What if I told you that we are all, knowingly or unknowingly, part of a huge, constant experiment in almost all parts of your life, relying on your demographic, personal as well behavioral information?

Well, like it or not; we all are. And this constant experiment is called marketing. And I would say it does not only benefit the companies trying to maximise market share and revenue but ultimately us, as consumers, by having more personal, relevant and delightful experiences with the comapnies and brands that we interact with.

A key factor in the success of the biggest companies these days – Netflix, Amazon, TikTok, Google, Apple, Microsoft – is their ability to test often and quickly, and optimise for a better customer experience, higher conversion rates and ultimately more revenue.

But how can you go about these experiments, whether you are a marketer in a big corporations, part of the growth team of a small start-up or a small business owner?

This article explores the experimental nature of marketing, the importance of understanding the buying journey, the role of customer journey analytics and marketing attribution, and the process of designing effective marketing experiments.

But let’s take it step-by-step.

Understanding the Buying Journey

The buying journey is the process that consumers go through before making a purchase. This journey is typically divided into several stages:

Awareness: The consumer becomes aware of a need or problem. To better understand this stage for your customers, look for answers on:
1. How do your customer describe the problem that they are initially facing?
2. Where and how do they look to identify which underlying problem they are facing?
Consideration: The consumer researches potential solutions to their need or problem. Key questions here are: Once identified, where and how do my customers look for possible solutions to solve that problem?
Decision: The consumer evaluates different options and makes a purchase decision. Ask yourself questions like: What are alternative solutions to the problem I solve? Based on which criteria do my customers decide for a solution?

Understanding this journey is crucial for marketers because it allows them to tailor their strategies to meet consumers‘ needs at each stage. By aligning marketing efforts with the buying journey, businesses can enhance customer engagement, increase conversions, and build long-term loyalty.

Understand the Touchpoints your Buyers have with Customer Journey Analytics

Customer journey analytics is the process of tracking and analyzing the interactions customers have with a brand across multiple touchpoints. These touchpoints include online interactions (such as website visits, social media engagements, and email communications) and offline interactions (such as in-store visits and call center interactions).

Customer journey analytics provides valuable insights into how customers move through the buying journey. It helps marketers understand:

Which touchpoints do customers have in their buying journey?
Which different channels contribute to the overall customer experience?
Where customers are dropping off in the journey?

By using customer journey analytics, you as a marketer can understand better which specific steps your buyers take (i.e. which touchoints they have with your company) on their “journey” toward the purchase of your solution throughout the awareness, consideration and decision stage.

Marketing Attribution: Identifying the Touchpoints that Matter

Based on your understanding of your customers’ journey and the touchpoints they have with your company on it, the remaining question is: Which of these touchpoints actually matter and influence the buying decision?

This is where marketing attribution comes into play.

Marketing attribution is the process of assigning credit to different touchpoints that contribute to a conversion. It helps marketers understand which channels and tactics are driving results and which are not.

There are several common attribution models, including:

First-Touch Attribution: Assigns all credit to the first interaction a customer has with a brand.
Last-Touch Attribution: Assigns all credit to the last interaction before conversion.
Multi-Touch Attribution: Distributes credit across multiple touchpoints based on their influence on the conversion.

Understanding marketing attribution is essential because it allows marketers to know which touchpoints (e.g. campaigns, posts, content pieces, marketing emails…) matter. In case you are not familiar with marketing attribution yet, check out this article to learn more.

As a marketer, you need to be in the places that are relevant but you cannot be everywhere all the time.

Hence, marketing attribtion enables you to allocate time, resources and budgets more effectively and focus on the touchpoints that have the greatest impact on conversions.

Experiment and Optimise the Touchpoints that Matter

So far, you should have gained a fair (first) understanding on the touchpoints your customers have with you and which of these touchpoints have an impact for conversion. Now, it is time to act and optimise these touchpoints.

Start with the most relevant touchpoints, or the ones that have the lowest conversion to brace for impact. Let’s explore how to run experiments to optimise.

Designing Effective Marketing Experiments

Designing effective marketing experiments involves a structured approach. Here’s a step-by-step guide to help you get started:

1. Define Your Goal

Before you start an experiment, it’s essential to define a clear and measurable goal. What do you want to achieve? Are you looking to increase website conversions, improve email open rates, or enhance social media engagement? A well-defined goal provides direction and helps you measure the success of your experiment.

For example: A popular e-commerce company wanted to increase its conversion rates, in particular reduce card abandonment in the check-out process.

2. Pick One Variable at a Time

To isolate the impact of your experiment, focus on testing one variable at a time. This could be anything from the color of a call-to-action button to the subject line of an email. By testing one variable, you can clearly see its effect without the results being muddied by other changes.

E.g.: To achieve the desired goal, the marketing team of the e-commerce shop decided to alter the check-out form to simplify the check-out process.

3. Define your hypothesis

Your hypothesis is a prediction of what you think will happen as a result of your experiment – basically the thing you would like to test formed in an assumption challenging the status quo. For example, „Changing the call-to-action button color from blue to red will increase click-through rates by 10%.“ Your hypothesis should be specific and based on informed assumptions.

E.g.: The marketers hypothesised that a single-page check-out form (compared to the current multi-page form) would reduce the card abandonment rate by 20%.

4. Create a Control and a Challenger

In any experiment, you need a control (the initial version) and a challenger (the new version to be tested). The control serves as your baseline, while the challenger is the variation you’re testing to see if it performs better (or worse).

E.g. To run the experiment, the ecommerce team implemented a more streamlined and simpler single-page form (challenger) to be A/B tested against the current multi-page form (control).

5. Split Samples Equally and Randomly

To ensure the validity of your experiment, split your audience into equal and random groups. One group will see the control, and the other will see the challenger. Make sure that both sample groups are similar in terms of demographical and psychographical (e.g. buying behavior). Here, random sampling helps eliminate bias and ensures that any differences in results are due to the variable being tested.

E.g. The team used an A/B testing tool to randomly direct 50% of the users during check-out to the simplified form (challenger) while the remaining 50% used the previous check-out form (control).

6. Determine the Sample Size

The size of your sample can significantly impact the reliability of your experiment results. A sample that is too small may not provide statistically significant results, while a very large sample may be unnecessary and resource-intensive. Use statistical tools or online calculators to determine an appropriate sample size for your experiment.

E.g: The e-commerce shop had several 100s of check-out each day, allowing them to create a sufficient sample size within a one week test period.

7. Run the Test

Once everything is set up, run your experiment. Ensure that it runs long enough to gather sufficient data for analysis. Depending on your goal and the nature of your experiment, this could range from a few days to several weeks. Given that time & seasonality can have an impact on the buying behavior, it is advisable to run both the control and challenger in parallel with A/B or split testing.

E.g: The shop ran the test for 7 days without alteration.

8. Analyze the Results

After your experiment has concluded, analyze the results to see if your hypothesis was correct. Look at the data to determine whether the challenger outperformed the control and by how much. Use statistical analysis to ensure that the results are significant and not due to random chance.

E.g.: The team compared the the conversion rates by looking at how much percent of the users abandoned the cart and how many went through the check-out process in the control and challenger group. The results showed that the single-page form significantly increased conversions.

9. Change, Iterate and Repeat

Based on your findings, make informed changes by implementing the better performing version your status quo. Remember, marketing is an ongoing experiment, so use your changes as the baseline for the next experiment and use the insights you have gaine to plan and conduct further tests to continue optimizing your strategies.

E.g. The team of the e-commerce store permanently implemented the simplified form and ran more experiments (such as adjustments of wording and CTA buttons) to further optimise the conversion rate.

AI generated image of ancient rome with a view fo the colosseum

Real-World Applications of Marketing Experiments

Now that you know the theory, let’s look at some additional real-world applications of marketing experiments to illustrate their impact and to provide you with some inspiration for your testing

1. Email Marketing

An online subscription service aimed to boost its email open rates. They decided to test different subject lines. The control group received emails with the original subject line, while the challenger group received emails with a new, more personalized subject line. The experiment revealed that the personalized subject line had a 15% higher open rate, prompting the company to adopt this approach for future campaigns.

2. Social Media Engagement

A travel agency sought to improve engagement on its social media posts. They experimented with different types of content: static images (control) versus short videos (challenger). By running this experiment, they discovered that videos garnered 25% more engagement, leading them to incorporate more video content into their social media strategy.

The Role of Technology in Marketing Experiments

Technology plays a crucial role in enabling and enhancing marketing experiments. Here are a few tools and technologies that marketers can leverage:

1. A/B Testing Tools

Tools like Optimizely, HubSpot, and Google Optimize allow marketers to easily set up and run A/B tests on their websites, landing pages, and apps. These tools provide intuitive interfaces, robust analytics, and statistical significance calculations.

2. Marketing Automation Platforms

Platforms like HubSpot, Marketo, and Mailchimp enable marketers to automate and test different aspects of their campaigns, from email marketing to lead nurturing. These platforms often include built-in A/B testing capabilities.

3. Analytics and Data Visualization Tools

Tools like Google Analytics, Tableau, HubSpot reports and Power BI help marketers track and analyze data from their experiments. These tools provide insights into customer behavior, campaign performance, and more, allowing for data-driven decision-making.

Conclusion

Marketing is indeed a big and constant experiment. By understanding the buying journey, leveraging customer journey analytics and marketing attribution, and designing effective experiments, you as a marketer can continuously optimize your strategies and drive better results.

For me, the most exciting aspect of looking at marketing that way is that it inherintly allows for failure, fosters constant improvement and removes the pressure of perfectionism. Adopting a mindset of constant experimentation is not just beneficial—it’s essential for long-term success. Remember, even rome was not built in a day and the same applies for marketing. Take small steps but ensure that you never stop.