According to Wikipedia, the goal of growth hacking strategies is generally to acquire as many users or customers as possible while spending as little money as possible. A growth hacking team is made up of marketers, developers, engineers and product managers that specifically focus on building and engaging the user base of a business.
The typical growth hacker often focuses on finding smarter, low-cost alternatives to traditional marketing, e.g. using social media, viral marketing or targeted advertising instead of buying advertising through more traditional media such as radio, newspaper, and television.
For the last decade, growth hacking teams have been focused on finding data points or opportunities to “hack” the system and acquire clients at a lower cost via product iterations or UX, to increase the LTV of users. All of these findings are great, but with the advent of Artificial Intelligence, Growth Hacking teams are struggling to integrate Data Science into their day to day process. The main reason for the struggle is companies are hiring Data Scientists, data architects and data engineers who live in silos within the organization, in part because this is a department that serves the entire organization, not just Marketing. But, wait, we had the same problem before we came up with Growth Hacking teams right? I mean, engineers and product managers were also part of different departments in the past and Startups were the ones who figured out that companies should bring this talent together into growth hacking teams.
Now, with Artificial Intelligence is not that simple because unfortunately, the talent is scarce. By 2025 there will be a shortage of 2.5 million data scientists according to Linkedin and Upwork. Usually the Data Science team, if it exists at all, is small and tends to cause inefficiencies. For example, imagine departments like operations, sales and marketing funneling all of their requirements at once, and to a single and small department. These departments become sort of “labs” within the organization.
As growth hacking teams understand the importance of Artificial intelligence in their operation to for example, reduce churn, predict user behavior, or increase LTV, some profiles inside the team have started to migrate into data analysts. So now, the problem is different. Tools in the market for AI are still too specialized and siloed, the same way organizations are. Existing tools are focused on Data Science and Data architects, the same way that in the 80’s airplanes needed to fly with a flight engineer. Functions of the flight engineer included inspecting the aircraft and overseeing fueling operations before flight. During the flight, the flight engineer monitored the performance of the engines and cabin pressurization, air conditioning, and other systems.
Bruce Dale/National Geographic/Getty Images
So, like flying in the 80’s, Growth hacking teams need to solve the problem the way the airline industry did 20 years ago. First we need developers, product and marketing people to get trained in data and Machine Learning. They need to understand how it works and what it can do for a business. The same way pilots had to be trained in avionics and engineering.
Second, we need the technology to catch up, we need no code tools that can be operated by non technical professionals, the same way pilots can now fly planes without a flight engineer. For example, having access to a software that can by default, set a Spark configuration to run a linear regression and then, automatically send the output to a business application.
If mitigating from a Marketing profile to a growth hacking professional was hard, upgrading yourself to a growth science professional will be 2x harder. A growth science team leader would need to have strong management and marketing skills, computer science background and data analytics capabilities. The team will now have to operate and execute a bit differently, first aggregating and cleaning large amounts of data. Analyzing and visualizing. Building theories and hypotheses on how AI and ML can solve specific problems at scale and preparing business applications to run the outputs of the models. It is just not anymore about working with Excel files or Google sheets and building rule-based systems. There’s a new business level, and professionals need to level up, fast.
Although we have started to see the shift, very few professionals are talking about it. Let’s start the conversation.
Check out our new platform Datagran 3.0 Beta below.
