5 Growth Hacking Best Practices
1.Establish a Data Foundation
This is paramount. You need to make sure that you have a good data infrastructure in place before you start thinking about Growth. Do you have proper event tracking? Is there good internal knowledge of what a sign-up means? Is there a decent data governance? These are all things that should be aligned before thinking about Growth. Not every metric needs to be perfect at this stage, but the key ones need to be clear, reproducible, and actionable. If you don’t know what your users are doing, how can you help them?
2. Feature Agnostic Approach
While you might have sunk a lot into certain features over others, it is important to remain feature agnostic when you are looking at the data. It is okay to look at certain data first, but make sure to allot time to have a clear overview of your features. This is where you will be able to pick out anomalies, changes and ultimately get closers to your “Aha” moment.
3. Establish a Process and Stick to It
Before you start iterating and testing to your hearts content, make sure that your team has agreed to a process and remain consistent in how you record that process. Finding Growth opportunities does not work when you can’t compare between different hypothesis. Further, you might want to go back to them later when a new feature is released or you have a new piece of information, and it will be very difficult to figure out what the data meant 3 months ago. Your product and data are likely changing constantly, making sense of the data is reliant on a clear and consistent process.
4. Your Data Taxonomy Should Be a Living Document
Tech companies are changing rapidly and it is integral that as you not only see your customers grow, you set up your company to grow with them. That means making sure that when new people are onboard to the company, especially jobs that are reliant on data-based decisions, know what everything means. A sign-up might have meant something when your product what in its first phase, but it might mean something different currently and it should be evident not only that there was a change in meaning, but also when that change occurred. Further, it should be widely accessible internally, though not easily changeable. If there is a change in meaning of anything integral to the business, it should be discussed with business stakeholders and engineering stakeholders alike.
5. Don’t Be Afraid of Third Parties
While vendor lock in is certainly a concern, third party data products almost always contribute positively to a company’s data infrastructure. This is because most startups and scale ups don’t often prioritize data from the get-go and have a whole host of legacy issues. Further, they often do not have the dedicated resources needed in house to maintain finely tooled data governance and infrastructure. If your company is trying to figure out what old data means, is experiencing internal misunderstanding of data or a misalignment in what data product managers and executives are basing decisions off of, then it usually a good idea to scope out what products exist on the market to help you.
