Looking through career pages of popular tech companies or browsing LinkedIn, it’s more and more common to see job listings for growth hackers or growth engineers. But who is a growth hacker and what kind of engineering team is growth? At first glance, it doesn’t sound like a more traditional team, which usually includes words like platform, infrastructure, database, security, frontend, backend in their titles. If you’ve come across growth hacking and want to know about growth engineering, you are in the right place.

Welcome to the AARRR Framework

SAAS companies build products that help people solve specific problems. Being able to match people with these problems to the company is the first step in acquiring customers. Every growth team works on one or more steps in the funnel known as the AARRR framework. Introduced by Dave McClure, founder of a San Francisco-based VC fund and accelerator 500 startups, AARRR is designed so that the more customers that use the product, the more the company can invest into features for the product.

Every mission driven company wants to get customers to use their product and for customers to get value out of the product. The ultimate goal of a growth team is to enable, activate, and retain paying customers on the platform who like the functionality so much, they’re willing to refer others to it—which creates a flywheel of satisfied customers and boosts revenue along the way.

Growth organizations are structured slightly differently at every company. Most companies will have a growth engineering team which closely works with marketing, data engineering, data science and the product teams. Early stage startups that do not have data teams usually start off with just a growth engineering team. As this team matures and the company grows, a part of the growth team spins out to become the data engineering team.

Getting to Know the User

The first thing a successful growth engineering team needs to do is to know their customers. Knowing your customer means knowing where they come from, their intent when they land on your homepage, how they interact with your site, how long they spend on your site, and whether they leave your website without navigating to another page (bounce). These are all important questions necessary to make informed decisions. Gathering this information is the first step the growth team makes. After all, when you know your customers well, you’re better equipped to serve them well.

To start out, marketing websites and applications need to be instrumented with analytic events so that the team can make informed decisions and measure the performance of features. The growth engineers then start instrumenting the startups marketing website and all their product applications to better understand visitors and customers behavior.

Frontend events

• Page Views
• UI Clicks
• Scroll Events
• Hover Events

Backend events

• Order Placed
• User Created
• Lead Saved

There are far more front-end events in a single page application (SPA) than just basic UI interactions because a lot of the application functionality is rendered on the client side instead of being done in the backend. For example in an audio playing application your events could be:

• Song Played
• Song Skipped
• Album Selected

Once the apps are instrumented and analytics events start flowing through, then we need a way to visualize and query the information we receive. To do this, use third party tools like Amplitude or Google Analytics. As your startup matures, the part of the growth team that worked on data analytics and instrumentation will end up growing and becoming the data team that will build out the company’s data warehouse. Care must be taken at every step and at every level as to what is collected and how it is stored. Ensuring visitors are anonymous and they are not identifiable from the start is a good way to ensure the data you’re entrusted with is kept secure and private.

Experimenting to Create a Delightful User Experience

Once we have instrumentation, then the growth team can start running experiments. Why should you experiment? Not experimenting is actually running a single experiment with all users all the time without having a way to measure success of the experiment. Any experimentation you do should be in service to the user interface (UI) with emphasis on conversions as the primary metric. That way, you can validate positive changes to the UI and provide customers with a delightful customer experience.

Experimentation plays a large part in features built by the Growth team. The Growth team quickly iterates by implementing hypotheses and experimenting the implementation with a user segment along with a control group. Doing it this way allows the team to compare the variations and have data to back up productionizing a winning variation or learning from the losing variation.

Every experiment is a learning experience. Not all experiments will win, but by experimenting you quickly learn what works and what doesn't. Growth Engineers build out the framework for running experiments in the company. Measurement of the experiment is also an integral part of the process and growth works with data scientists and analysts to ensure experiments are measured correctly. Automation is the key of the game here and high performing growth teams work on automating the experiment framework.

The beauty of experimenting is that it provides a framework for anyone with an idea, to intentionally and efficiently formulate an experiment and test out their hypothesis. At Gusto we have an open culture for running an experiment. We have an experiment idea review meeting where anyone from marketing, design, data science or engineering can bring their ideas and can discuss the validity of their hypothesis. Using our experimentation framework we can then convert the idea into an experiment. Experiments are then scored based on the ICE framework. The top scored ideas are then run as experiments.

Engaging with Users

Engagement is another area that a lot of growth teams work on. Most applications communicate with users in three ways:

• Transactional messages are messages sent to the user as a response to a user action—for example, a confirmation email when a user makes a change in the application.
• Marketing messages are usually emails sent by the marketing team to a wide circulation list promoting the company or product.
• Engagement messages are usually data driven messages that are targeted to certain users based on their usage of a product, like an email sent to a user who has not logged into the application for some time. Engagement can also happen inside the application. These can be targeted informational messages or call to actions shown to the user via UI components like cards, modals, banners.

Growth teams build the framework for engaging with the user. These include the communication channels to engage with users, both in the product and via email, sms or push notifications. Growth teams success metrics are centered on driving engagement, rather than on volume, and are used to ensure users are not too overwhelmed with notifications. Being able to intelligently communicate with users at the right time is also an important tool.

Personalizing the User Experience

Selection of users, also known as audience segmentation, is also a key component of growth engineering. An example of this can be seen in a lot of mobile applications. In popular ride sharing apps like Uber and Lyft your friend might have a promo code with a name like “30OFF” that gives them 30% off on rides. When you add that same promo code to your app, you get an error message saying you are not eligible for the promotion. Why does this only work for your friend and not you? Who decides this eligibility and why does the app like your friend and not you? This is known as targeting and your friend was part of a promotional user segment. Another example of this is shopping at an online clothing website with a friend on multiple devices and seeing different prices for the same item. Or getting an email with a personalized “just for you” 20% coupon code.

Sophisticated, responsible targeting and personalized recommendations is what most mature growth teams want to be able to offer. This is where you can fine tune experiments for specific segments of your users based on certain actions they’ve done in the past and use this segment to run an experiment. Personalized recommendations, when done responsibly, are beneficial for both the company and the user. Choosing a user who is on the fence at buying an item with a discount code helps the user as well as the company. Pairing users with messages aimed at improving discoverability of the application, helps the user navigate and use the application more effectively.

For personalization, growth engineers work with data scientists to build intelligent recommendations using machine learning models. With the experimentation framework the team can test out different models for recommendations and see which works best for each application of recommendations.

Summary

The goal of a growth engineering team is to use their technical skills to achieve all of the above. The growth team is very metric driven and all changes made are measured to understand the impact of the change. By having a growth mindset, the team makes data driven decisions and can iterate quickly by experimenting with different ideas.

At Gusto we have several growth teams working throughout the funnel to improve all the AARRR framework metrics. We deal with many challenges in reducing friction in our onboarding process, engaging users in our application and being able to help users with discoverability. At the top of the funnel the growth acquisition team works by experimenting on the homepage, lead creation forms, SEO optimization to name a few. Understanding users intent and customizing the top of the funnel experience based on that intent provides the best user experience.

In the application the growth team iterates on the onboarding process to reduce friction for the user and improve the speed at which they can run their first payroll. Another Growth team works on building out product recommendations and call to actions for customers to drive feature adoption and upsell. Another Growth team is working on optimizing engagement and communication to the user in the application and via channels like email and text messages. If you are interested in these challenges and have a growth mindset in being metrics driven, reach out to us on our career pages.