09 Nov Interview with an Engineer that Knows Marketing
He had me at “We need to dynamically append sharing urls with utm codes.”
For those of you in marketing who have ever met someone like Pete O’Leary of Chefs Feed, you know what I mean. Pete is one of those rare, mythical – better than the now common unicorn – creatures: he is an engineer that understands the technical needs of a marketing team. Pete builds product and databases with the understanding that someone else will need to get to the data and utilize it for reporting results and for attributing new users and engagement to campaigns throughout the user lifecycle. Working with an engineer like this is what it takes to be successful as a marketer and is quantifiable by the amount of valuable time you will save gathering insights and making insightful decisions on the next steps for your marketing team. (marketing nimbleness) These types of engineers know the questions the marketer, leadership team and investors will ask before the product is built. They have experience building apps, websites and systems with flexible code structure, acquisition, SEO and proper indexing in mind. They know that you have to build a platform that is nimble enough to respond to changes in marketing strategy.
If you are an engineer at a high-growth company Pete recommends you consider these 5 things before you build:
#1 Standardize: there are lots of marketing systems and each one reports different numbers, you have to develop your own source of truth and trust it.
#2 Simplify: major brands with huge marketing budgets can afford to track dozens of metrics each day, you cannot. Work with your team to pick a small handful of metrics – 3-6 is ideal – then make sure they are correct and everyone understands them.
#3 Get lots of (internal) customers: pick your technologies carefully so that everyone in your company can get access to the data themselves. [inlinetweet prefix=”” tweeter=”” suffix=”- @subtractioncap @pete_o”]If you have to write SQL commands to get reports, then you will be the bottleneck for data[/inlinetweet]. Get a simple tool, train people on how to use it and evangelize its use. The more people use your analytics, the better.
#4 Get all your data and keep it: there are lots of data sources and lots of data. Get access to all the raw data from every system you depend on and keep it, forever. Storage costs almost nothing these days. You will find as you learn more about your data and get more requests from internal customers, you will need to go back and build new reports over old data to do year over year comparisons or test data quality, etc. You can’t do this if you got rid of your data to save a few bucks.
#5 Build your infrastructure early: the cost and complexity of using very powerful analytics systems has plummeted in the past couple of years. You should build a data warehouse with advanced analytics capabilities before you launch your product, not 1 year later when you finally realize you can’t figure out what the heck is going on.
Pete O’Leary, now CTO at ChefsFeed, joined his first startup in ’91 as a Macintosh programmer. He has worn many hats over the years but always ends up working with engineers at startups. In his spare time he codes and cooks at his home in Noe Valley, where he lives with his wife, his son and his dog.