AI will predict how ads perform, before you ever spend a dollar. Learn what else our Voyager Release can do. 🛰️

The 7 Necessary Steps for Creating a Growth Marketing Machine That Works

Guest Author May 8, 2017

Gone are the days when titles such as Growth Manager or Growth Hacker were solely reserved for giant corporations or cutting-edge Silicon Valley unicorns. Today, more and more companies are investing in growth departments to explore new channels and diversify revenue streams.

If you’re still skeptical about the rise of growth marketing, just look at Optimizely’s acquisition of Experiment Engine, a SaaS experimentation platform that helps accelerate growth initiatives and democratize data and learnings. Or Quantifi, a marketing R&D platform for digital ads that allows companies to experiment at scale. Not only are companies investing in growth, but new products are being built to help facilitate it.

That being said, growth can be messy. Assuming you already have a good product/market fit, you need to formalize a growth process that’s repeatable before you can start to see the benefits. To get started, here are the seven necessary steps to building a successful growth marketing machine.

The New New Thing Podcast: How a Dedicated Growth Team Can Transform Your Business

1. Determine Who Will Be on Your Growth Team

In a perfect world, you would have a dedicated Growth Manager who works with the Growth Team to implement campaigns across your entire business. This is, of course, not always the case.

Start by asking whether you want someone (or a few people) focusing all their energy on growth, or if growth will be a separate responsibility in addition to day-to-day employee duties. Keep in mind that even if you have one person devoted to developing the growth strategy, other parties with vested interest are needed to implement the actual test. Your Growth Manager will likely be the prioritizer and organizer of ideas who pushes the team forward and ensures the team hits a specific testing cadence, whereas employees within the experiment’s department will be responsible for actually pushing the test.

On the other hand, you could choose to instill a growth mindset within different departments across the business. In this case, you might not have a true Growth Manager who controls all the experiments, testing, and analysis, but instead the responsibility would be dispersed among employees within specific departments.

The key to this approach is holding the department accountable to growth initiatives on top of their day-to-day responsibilities. Without accountability, growth can easily be neglected. Instead of saying “we should test more,” set specific goals like “we will execute X number of growth experiments every week, month, quarter, etc.”

2. Source Ideas

Once you’ve determined who is responsible for growth initiatives, you need to gather ideas. The Growth Manager should not be solely responsible for coming up with all the ideas your company will test. Not only would this be a ton of pressure put on this person, but you’ll also miss out on many great perspectives and ideas that come from other parts of the business.

Generating Ideas Through Your Employees

Your employees have different perspectives on your product or service. Take advantage of this. Company leaders’ growth ideas versus people in operations, technology, or customer service will provide a good holistic view of your product or service. As an added bonus, encouraging them to put ideas forward will help instill the growth mindset into the company as a whole.

Looking at The Competition

Stalk your competitors for new growth ideas. If they’re dipping their toe in a new channel, maybe your brand might want to test it as well. Even if you don’t want to blatantly steal their ideas, getting a little inspiration can help jump start your brainstorming and open your mind up to new possibilities you may not have previously considered.


Stay up-to-date on new ways people consume data. This includes reading industry-leading blogs and news, as well as listening to podcasts. Another great place to monitor new opportunities is through mobile app stores. Often apps can provide a new way to market to customers, and investing early could lead to a first mover advantage and higher returns.

Formatting Your Ideas

Finally, make sure you write each idea in a provable format (similar to a hypothesis) – if we do X, then Y will happen. This is important because it will both help you determine the success of the test later on and educate different parties within the organization when you share results.

3. Prioritize Ideas by Merit, Not Stakeholder

Do not play favorites when deciding which ideas to test. Just because an idea comes from the CEO does not mean that it’s inherently better or worse than an idea from the intern. To make decisions in a nonpolitical way, develop a process to rank your ideas.

Although there are a ton of ways to prioritize your testing, one popular method (and good place to start) is the Impact, Confidence, Ease/Effort (ICE) scoring, developed by Sean Ellis. The components are as follows:

  • Impact: The overall potential return of the idea
  • Confidence: How likely you believe the experiment is to succeed or fail
  • Ease/Effort: How easy or difficult the idea is to implement

Rank your ideas on a scale of 1 to 10 for each element, 10 being the best and 1 being the worst, then find the average. Move ideas that score the highest to the top of your list and low-scoring ideas to the bottom. Once sorted, begin testing and experimenting with the ideas that score the best and backlog the tests that scored lower. You can also begin to categorize ideas into buckets such as email/retention, ecommerce, referral, and acquisition to improve specific areas of your company.

4. Build a Testing Plan

As mentioned previously, make sure you have a set cadence you want to hit (most experts suggest a high tempo testing rate to find learnings quickly). Do not be discouraged if your first few experiments fail. In fact, many growth marketers encourage failure. But if you don’t set a cadence to your tests, you’re more likely to push growth to the back burner.

After you decide on ‘X number’ of tests each week or month, make sure you have the framework in place to measure the experiment’s success. This means deciding which specific group you’re going to test and which pieces of technology you’ll use to track the outcome.

For example, some companies with customers that have longer buying cycles might consider segmenting those people into cohorts depending on which part of the funnel they’re experimenting with.

Types of Tests

According to Ellis, there are two main types of tests:

  1. Tests to discover (pings): These tests are created to identify new channels or opportunities to market your product. You’re looking to put something out there to see if you get a response, and ultimately assess the potential of new channels.
  2. Tests to optimize (A/B tests): These tests build upon what you already have, be it landing pages, content, product pages, etc. A/B testing helps lift conversion points by exposing different variations of whatever you’re testing to the consumer. It’s important to keep in mind that you need a decent amount of traffic to successfully A/B test, so if your site is low on visitors you should work on that first.

Finally – and this is often overlooked – make sure you think about your brand and what it stands for. Ask yourself, “If this experiment works, would we realistically implement it?” If the answer is no because it goes against who you are as a company, then don’t bother testing it.

5. Execute Your Plan

If you’ve put in the work beforehand, prepared how you’ll measure success, and determined who will be involved, how long the test will run, and how you’ll communicate to vested parties about the status of the test, execution should be pretty straightforward.

The difficulty, however, arises when you have a lot of different tests happening at once that may or may not interfere with each other. For example, if you’re testing multiple elements on your homepage and conversions increase, it can be hard to tell which test was responsible for the lift. To do your best to avoid confusion, keep things organized and update everyone on a regular basis.

6. Evaluate (and Share!) the Results

Consistent communication between your team and the appropriate players is critical. You might consider a weekly standup meeting where you discuss what’s being tested, why it’s being tested, and what you may have learned from the test to help keep others in the company in the loop. Sharing the results (even if they’re failures) also helps your co-workers learn with you. No matter how terrible the outcome of the test, democratizing the data will help ensure the same mistakes aren’t made repeatedly.

7. Iterate, Repeat

Take your learnings, iterate, and repeat – growth is a cycle. For example, let’s say you decide to test a new checkout process on your website. You find that the new flow increases conversions so you decide to implement it. Great! Now that you saw success in this area, maybe you test a similar flow when customers sign up for an account on the website and so on and so forth.

The same applies to something that’s harming conversions. Maybe adding video to product pages decreased conversions, so now you want to test some of your landing pages without video. Or perhaps you notice that video actually converts better on desktop but worse on mobile, so you test static images in place of video across your site on mobile devices. The point is: continue to take nuggets of information away from each test to continue the improvement process.

Growing is a Mindset

Whether you have a dedicated growth team or not, the chances of success are far higher when every employee buys in to the growth mindset. Your employees interact with your product in vastly different ways – encourage them to share their knowledge! And once you create an environment where departments feel empowered to take smart risks, don’t be afraid of failing. Keep the testing cadence up, learn from the results, and do it all over again — it will pay off.

Featured image: Corban OneSource