I’m an ex-London, now San-Francisco based developer.  My day job is at Google, where I work as the Tech Lead/Manager of the Content AdWords Frontend - the parts of AdWords you use to create, manage, and view everything from demographics and audience breakdowns on Display campaigns.

(In other words, all those ads you see all over the internet are, at the moment, my fault.)

I've got multiple teams, working in different areas on things unique to the display advertising business - ad construction, creatives, display targeting, audience, and remarketing.  Our business is divided into a couple of different human-readable segments:

  • Performance advertising - ads where the advertiser pays only for measurable results (a view, a click, or a conversion), the heart of Google's business today.  We do stuff like Ready Creatives - give us a website, we'll scrape it for images and text, and build you a ready-to-use ad, and Search Network with Display Select - a high-performance hybrid search/display campaign where we match the performance of ads across the entire internet to match the performance of your ads on Google Search.  Or remarketing campaigns, so that pair of shoes you almost bought will haunt you anywhere you go on the internet.  Or you can target only in-market users, people who've declared through their actions on the internet that they're in the market to buy a car, or a house.  Magic.
  • Brand advertising - Whether it's pay-per-engagement of a hover-to-play, or a hover-to-expand, or interacting with an embedded game, engagement ads engage the user with the creative without taking them away from the site they're on.  Buying only viewed impressions upheaves traditional CPM advertising by pointing out the obvious - advertisers shouldn't have to pay for an ad that wasn't seen by a user.  And affinity segments goes beyond coarse demographics to let you target users by what they're actually interested in, so advertisers can ensure that their brand has share-of-voice in any future purchasing decision.
  • Mobile focus - as we've recently announced publically, we're focusing most of our efforts on helping advertisers understand the user shift of eyeball-minutes to mobile devices, and building focused advertiser tools to help them reach this mobile-native audience.  Our first big launch is a revamped and reimaged App Install and App Reengagement flow for AdWords, to help advertisers use their money to reach the audiences they can't find through the iOS App Store or Google Play Store to drive installs.

It's more complicated than that, of course.  Audience buying, as a business in the advertising world, consists of everything from TV's Gross Rating Points to media's basic demographic components (age, gender, etc.) into behaviour-based targeting systems like interest categories, in-market segments, and remarketing.  We then use mathematical magic to use what we learn from users that convert or engage into targeting signals to help us find more users like that, through similar audiences, automatic targeting, and conversion optimization.

Coding at work is primarily in R these days - I try to ensure we're making data-driven decisions, and spend my days doing impact analysis, forecasting, and helping ensure our product continues to provide users with a great user experience and advertisers with great performance.  Together, we turn backend mathematical models, prediction systems, and feedback loops into controls our users understand how to use and reports our users know how to read.  AdWords has a long way to go, but we're making great progress.

When I’m not at work, I play video games, and listen to an awful lot of gaming podcasts (Rebel FM, I choose you). I also have a dog named Daisy who takes up what free time isn’t absorbed by the previous two.

But if your’e here, you’re probably here because I write stuff in my spare time. At the moment, I’m working on:

  • Deckalytics, a deck analysis package for Hearthstone players that teaches deck constructors how to build competitive decks, and to do the heavy mathematical lifting required to properly analyze how a deck will perform before playing it.
  • AdSense Dashboard, a dashboard of performance metrics for 2.1-based (and beyond) Android devices, with dedicated layouts for tablet users.  This is on life support - I'm asking users who can (Android 4.x+) to switch to the official AdSense app I worked so hard to get staffed at Google.

Defunct apps include:

  • Weathrman, a live wallpaper for 2.1-based (and beyond) Android devices

My languages of choice at the moment, in order of frequency of coding, are Go, R, and Java.  (in order of experience, that's Java, R, and Go)