Skip to content

mbfinotti/advertising-skills / skills/advertising-hiring

advertising-hiring

Verified

Build the complete hiring packet for a paid-advertising role - the rung decision (media buyer vs performance marketer vs growth marketer; creative strategist is its own role), an outcome scorecard, a 2-3 round interview loop with a judgment-first question bank, a time-boxed in-interview work sample, and a 30-60-90 ramp with staged spend authority. For the marketing leader or founder recruiting. Use for "hire a media buyer", "interview questions for a performance marketer", "paid media work sample", "scorecard for our first ad hire", or "30-60-90 for a new paid media hire". Does not source candidates, post jobs, run applicant tracking, or write offers, and does not cover the candidate side - that is the sibling mbfinotti/advertising-skills@advertising-career.

references/question-bank-and-scorecard.md

View on GitHub
# Question Bank and Scorecard

## Loop shape

Real loops in this field are short. Duane Brown (Take Some Risk), from the PPCChat "Hiring & Getting Hired in PPC" roundup (hosted by Julie Bacchini, April 19, 2022): "We redid our whole process in the summer 2020. Went from 3-4 interviews down to 2 interviews. Standardized all questions AND ASKED BETTER QUESTIONS."

Governance patterns from the same roundup:

- "I have a team that does the screening, and I get the final say" (Jon Kagan).
- "At all levels, we create a hiring committee (including seniors)" (Pauline Jakober, Group Twenty Seven).

Navah Hopkins (Search Engine Journal, September 4, 2025) frames evaluation as three buckets rather than named stages: paid search, paid social, and general marketing plus culture fit. Practical shape: 2-3 stages - screen, time-boxed technical/case, peer or founder round.

## Reasoning over recall

The core scoring principle: reasoning over recall.

- Navah Hopkins: "You might be asked, 'Why would you structure a campaign/account this way?' Never cite 'best practices' or default methods as your rationale. Interviewers want reasoning rooted in context, goals, and a test-and-learn approach."
- Kasim Aslam (Solutions 8) builds the trap into the question itself: "What's your favourite bid strategy? It's a trick question, it depends on the situation."
- Marc Barraza frames what to reward: "I care less about someone knowing the formula for CTR and more about how they think about the relationship between CTR and other metrics."

The strong answer is "it depends, here is my reasoning"; the weak answer is best-practice recitation.

## Spending the interview minutes

Order the bank by judgment revealed per minute asked: `two senior filters > exaggerator probes > killed-campaign and incrementality probes > named situational questions > commitment screen`. The senior filters lead because they decide the hire, not because they are cheap - they are the most expensive questions here to ask and to grade. The axes disagree:

- value (separates a senior operator from a merely fluent one): `two senior filters > killed-campaign and incrementality probes > named situational questions > exaggerator probes > commitment screen`
- effort (minutes asked, plus the interviewer expertise needed to grade the answer): `two senior filters > killed-campaign and incrementality probes > named situational questions > commitment screen > exaggerator probes`

A default, not a law - re-rank it against the pool and the panel you actually have:

- An unvetted inbound pool promotes the exaggerator probes into the screen; a referral, or a bench left over from a previous search, demotes them.
- A panel with nobody able to grade a ROAS-drop answer moves the senior filters into the work sample rather than the interview.
- A seat that will not touch measurement for six months drops the MER filter below the situationals.

Structure per competency: 2-3 behavioral questions, 1-2 situational, plus follow-up probes ("What was the outcome?", "What was your specific role?", "What would you do differently?"). Add candidate-specific questions generated from each resume's gaps.

## 1. The two senior filters

**Diagnosing a ROAS drop.** Expect the fixed diagnostic order:

1. Measurement: pixel/server-event deduplication, platform-versus-backend revenue gap.
2. Media: CPM, frequency, audience saturation.
3. Creative fatigue: hook rate, hold rate, CTR by unit.
4. Post-click: landing page, conversion rate.

The discriminator (Bigeye Agency): "Low ROAS with flat backend sales points to true underperformance. Low ROAS with stable backend sales points to tracking or attribution gaps." Jordan Glickman assigns the audit across roles with a "within 24 hours of a ROAS alert" completion target - useful for probing how a senior candidate coordinates, not just diagnoses.

**MER versus ROAS.** The strong answer translates platform ROAS into business reality:

- Break-even ROAS: the per-channel minimum.
- MER (total revenue over total marketing spend): the business-level truth.
- Acquisition MER (new-customer revenue over total spend): the senior nuance.

The weak answer cites platform-reported ROAS uncritically, missing that platform ROAS is structurally biased upward because every platform claims credit for the same conversions.

## 2. Exaggerator-catching probes

A minute each, gradeable by anyone on the panel, and they eliminate a whole failure class - which is why they sit this high despite being the weakest signal on the list about a genuine candidate.

- Andrew McGarry (beyondcontent): "Killer question is always 'who do you admire or look up to in PPC and why?' It tends to reveal who is exaggerating their experience."
- Melissa Mackey (beyondthepaid): "I always ask 'What industry publications do you read?' in every interview at every level."
- Jon Kagan: "If they tell me they understand [the newest automated campaign types], I know they are lying." Claimed mastery of things nobody fully understands is itself the signal.

## 3. Killed-campaign and incrementality probes

The pairing that makes a spend-under-management screen mean something (Constant Hire): walk me through a campaign you killed and why, and how you separate incremental revenue from attribution-modeling noise. Never ask the spend question without these.

## 4. Named situational questions

Each is the named practitioner's stated go-to (PPC Live UK collection and the PPCChat roundup):

- Fraser Andrews (HDY Agency): "What is your current understanding of attribution and how it has changed over the last 3 years?" and "How would you go about structuring an account?"
- Menachem Ani (JXT Group): "How would you improve lead/call quality on a lead-gen campaign?"
- Glenn Schmelzle: "What's your approach to analyzing multi-touch attribution as a means of evaluating your media budget allocation across 2 or more channels?"
- Akvile DeFazio (AKvertise): "What metrics do you track on a regular basis and how/when do you inform your team of results and next steps?"
- Navah Hopkins, behavioral-strategic: "How would you prove that your campaign delivered results if the client questions the attribution?"

## 5. Commitment screen

Kasim Aslam: "Flash me forward 5 years, and the world is perfect. Where are you?... there are a lot of 'tourists'." Last on the order because it predicts tenure, not performance - cut it first when the loop is short.

## Scoring rules and rubric weighting

The published rubric to anchor against is vendor content. Pare (George Reid, October 22, 2025) uses a 1-5 scale ("1-2 superficial or incorrect; 3 core concept understood; 4-5 deep, structured, commercially aware") weighted Technical 40 / Communication 30 / Commercial insight 30, in a 45-60 minute structured interview plus a 15-minute sample-account audit. Label it as one vendor's internal process, not surveyed employer practice, and note that it puts 60% of the weight on non-technical judgment.

Rules that hold regardless of scale chosen:

- Write a role-specific behavioral description for every anchor point; a bare number or a generic label collapses into gut-feel scoring.
- Every score cites evidence - a quote or a concrete example. No evidence, no score.
- Interviewers submit scores before the debrief; debrief time goes where scores diverge; disqualifying signals are recorded separately from the score.
- Weight commercial reasoning and communication over technical recall.
- Any numeric weighting emitted is a house convention, not a published rubric; say so in the packet.

## Certifications: floor, not signal

Platform certifications - Google Ads (Skillshop), Meta Blueprint, and peers - clear applicant-tracking screens and roll up into agency partner-program status; they do not predict competence. Certifications expire and need annual renewal.

The Google Partners requirements explain the contradiction between job listings demanding certifications and practitioners dismissing them - an agency operational requirement, not a quality bar:

- Keep at least 50% of designated account strategists certified in Google Ads, with at least one certification per product area.
- Spend $10,000 across managed accounts in 90 days.
- Hold a 70% optimization score.

Justin Brooke (AdSkills) names the incentive problem: "Google teaches you how to use Google. Meta teaches you how to use Meta... this teaches you how Meta wants you to use the platform. And Meta's interests don't always align with yours as an advertiser."

For the hirer:

- Never weight certifications in competence scoring.
- Never reject a strong candidate over a missing one - the free ones can be earned after hire in days.
- If partner compliance genuinely requires certification, state that reason explicitly in the req so it is not mistaken for a skill bar.
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 323kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 323kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 323kUpdated Claude CodeCodexVerifiedInstall