Issue #6 · May 29, 2026

Analytics Leadership Has a Build Brief and a Scale Brief. They're Not the Same Interview.

Career Snapshot

The analytics leadership pool is about to get bigger. Meta’s notice periods clear in mid-June, and most of that clearance lands at the Director and Head of Analytics tier. The demand side has not adjusted to match. That is the thing nobody is saying.

Roblox is the social-ad-platform story this week. They are hiring measurement analytics and BI leadership roles for an advertising marketplace that does not yet have a measurement framework. No historical ROAS benchmarks. No attribution methodology for immersive ad formats. A user base that skews under twenty-five. Roblox is not looking for somebody who knows how advertising measurement is supposed to work. It is looking for somebody who can decide what it should mean on this platform.

LinkedIn has active Director and Head of Analytics reqs in the advertising division. Comp at the Director tier: $220K–$290K base. TikTok US has analytics leadership reqs live in measurement and advertising, and the search is active regardless of the acquisition noise.

Hot this week:

Director / Head of Analytics. Build briefs and scale briefs are both live. They are not the same search. Details in Skill Market Shifts.
Measurement Analytics Leadership at social ad platforms. Roblox (zero-to-one build). LinkedIn (govern at scale). Different problems.
BI / Analytics Leadership at AI-revenue SaaS. The companies that declared AI revenue in Q1 are hiring the analytics leaders who will report whether it held in Q2. Director comp: $200K–$270K base.

Reading the tape: more candidates, more specific briefs. The supply side got bigger. The screen did not get easier.

Skill Market Shifts

1. BI Director JDs now include an AI/BI governance philosophy screen — not a tooling checklist.

The shift: For most of the last decade, analytics leadership interviews screened for tool fluency and team structure. H1 2026 JDs at companies running Snowflake Cortex Analyst or Databricks AI/BI in production have added a new filter: governance philosophy. How do you make sure an AI-generated analytics output is trustworthy before a VP acts on it. The JDs say “AI-assisted analytics governance” and “output validation framework.” Those are not tooling questions. They are design questions, and the difference is the whole point.

The evidence: Snowflake Summit (June 2–5) and Databricks Summit (June 9–13) will keynote governance features for their AI/BI products. The JDs at companies already running these tools reflect the shift. The Summit confirms the direction. It did not set it. The JDs did.

The career translation: “We use dbt tests and data contracts” is the Staff AE answer to this question. The BI Director answer names a governance philosophy at the output layer — a specific failure mode the AI tier creates that automated testing misses, an escalation path, and an acceptance criterion before a number lands in a VP dashboard. The candidate who has this written down before June 2 is ahead. The Summit audience will hear it as news.

2. Build briefs and scale briefs now signal which one they are in the JD. Most candidates give the same pitch for both.

The shift: “Stand up,” “define from scratch,” and “build the analytics function” are build signals. “Lead an established team,” “drive organizational maturity,” and “modernize existing infrastructure” are scale signals. Both types of Director role have always existed. What changed in H1 2026: the JDs are explicit about which one they are running. The signal is in the language. Most candidates do not read that far.

The evidence: Director JDs at AI-revenue companies in H1 2026 — Salesforce, HubSpot, Roblox, growth-stage SaaS with new analytics functions — use build-signaling language at a higher rate than prior years. Scale briefs dominate at companies where the BI stack already exists and governance is the active problem.

The career translation: Read the JD for the signal before you customize anything else. Build brief: your answer to “how would you structure this org?” should be first-principles — team shape, tooling philosophy, one explicit decision in month one. Scale brief: your answer to “where’s the debt?” names a specific failure mode in the current stack and a logic for retiring it. Build pitch for a scale role is the wrong answer. Scale pitch for a build role is the same problem in the other direction.

3. Measurement analytics leadership at emerging social ad platforms now requires benchmark-first design — there is no benchmark to inherit.

The shift: At established platforms, the measurement analytics leader inherits an attribution methodology. Flawed, contested, but inherited. At Roblox, Spotify’s video ad formats, and TikTok’s shop integration, that methodology does not exist. The leader has to design it. The JD language is telling on itself: “novel format attribution,” “advertiser measurement education,” and “baseline methodology design” appear at Roblox and similar platforms. Those phrases are absent from comparable JDs at Snap or Pinterest. Not the same job.

The evidence: Roblox’s measurement analytics leadership reqs in H1 2026 specify benchmark-design responsibilities and advertiser education ownership — language that signals a zero-to-one problem, not a scale-and-refine one. Spotify’s advertising analytics briefs carry similar requirements for video and podcast formats where the historical comparable does not exist yet.

The career translation: The portfolio artifact for a Roblox or Spotify search is not a ROAS dashboard. It is a benchmark-design document: given a novel format with no historical data, here is the measurement framework — what I’d measure, why, what confidence threshold makes an advertiser trust it enough to scale spend. Most candidates applying for these roles have not written this document. The JD has been asking for one.

Next Action Plan

Portfolio. Target: Head of Analytics or BI Director, build brief. Write a one-page AI/BI governance framework. Three components: the specific failure mode your AI analytics layer would produce that automated testing wouldn’t catch — one example from your own history; the escalation path when the AI tier produces a confident wrong answer for an exec; the acceptance criterion before a number goes into a VP dashboard. By Friday you have the artifact. The Summit audience will hear this framing on June 2, which is the thing.

Networking. One message to a BI or analytics leadership contact running Snowflake Cortex or Databricks AI/BI in production. Three sentences: “I’ve been thinking through how analytics leaders handle governance review before AI-generated outputs reach exec stakeholders — is that process living in the BI platform, in dbt contracts, or a separate review step?” That question has a real answer. “Thoughts?” does not.

7-day micro-skill. Run Snowflake Cortex Analyst or Databricks AI/BI on the free tier against a dataset you know well. Ask ten questions. Find the one it gets wrong. Write three sentences: the question, the wrong answer, the governance step that would catch it. By Friday: one concrete failure mode you can name in a Director-level governance conversation. The keynote will give you the demo story. You want the audit story.

Sources

Crafting runs weekly. The job is to be useful, not to sell. If a line in here is close to a move you're working on, the email reaches a person.

Hiring rather than hunting? Signal is the letter for your side of the desk.

Get next week's issue Email me

← All Crafting issues · Drafted with Claude · Edited by Paul Brown