Sean Ellis coined the term “growth hacking” way back in 2010. Since then, the term has taken on a life of its own.
It’s the focus of dozens of books, new roles, new departments and teams, new methods of thinking, hundreds of articles, hundreds of guides, hundreds of webinars… you get the idea.
Yet, it still feels very elusive. High-growth companies simply have something most companies don’t, right? Some secret growth hack or silver bullet that skyrocketed them to household names.
Wrong. The truth is, they simply had a solid growth marketing process.
Growth Process > Growth Hacks
If you’re familiar with CXL, you know that we preach process over tactics to optimizers. Growth experts largely believe the same. One of my favorite quotes on the topic comes from Paul Graham of Y Combinator who once said…
Yet, marketing teams continue to make decisions based on what has worked in the past, what has worked for other people and what appears to be working for competitors.
In our 2016 State of Conversion Optimization Report, which many people in growth roles completed as well, we found that…
- Only 32% of small companies (less than 500 employees) say they have a documented, structured optimization process.
- Even at large companies, that number only increases to 42%.
- 26% of all companies meet with their optimization team “only when necessary”.
The takeaway is that process, though supported by the experts, is still not the norm. Rob Sobers of Varonis confirms…
Worse, the idea of golden “growth hacks” is still alive and well…
Growth without effort, growth without resources, growth without spending money. You get the idea.
Morgan Brown, co-author of the upcoming Hacking Growth, says it best…
You’re not going to find your next major growth breakthrough in a growth hacks listicle. But since a silver bullet that works for everyone all the time is so appealing, it doesn’t stop many people from looking.
As David Arnoux of Growth Tribe explains, the successful growth hacks you’ve read about from tech superstars are merely the result of a finely-tuned process…
Though, as Morgan explains, a major growth breakthrough is rare. Typically, it’s a game of continuous small wins that add up over time…
So once you find that golden nugget, silver bullet that works perfectly, as David said, it’s back to the loop, back to the process.
It Starts with the Product
It should go without saying that growth starts with the product. If you don’t have a validated product or business model yet, stop here. You can’t growth hack your way into problem-solution fit or product-market fit, so really focus on that first.
Make sure you have a good understanding of your value proposition, cost structure, revenue streams, customer segments, etc. before continuing.
Ideation x Execution = ROI
The ROI of a growth process is ideation multiplied by execution. If you have a great idea, but it’s executed poorly, it isn’t working to its full potential. If you have a bad idea that’s executed flawlessly, it won’t work.
That’s why growth processes are designed to make both ideation and execution smarter and more efficient.
Of course, there a lot of different growth processes. Just as there are a lot of different conversion research models you could use or prioritization models you could use.
Here’s Growth Tribe’s G.R.O.W.S. process…
And the process Brian Balfour teaches…
And the process Morgan teaches…
But, you’ll notice that they all follow a similar flow…
- Come up with ideas.
- Prioritize those ideas.
- Plan out the execution of the highest priority idea.
- Execute the highest priority idea.
- Analyze the outcome of the idea.
Then, you start over again, armed with the knowledge from your previous experiment.
You’ll see how steps three and four focus on smarter execution while steps one, two and five focus on smarter ideation.
Set the goals and the timeframe.
Let’s back up a bit first, though. Before you launch a growth process, you’ll want to consider your goals and your timeframe for each goal. That starts with setting your objectives and key results (OKRs).
There are a lot of different ways to approach this, depending on your business model and your goals. But I really like a concept that Drew Sanocki of Empire Growth Group teaches in CXL Institute’s Ecommerce Growth Masterclass…
So, to run with what Drew teaches, you could set three key results that, when multiplied together, would double your revenue. Drew talks about this in the sense of ecommerce (number of customers, AOV and retention), but you’ll find that the formula works for a lot of other business models as well.
Hyper-Focused Growth Cycles
Now you have three core goals for the year. Then, you introduce the sprint mentality. If you break the year into four 90-day, hyper-focused growth cycles, you have more than enough time to focus on each of the three multipliers.
So, why 90 days?
The 90 days is perfect because you can’t really procrastinate but it’s also not absolute panic.
It’s a short enough timeframe that you feel the pressure, but not so short that reaching your goal feels impossible. If you don’t hit your “annual size” goal in 90 days, you have the fourth 90-day period to fall back on.
But the beauty of combining these two concepts is that you aren’t dealing with “annual size” goals. You have the Drew Sanocki-inspired multiplier goals, which are, by design, more easily attainable.
(Note: Drew suggests starting with the retention goal first.)
It can help to be laser-focused for a full three months, but you can alternatively focus on all three core goals throughout the year. Just ensure each new idea added to your backlog (more on this shortly) is tagged with the correct multiplier.
Your Tempo and Resources
You’ll also want to ask yourself…
- How many experiments will you run per week?
- What resources do you have available to you during this timeframe?
As Sean Ellis, co-author of Hacking Growth, explains, that second question is vital…
Be realistic from the beginning about what resources you’ll need and what resources will actually be available to you so that you can choose growth ideas and execute accordingly.
Come up with experiment ideas.
Now, you can move on to ideation. It’s the first step of the G.R.O.W.S. process you saw above…
What’s great about the method David describes is that it doesn’t rely on the traditional brainstorming model, which is surprisingly ineffective. You’ve definitely been part of the traditional model. A whole team around a desk, calling out ideas on the spot.
Instead, David’s mindstorming has everyone come up with the ideas separately and then bring them together (and build on them) in a meeting.
It’s important, though, that this step is done as a group. As Sean explains, problems arise quite quickly when only one person is responsible for ideation…
Opening ideation up to the entire team can provide you with new perspectives. It’s the same reason copywriters turn to sales teams and support teams when they conduct their research.
Introducing… the Growth Team
But let’s back up a bit to the idea of a growth team. What exactly goes into a growth team? Is it the same thing as a marketing team? Morgan explains it really well…
The big takeaway here is that it isn’t just digital marketers. So a traditional marketing team and a growth team are not one and the same. You have growth-minded experts from a number of different areas of the company.
This speaks to how modern teams are changing. We are slowly emerging from our narrow silos, according to Morgan…
Forming a growth team or even getting the whole company involved with ideation starts with a culture of experimentation. I recommend reading this article by Alex Birkett if you are trying to foster that type of culture.
Creating Your Backlog
Once you start collecting ideas, you’re going to need a place to store them. This is often called a backlog. Sometimes they’re in Excel…
(You can grab this template by clicking here. To change the drop-down items, highlight the cells and select “Data validation…”)
Sometimes they’re in a third-party tool, like Projects…
Note that Projects has prioritization, which we’ll get to shortly, built-in.
In the Excel template, you’ll see that I label ideas based on product, multiplier and channel. Each tool you come across will likely do things differently. For example, Projects asks you to pick a growth lever…
You’ll also notice the tags section. This all speaks to the importance of personalization. You might label ideas in your backlog based on multiplier, channel, product, stage of the funnel, growth lever, etc. Choose what works for you and your team.
Filling Your Backlog
When you’re first looking to fill your backlog, I recommend going through the full conversion research process. (You can learn how we do it at CXL by reading about the ResearchXL model here.) This is another example of how growth and optimization overlap and work together.
Beyond that, you’ll need to rely on your team for constant ideation. Of course, that can slow down as busy teams get, well, busier. Sean explains what they did at GrowthHackers to keep ideation going…
Prioritize your experiment ideas.
So now you have a massive backlog of ideas. What do you do with those dozens of or hundreds of ideas? As David explains, the next step is to prioritize them…
This is what the number beside my ideas in Projects is, a prioritization score. If you recall, they have you rank the idea based on impact, confidence and ease. This is called the ICE model, but there are others…
ICE and PIE are simple and commonly used by tools, making them more accessible. However, they are subjective, especially when you have the entire team submitting ideas.
For example, a colleague might consistently overestimate the potential of his ideas while another might consistently underestimate the potential of his ideas. A colleague might think all of her ideas will be easier to implement than they really are while another colleague might think all of her ideas will be harder to implement than they really are.
As you can imagine, serious consistency and objectivity issues can arise quite quickly.
At the end of the day, you’ll want to choose a prioritization model that works for you and your team. Aim for it to be as objective and customizable (e.g. perhaps branding is super important to you right now or the PPC channel) as possible.
Creating Your Experiment Doc
So, what happens when you prioritize your ideas and you’ve selected an idea to start with? According to David, it’s time to outline your experiment design so you nail the execution…
An experiment sheet or an experiment doc should outline…
- The objective you’re serving.
- Your hypothesis.
- Your start / end dates.
- The metrics you’ll be using.
- The pages that will be affected.
- Who you’re targeting.
- What the variations will look like.
- Your analysis and conclusion.
- The next iterations / steps.
Here’s an experiment doc example from Morgan, which I really recommend. The idea is that you’ve…
- really thought through the execution of the experiment,
- considered the external factors,
- and created something that anyone looking back on this experiment will be able to use.
Creating Your Hypothesis
When you’re putting together an experiment doc, one of the first things to consider is the hypothesis. You’ll have to take your idea from the backlog and turn it into a real hypothesis to explore. Morgan has some sage advice…
Here’s the formula Craig Sullivan, a veteran optimizer, uses…
We believe that doing [A] for people [B] will make outcome [C] happen. We’ll know this when we see data [D] and feedback [E].
And the formula Drew uses…
If successful, VARIABLE will increase by IMPACT because ASSUMPTIONS.
So, your hypothesis might look something like this…
By reducing distractions on the product page, our target audience will better understand our products and take the logical next step (purchasing). We’ll know this by observing revenue per visitor.
Experiments should be designed in a way that means you will learn something concrete, regardless of the outcome. If your hypothesis is not proven successful and the experiment is considered merely time and energy wasted, you’ve done something wrong.
Run the experiment.
Finally, the fun stuff. As David (and Rihanna) put it, this is the work, work, work step…
More or less, this is the part where you do what you said you were going to do in your experiment doc. But, there are a few things to consider.
- Read about smoke testing to save yourself (and your developers / engineers) a lot of grief. You’ll be able to validate your ideas quicker while consuming less resources.
- Be aware of external factors that are at play and that could be influencing the results of your experiments. Do what you can to limit them, but know that you can’t eliminate them. This article on sample pollution will give you some ideas. Always add external factors to your experiment doc.
- The more rapidly you can move through this growth process, the better. Consider high velocity testing / experimentation.
- Experimentation is not synonymous with A/B testing. You can run an experiment without running a test. However, that doesn’t mean that your experiments should never include testing. Take advantage of it if you can, but only when it makes sense.
Analyze the results.
After the experiment has concluded, you’ll of course want to analyze the results. For David and his team, this is the last part of the process…
Here are some questions you might ask of your experiment…
- Did you execute the experiment properly?
- If your hypothesis was incorrect, why?
- Would the results be different if you segmented the data?
- Should you optimize the experiment and run it again?
- If your hypothesis was correct, why?
- What can you take away from the experiment to run smarter future experiments?
If an idea works, optimize it and double down on it. How might you be able to apply what you’ve learned from the success elsewhere?
If an idea doesn’t work, determine whether it’s worth optimizing and trying again, or if you should move on. But don’t get too attached to an idea. If it’s just not working, know when to let go and move on. Speed is, largely, the name of the game.
Weekly Growth Meetings
Every Monday, meet with your growth team to discuss the previous week as well as the upcoming week. You’ll want to cover…
- How many experiments you were able to run last week.
- How many scheduled experiments weren’t run last week.
- New learnings from concluded tests.
- How you’re performing against your OKRs.
- What experiments you’ll run next.
Morgan has created a great weekly growth meeting agenda template that you can steal. Please don’t be the type of growth marketer that meets with her team “only when necessary”.
Trends Over Time
Over the long-term, you’ll want to ask yourself…
- Are your hypotheses getting more accurate?
- Are you getting more wins?
- Are the quality of your insights improving?
- Are you running more tests in a week?
Chart this data over time to see how your growth process is improving (or not). The longer it runs, the more efficient and smarter it should become.
Archive and distribute the learnings.
After reflecting on your experiment, you’ll no doubt record what you’ve learned in the experiment doc. That’s half the battle! Approximately 22% of the people who responded to the 2016 State of Conversion Optimization Report said they don’t archive results at all.
Morgan clearly explains the need for a central experiment doc repository…
Why You Should Archive
A well-maintained, easily accessible, easily searchable archive is important for a few reasons…
- You’ll be able to create clearer, more robust reports for internal distribution.
- You’ll avoid running the same experiment twice or learning the same lessons over and over again. (It happens, especially on high velocity teams.)
- You’ll be able to bring new hires up to speed quickly.
- You’ll run smarter tests in the future.
Note that Brian has a “Systemize” step in his growth process that you saw above…
There are two ways we can systemize. We try to systemize as much as we can with technology and automate things. Certain things we can’t automate. Certainly things in content marketing and stuff just require a human involved, and for those things we write playbooks.
You would automate or create playbooks for the same reasons.
We’ve written an entire article on the art of archiving results, so I’ll leave it at that.
Why You Should Distribute
Distributing your newly acquired insights is equally important as archiving, though. You want to share your learnings company-wide for a few reasons…
- To prove the value of your growth process and get buy-in.
- To ensure your insights can be used by other people in other departments.
There are so many resources out there for presenting data and insights in a persuasive, meaningful way. If you’re presenting A/B test results, I recommend reading this. For everything else, I recommend reading through the lessons here.
This growth process can work for any type of business and any function of marketing as long as you’re purposeful and diligent. And, of course, as long as you’re willing to shake that growth hack addiction.
Here’s what you need to remember…
- Validate your product and business model before launching a growth process.
- Choose a repeatable, data-driven process over throwing tactics at the wall.
- The ROI of your growth process is ideation multiplied by execution.
- Set the goals and the timeframe.
- Come up with experiment ideas.
- Prioritize your experiment ideas.
- Run the experiment.
- Analyze the results.
- Archive and distribute the learnings.