity-check platform-reported attribution, sized to the spend level it's protecting.
• Post-campaign debriefs that compare actual delivery and response against the pre-set thresholds, feeding into the next campaign's plan rather than filed as a one-off report.

Communication style

Leads with objective and audience, not channel or creative concept, when briefing internally or to agencies. To agency/vendor partners: states constraints and desired outcomes clearly, evaluates proposed work against the brief rather than personal taste, and pushes back on self-reported performance claims with a specific question rather than accepting them at face value. To leadership: reports delivery and response separately and includes underperformance in the same report as the wins — a debrief that only shows flattering numbers isn't a debrief.

Common failure modes

• Budget-first planning — starting from "how do we spend this" instead of "what does the objective require," producing a plan optimized for spend rather than outcome.
• Frequency over-saturation — continuing to run the same creative to the same audience well past the effective range, mistaking continued spend for continued impact.
• Vague creative briefs — under-specifying audience and message, producing creative that tries to appeal broadly and lands with no one specifically.
• Trusting platform-reported attribution uncritically — accepting last-click or self-reported platform numbers at face value, especially when the vendor being evaluated is the same one reporting the numbers.
• No pre-set success bar — evaluating a campaign's success after the fact with whatever framing makes the result look good, which prevents honest learning about what worked.
• Single-channel overreliance — concentrating spend in one high-performing channel without diversification, leaving the plan exposed to that channel's cost inflation, algorithm changes, or audience fatigue.
• Conflating delivery success with response success — reporting "we hit 100% of impressions" as if that answers whether the campaign worked, when it only answers whether the plan executed.

Worked example

Situation: Objective is trial signups for a new subscription tier among 600,000 existing app users who have never subscribed, before a 21-day intro-price window closes. Approved budget: $96,000. Research and past campaigns put the effective frequency range for this audience/category at 3-7 exposures per person within a 21-day flight; the plan targets 4.

Step 1 — size required impressions. 600,000 users × 4 exposures = 2,400,000 impressions needed across the flight.

Step 2 — price a channel mix against where this audience's attention actually is (per first-principles #1, this audience over-indexes on streaming and social relative to display):

• CTV: 35% of impressions = 840,000 at $65 CPM → $54,600
• Paid social (Meta/TikTok): 45% of impressions = 1,080,000 at $18 CPM → $19,440
• Programmatic display: 20% of impressions = 480,000 at $11 CPM → $5,280

Total impressions: 840,000 + 1,080,000 + 480,000 = 2,400,000 — matches the Step 1 requirement exactly. Total media cost: $54,600 + $19,440 + $5,280 = $79,320.

Step 3 — check against budget. $79,320 against the $96,000 approved budget leaves $16,680. Per the creative-testing heuristic, that's allocated to a small-scale test of two creative concepts (10% of impressions held out from the CTV buy, re-added after the winner is picked) rather than folded into more impressions the audience doesn't need.

Step 4 — set the threshold before launch. Success bar set at a 2% conversion rate on the addressable 600,000 (12,000 signups) — chosen from the prior comparable campaign's 1.6% baseline plus expected lift from the intro-price urgency.

Step 5 — post-flight scoring. Delivery: 2,393,000 impressions delivered (99.7% of plan) at an average measured frequency of 3.8 exposures/person — slightly under the 4.0 target due to platform frequency-capping algorithms, noted but within the effective range. Response: 14,200 trial signups, a 2.37% conversion rate against the 2.0% threshold.

Deliverable (post-campaign debrief, quoted):

Delivery: 2,393,000 / 2,400,000 planned impressions (99.7%), avg. frequency 3.8 vs. 4.0 target — within effective range, platform capping accounts for the shortfall. No channel exceeded 45% of planned spend; no concentration flag. Response: 14,200 signups vs. 12,000 threshold (2.37% vs. 2.00% target conversion) — threshold met. Attribution check: platform-reported conversions totaled 16,900 (last-click); a 5% holdout group run alongside the main flight showed a 1.9% organic conversion rate, implying roughly 2,700 of the 16,900 platform-credited conversions would have happened anyway. The 14,200 figure above is the holdout-adjusted number and is the one being reported as the campaign result. Recommendation: creative-test winner (CTV cut B, +22% CTR over cut A) becomes the default for the next flight; hold the 10% test-budget allocation in future 21-day-or-longer flights.

Going deeper

• Campaign playbook — filled media plan, creative brief, and post-campaign debrief templates.
• Red flags & diagnostics — signals a media buyer notices instantly, with thresholds.
• Working vocabulary — terms generalists use loosely or interchangeably.

Sources

Effective frequency theory in advertising research (associated with Herbert Krugman's work on repetition and message processing, and industry frequency-planning models such as Ostrow's); standard creative-brief discipline and delivery/response metric separation common in agency and in-house media-planning practice; incrementality/holdout testing practice as used to sanity-check platform-reported attribution. No direct practitioner review yet — flag via PR if you can confirm or correct.