GROWTH HACKING SERIES: Research and Testing

Welcome to another growth hacking series well this article will going to dial into research and testing key important aspects of growth hacking. Growth hacking most times does not come out of the blue and is not a one-off event.

90% of the time growth marketing is a continuous process of iteration, testing drafting hypotheses and measuring the effectiveness of each test. There is a lot of misconceptions on what growth hacking or growth marketing, I advise you or take up CXL’s growth marketing course to get the right orientation.

Most times business owners think it’s just being in an isolated room and coming up with one big idea to achieve Dropbox’s results -1300% growth in 15 months.

But that’s not the case it’s all about the company building a continuous process of testing and iteration. It all starts with all the optimization, which means making the experience of the products and website better, optimising the language to achieve language market fit, optimising the communication and the channels to reach your target audience.

So let’s talk about Testing and Research since it's one of the foundations of growth marketing. A lot of questions on how do we know what to test?.

So how do we know what to optimise how to know what to test that’s where research and testing come into play. An effective research process will help you outline what to test, draft out hypotheses and prioritize those hypotheses on what to test now.

Research and testing

Optimise your optimisation process is done effectively with three objectives

1. Test (or make more effective changes) testing stupid stuff through running out of business
2. Reduce the duration( can cost optimisation)
3. Improve the speed of her experiment.

How to optimise a website objective;

Let’s say you were given a target to increase conversion rates by 20% at the end of the year.

How would you go about it?

Most people go to Google and search for conversion optimisation Tactics or strategies and get lots of tactics and strategies. 2 problems can arise from here.

• You can decide to do all tactics at the same time which will not be effective. I mean you just don't want too many tests at once, it’s a nightmare.
• Or decide to test all the tactics one by one. And if you are in an article with 100 tactics to test from then you are in a long way

The best way to go when developing a new product is to start building with best practices in your industry and optimising from there.

If you have no idea what to test it’s a sign that your optimisation process is completely broken.

Conversion optimisation is a process it is a systematic way of finding opportunities for growth and developing data-based ideas for how to build open these opportunities.

What does a good optimisation process look like?

1. It tells you where are the problems in your website/products(where the leak is the biggest)
2. What are the problems
3. Why is this problem a problem
4. And turn them into hypotheses and what we can do to fix this problem
5. Prioritise test and instant fixes.

Discovery of What Matters

Know what matters to users about your product, what are their pain points, what kind of emotional needs does the product satisfy. When you have this information you can put forth messages that resonate well with your target market.

Highlight features that they care about and remove features that don’t resonate well with them. What actions how your target market doing when they land on your website how are they making a purchasing decisions, who are they consulting with how many different sites they compare before making these decisions, what part of their website actually helps you increase the conversion rate and which part lowers the conversion rate. The Discovery of what matters is the central piece of any optimisation work.

How do we figure out what the real problem is? Two ways

1. Testing: let’s see if this thing is a problem let’s test a different way of the landing page or sign up process and see what works see if the idea makes it better or worst
2. Start with research instead and not just randomly testing to increase the odds of having a high win rate. User research and conversion research has to be the foundation of everything.

Research is what helps us identify where the problems are, what the problems are. You don’t want more data but better data you need data that answer business questions. Data is passive it doesn’t tell you anything it is up to us to interpret the data to make sense of the data.

Start the optimisation process with a list of questions; what do they want? what did they meet? What are the problems? who is their target audience? Your data in your Analytics should answer these business questions which makes being data-driven easier.

“get over analysis paralysis” if you touch your data long enough it will confess to everything — Peep Laja.

The research XL process

The process helps you to gather 6 types of data to help you make great optimisation decisions and come up with tests that tend to win more effective and have a bigger impact.

All about identifying problems and coming up with the best possible ideas to test.

How to figure out what you should be testing and test those ideas.

Step 1: Technical Analysis

The first step in finding out what to test or drafting out hypotheses is by doing Technical analysis. You need to understand what is broken in your site if your site works flawlessly with every single browser and device combination. You can get this info from your web analytics tool. Compare your key metrics — revenue per visitor or conversion rate per browser device.

The easiest way to make money is to fix all technical issues you might have — Peep Laja

Step 2 heuristic analysis

1. Relevance: Google search term 10 Year Show on the landing page
2. Clarity: honest and specifically what the website is about 2 it is for
3. Motivation: what is done on the page to increase motivation to take action
4. Friction: what makes it difficult to take action.

We need to use our experience in the failed to assess what might be working and what might not be working.

Go through your website’s mobile and desktop view separately, page-by-page with a team of people and critique every single screen that we see based on fixed criteria called heuristic analysis.

This analysis is a way to superfast identify possible problems on the website but the results are not scientific but suboptimal.

Simply put, it is a quick way to figure out possible problems that you will now go and gather data for to either validate or invalidate some of the assumptions.

Motivation is an important aspect of this analysis. You should give people specific reasons why they should buy from your website and not competitors websites and it should be on every landing page.

When accessing pages on the website every single page should have one desire action on that page.

Fogg Behaviour Model

The framework of Fogg behaviour model says that desired behaviours happen when these three things occur;

1. There is a high motivation
2. Ability ( how is it is to take action)
3. Trigger/Prompt: when you do you present a call to Action(you should not reveal the call-to-action until their motivation is higher).

Step 3 Digital Analysis

• Where are the leaks
• Which segments
• What are uses doing, which action correlates with higher conversion?

Look at your funnel and find out where are the biggest lakes in this funnel and make different segments (mobile users or desktop users etc).

Most startups are probably doing Data Analytics wrong. When you are not tracking the right information the data is either wrong or missing. Micro-actions are not being measured by startups or companies.

Examples of micro-actions are Scrolls depth, clicks, view rate on the website. We need to understand which actions correlate with high conversions rate then we want to get more people to take that action.

In the next article, we’ll continue the other three stages of the research Excel process.