Mike Andrews Ph.D is Chief Scientist (Forensiq) at Impact Radius, and is carrying out some fascinating work around digital marketing and ad fraud.
We caught up with Mike to ask him about his work around ad fraud, and what a typical working day is like…
Broadly speaking, what are your key business goals? And what are the most useful metrics and KPIs for measuring success?
Fighting fraud and identifying emerging threats. Educating the industry and our clients in collaboration with partners including academia.
KPI’s are more for the engineering side, rather than for science per se, but still, at least on the data science side, having more folks making more use of our various fraud databases over time is a good indicator of how well we’re doing.
What is the biggest challenge in your role?
Making sure everything we do is as scientific as possible, based on state-of-the-art understandings and techniques, “standing on the shoulders of giants” as it were. It’s a challenge because that’s not always the most expedient way of doing things, there’s no shortage of tempting shortcuts or hacks that just don’t work out well.
Science means simple, evidence-based, predictive, and experimental; aka “the scientific method.” also: objective as possible, although i am a big believer in “gut,” at least in the face of uncertainty.
How is the function of marketing evolving within your organization?
It’s far more than just social media management or branding; one direction it’s evolving is towards promoting unique innovations we make within the company to the market, for example providing thought leadership in the form of studies, like our mobile device hijacking work.
On the pure marketing side separate from our fraud fighting efforts per se, there’s a lot we’re doing on the data science side of marketing campaigns, such as tracking and optimizing engagement.
Which tools or software do you find most useful in your role?
Browsers are the greatest things since sliced bread for user interfaces, and I use them that way quite often, although i’ll take a good command-line interface over a graphical one any day.
What do you see as the most significant trends in digital marketing over the coming years?
Ever-deeper understandings of the customer and their state of mind, including being able to predict relevant stuff about what they’ll do generally, especially in a causal way, meaning beyond just finding correlations but actually understanding likely responses to interventions.
Fraudsters will continue innovating and thus so must we in countering.
What kind of skills do you need to be effective in your role?
broad and deep understanding of science & technology, an ability to make complex customized things from scratch, getting along well with both business and technology folks.
A sense of where the industry might be headed to try and get there first, and an ability to assess prospective folks’ talents for doing well on a team.
Tell us about a typical working day…
Starting at 4am, drink some coffee while catching up on the latest in science and technology, which is mostly scanning all the new scholarly & technical papers from universities and tech giants alike; usually there are a handful of significant ones.
Also taking care of most urgent items I’ve “slept on,” like project plans, etcetera. Join a class at the gym 5-6am, then from 6-7am continue working on my highest priority items, like sending some emails i’d like answered later that day. 7-8’ish: breakfast with family and get the kids to school.
8am-5pm: the bulk of the day, may or may not include 3 hours of commuting by foot and train between NJ & NYC; if so, looking forward to a day tilted towards meetings, but if not, WFH and get some seriously productive uninterrupted coding or similar work done.
6-7’ish: have dinner with the family, then afterwards check kids’ homework, make sure they get ready for bed, and hopefully turn in by 8pm the latest!
Do you have any advice for people who want to work in the digital industry?
Really try to understand things from soup to nuts, devil-in-the-details style, as technically and as broadly as possible.
Build working prototypes of your concepts, deploy them if possible too. Do your best to understand the history of how things developed, and keep abreast of ongoing technical conversations and new ideas.
There are no dumb questions! The more you ask, the more you learn.
2017 will be a watershed moment for video, as consumption moves from the TV to other devices.
As it prepares for a 2017 IPO that could be the largest in the social media space since Facebook went public in 2012, all eyes are on Snapchat.
New Top-Level Domains (TLDs) have become more popular in the last couple of years, so here’s everything you need to know about them.
When measuring the effectiveness of discount codes, retailers often get it wrong. In this article, we'll look at how data-driven attribution can help businesses better understand where discount codes produce the best ROI.