Golden Data

Issue #7 · June 5, 2026

The Restructuring Wave Is Running Two Operations. Most Résumés Are Only Reading One.

Paul Brown Writes Golden Data from Ogden, Utah 5 min read

Every Friday I look at what the analytics job market produced that week. This one produced a rising layoff count and a rising Staff-level open req count simultaneously, and most of the analytics people reading the first number are not registering the second.

The 2026 restructuring wave — Coinbase cut 14% of its global workforce earlier this month, 144,000 tech workers affected year-to-date per TrueUp — is being processed as a single analytics-career signal. It is two signals. The exits are mid-level: business intelligence developers, junior analysts, and data scientists without production AI feature ownership. The opens are senior and Staff: roles where the job is to validate what the AI generates, audit what it costs, and tell a VP their number is wrong.

The two résumés are not the same résumé. Most of the coverage is describing one of them. The career decision is knowing which one you’re holding.

What’s actually moving in the market

Coinbase cut 14% of its workforce on May 5, approximately 700 workers, citing AI-driven restructuring. CEO Brian Armstrong said AI tools now allow non-technical teams to ship code and automate tasks with smaller, focused units. Data and analytics were in scope. In the same reporting window, Staff Analytics Engineer and Principal AE reqs at companies running production AI query layers held and in some cases added headcount. The exits and the entries are in different job families. Most layoff coverage is treating them as the same.

The 2026 aggregate — 349 tech company layoffs, 144,800+ workers affected through late May per TrueUp — is weighted toward IC2–IC4: BI developers, junior and mid-level analysts, and data scientists without production AI feature ownership. A widely circulated May 2026 analysis of 92,000 data-adjacent layoffs describes that mid-level cohort. Staff and above is not the cohort clearing desks.

dbt Labs and Fivetran signed a merger agreement in October 2025, expected to close mid-to-late 2026. Fivetran raised prices 4–8x for some customers in 2025. That pricing pattern does not stop at closing. The analytics org that enters the post-close renegotiation without a documented cost audit negotiates from weakness. That audit is a Staff AE deliverable. The window before closing is the negotiating window.

Meta’s advertising analytics engineering org is posting Staff and Principal AE roles — cost governance, semantic layer ownership, AI output validation — while its broader headcount sits below 2023 peak. The restructuring and the build-out are running simultaneously at the same company. That is the full picture in one frame.

Senior Analytics Engineer comp in 2026: $183K average base; Staff and Principal bands at $190K–$260K at major tech, per Glassdoor. At orgs with mature dual-track systems, Staff+ IC compensation matches Director-level. The IC-versus-management comp gap closed materially in the past three years.

Snowflake Summit 26 opens June 1 in San Francisco. The platform keynote on June 2 announces what becomes the new default screen for analytics engineering roles by Q3. The JDs follow announcements with a six-to-eight-week lag. The candidates who read carefully hold the gap.

What I’d do this week

Run a vendor-consolidation audit before the dbt/Fivetran window closes. The dbt Labs/Fivetran merger gives every analytics org a legitimate negotiating moment before close. Snowflake Summit 26 opens Monday, and the platform keynote on Tuesday sets the next six months of AI query pricing conversations. The org that arrives at those conversations with documented cost data holds the position. The org that doesn’t is explaining its usage to a vendor that already knows it.

  • The user moment: Monday morning, when the Summit keynote drops and someone asks what it means for your stack.
  • The shape: Cost audit — Fivetran connector spend, dbt Cloud seats, Snowflake credits against AI query volume. Not a migration plan. An inventory.
  • The time budget: Three hours across two days. Two to pull the numbers. One to write the one-page summary.
  • The artifact: One page: current vendor spend, AI query volume, one alternative you ran at least a minimal test on. The document does not need to be beautiful. It needs to exist.
  • What success looks like: You are the person in the vendor renegotiation meeting who already has the number. That meeting happens whether you prepared or not. The preparation changes whose position it is.

Write the sentence that tells you which side of the restructuring split you’re on. The Coinbase 14% and the 92,000 data-adjacent layoff aggregate are being read as one signal. They are two. Write one sentence: the deliverable my role produces that would still require validation before a VP would act on the AI’s version of it.

  • The user moment: Before your next 1:1 with your manager. Not to discuss it — to know the answer before someone else is in the room asking it.
  • The shape: One sentence. Written down. Not a document for sharing — a role audit for yourself.
  • The time budget: 20 minutes. Alone.
  • The artifact: One sentence. If you cannot write it in two minutes, the sentence is not there. That is itself the finding.
  • What success looks like: By next Friday, you know which side of the restructuring split your current role sits on. If the answer is the wrong side, that is the most useful thing you found this month. You still have time to move it.

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 column for your side of the desk.

Email me Read more Crafting

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