About
I've spent 10+ years turning data into an unfair competitive advantage.

I'm Malcolm Angus, in San Francisco
. I've never fit one title: analytics engineer by trade, data product manager by instinct, forward-deployed engineer by necessity. I've done the work at every altitude: keyboard-level pipelines, principal-level analytics at Atlassian, founding and running the data team at an SEO agency, and now building agentic data systems at Retool. I work both ends of the mess: greenfield 0 โ 1, the first data product where there is no pipeline, no agreed metric, and no obvious place to begin; and brownfield data work, wiring new AI and data into a company that already has systems, politics, and a real process that lives in people's heads instead of the documentation. The through line is ownership: being the one person accountable for whether a data product changes a decision, not just whether it ships.
Along the way I kept meeting the same gap. Companies collect data for years, call it a moat in the pitch deck, and never build the loop that would make it one. The engineering is rarely the hard part. The hard part is strategy: choosing the decision worth improving, instrumenting outcomes instead of events, and resisting the platform until it hurts.
That gap is what this site is about. I write essays on data products, moats, flywheels, business strategy, and how markets really work, the same questions my day job keeps circling back to.
The bull? Angus. Family name. You can thank my ancestors for bringing the steak ๐ to the USA.
Where I've done the work
2024 to present
2+ yrs
Retool ยท Data & Analytics Engineer, PLG & GTM
The B2B SaaS platform for building internal tools fast.
Building the data models and the agentic data agent that power on-demand quantitative and qualitative insights for product and go-to-market teams.
2020 to 2024
4 yrs
Organic Growth Marketing ยท Head of Data & Internal Tools
An SEO and content agency for B2B SaaS companies.
Founded and ran the data team: the SEO data pipelines, measurement systems, and analytics products used across client engagements.
2016 to 2020
4 yrs
Atlassian ยท Principal Data Analyst
The B2B SaaS maker of Jira, Confluence, and Trello.
A four-year run from senior marketing analyst to principal: PLG marketing analytics for devtools, attribution models, channel portfolio management, and search analytics, then marketing data ops, web scraping, and competitive intelligence at principal level.
2014 to 2016
2 yrs
LiveCareer ยท Marketing Acquisition Analyst
A B2C SaaS company behind consumer resume and job-search products.
Started where most data careers should: close to the spend, where the numbers had consequences the following Monday.
Selected public work
Products
Divisadero divisadero.co โ
Ad intelligence
Most ad-creative testing is paid guesswork: teams burn budget re-testing ideas competitors already proved don't work. Divisadero is the ad-intelligence platform I built to fix that. It ingests and AI-scores thousands of competitors' Meta ads into a searchable archive of the messaging, formats, and hooks that actually work, so a DTC wellness brand starts from proven concepts instead of a coin flip. Better creative, lower CAC.
Thunderdome thunderdome.io โ
AI visibility index
A free index of which tools ChatGPT, Claude, and Gemini actually recommend to buyers, across dozens of B2B-software categories and refreshed monthly. I built it end to end, the three-model measurement pipeline and the site, with no paid placement so the ranking is the product.
Nopalito nopalito.io โ
Brand data
A 0 to 1 data product I built that pulls a brand's full identity, palette, type, voice, and a design scorecard into structured, reusable data.
Menuomics menuomics.com โ
Menu engineering
A 0 to 1 data product I built: behavioral-economics menu engineering that rewrites a restaurant's wording, pricing, and layout from cited research. Sixty-plus published breakdowns.
Writing & talks
How a Data Team Built a Production-Ready Customer Insights App Retool webinar โ
Building an internal tool that sales, success, and marketing actually use without borrowing a sprint from product. Read the write-up I authored โ
How to build data products for the next decade Retool blog โ
A guest feature on the data products I built at an SEO agency: dashboard-discovery search, findable dashboards, and read-only reports turned into interactive tools.
Demo: How to build data apps and products for the next decade YouTube โ
A live demo of building data apps and products for the next decade, shown end to end.
Building an SEO Data Pipeline Medium โ
The engineering write-up on turning search data into a pipeline a team can actually make decisions with.
Data Moats & Flywheels: a field guide LinkedIn โ
A ten-part visual guide to spotting real competitive advantage in the age of AI, apart from a big pile of data.
Elsewhere
Daily notes on LinkedIn, longer archives on Medium, or send a note through the contact form.