Growth startup (3/12): running growth experiments with research and testing review

The first two posts of this series of growth marketing, based on my learnings from CXL’s Growth Marketing minidegree, were focused on growth marketing foundations.

It's a useless effort trying to grow if you don’t have a good foundation in growth marketing.

GROWTH STARTUP SERIES PREVIOUS POSTS:

Growth startup (1/12): open-minded growth vs. traditional marketing review

Growth startup (2/12): building a customer-centric growth process review

These previous posts are very important because they prepare us for what’s coming next: running growth experiments. This is where things get more stimulating - and usually more frightening for most companies. 

Let's move on?

A note before we take off

We depend on our visitors and customers to grow. Let’s always keep that in mind. All the efforts we take must be focused on them, not just to guide them to take action, but also for their optimal experience.

This is mandatory for them to decide to invest (not spend money) in our services/products/brands. If they don’t feel comfortable, confident and rewarded, it will be very hard to improve growth for our services/products/brands.

Having that in mind, let's ask ourselves:

Do we know our channels?

or

Do we know how to recognize the value of our channels?

We can say that there are 5 great channels with a lot of opportunities for growth strategies:

• SEM
• SEO
• Email
• Social and display ads
• Content marketing (forms, LPs etc.)

It's important to identify one to three channels that are working for you. You don't have time and resources to spend on all channels, then select the channel(s) where:

• Your audience is interacting and engaging more.
• There is a clear mismatch with your brand.
• You can be more competitive (if you are prepared to fight for organic traffic, let ads for another moment, for example).

The power of meaning and motivation

The idea is to find a good balance between ROI and scale. And if you see that your data is not good (low ROI or weak engagement), maybe it is better to let the channel (instead of keeping that because it is popular).

And if you feel thrilled to test emerging channels (VR, AR etc.) try to check what other companies are doing and start testing.

Social media marketing, for example, must generate demand and drive leads. But if it is only generating likes and comments (many still depend on them for objectives and goals), maybe it is not your most strategic channel for the moment.

We will talk in future posts about something very, very important for visitors and customers: motivation... But before we deep into that, watch the video below:

Running growth experiments

Growth does not arrive attracted by luck, assumptions, bets or feelings. You should already know this. And what you should also know is that it all depends on process, discipline and analysis.

Let's try to put some logic on that considering two main things: research and testing.

Research and testing are linked to the same objective: optimize for better results (and growth). And the probabilities of failure are high if you don't have a process. Ok, now you must be asking yourself...

What would be an optimization process?

There is a lot of content on that. Charts if you prefer. Or infographics. You pick what's is best for you. Here, we'll think of some questions you should answer when you find yourself trying to establish (or improve) an optimization process:

• Where are the problems?
• What are the problems?
• Why is this or that a problem?
• Turn know issues into test hypotheses.
• Prioritize.

But what to measure to understand if you have a good optimization process?

• Test more effective changes (what really matters for the strategy to evolve).
• Reduce the duration of optimization.
• Experiment faster.

"(...) conversion optimization is a process. It's a systematic way of finding opportunities for growth and developing data-based ideas for how to build upon those opportunities."

Peep Laja, CXL fouder

With a process, it is easier to discover what the real problems are. And to solve them, the path is:

Test

+

Research

But what to consider when starting to research for problems to solve?

• Technical analysis
• Bird view analysis
• Digital analytics
• Heat maps and form analytics
• Strategic surveys
• User testing

Once you have clues about what's going wrong, the challenge is to choose the thing that really matters for your growth to happen.

• Don't test what doesn't matter
• Avoid the temptation to copy someone else's tests
• Give your tests the time they need (business cycles, for example)
• Compare results with different/complementary analytics tools
• Don't give up if your hypothesis fails
• Don't ignore small gains
• Keep testing a routine

Intro into conversion research

We need to gather data that we can read and analyze, pull in this other data, and then say, Okay, I think I know what we should do.

And to get a better sense of this data, it is necessary to understand the user's experience on your website, blog, LP etc.

What are the distractions?

How much friction do they face?

There are some other elements you might wanna research, such as 5 second test, technical issues, analytics, mouse tracking, user testing etc.

For example: nowadays we're all multi-device users and sometimes websites, blogs and LPs are not responsive to tablets, desktops or mobile phones (or even different browsers).

Don't you think that issues with responsiviness on different browsers and devise are also a conversion problems?

You'll find different issues on usability and conversion that you have to prioritize. Don't distract yourself: you better start in the bottle by fixing things that keep money away. 

Categorize and prioritize

Categorizing the issues you found is a good start, but just the beginning.

The next movement is to find a way (a better way for you) to prioritize them by two possible criteria:

• ease of implementation,
• opportunity (what is gonna change for users and generate sales), taking into consideration the complexity, impact, time spent to fix etc.

And when you start doing your tests, don't waste time: you'll need to find out if your process is effective.

For this, pay attention to:

• testing velocity,
• the number of tests that make a difference,
• the impact of each successful test.

That's it for today... See you next Sunday. ;) <3

To be continued: optimization process frameworks, A/B testing mastery and statistics.