More and more companies are investing in data-driven working and the agile way of working. Despite many similarities between the two methods, they are often still used separately within companies. As a result, companies miss opportunities to grow faster and better meet the needs of their customers.

How do you combine both approaches and integrate them into your organization?

What are CRO and scrum anyway and why combine them?

To begin, let's look at what both methods entail and what similarities they have. 

Agile is a philosophy and way of thinking made up of a set of values and principles. In general, agile stands for an iterative approach to project management and software development that helps teams deliver value to their customers faster. Instead of putting everything into a big launch, agile teams work in small product updates. These updates are released as quickly as possible. With each release, a team learns what works and what doesn't based on the response of their users.

Scrum is an agile framework for developing and delivering products. A scrum team normally consists of a product owner, scrum master and development team. A scrum team works in so-called sprints. A sprint is a period of one to four weeks in which a potentially useful product is delivered.

Inside Conversion Rate Optimization (CRO) works in a similar way. A list of substantiated ideas is created, based on thorough research. These hypotheses are validated using experiments (A/B testing). For CRO teams, it is essential to develop and execute new experiments at a rapid pace. The more ideas that are validated, the greater the success of the CRO team.

Within both scrum and CRO, work is done on the basis of iterations that follow each other in rapid succession, and learning and improvement are central. With CRO, you learn, iteratively, about your visitors and improve your product. With scrum you build, also iteratively, a usable and tested product.

To achieve faster growth and better alignment with customer needs, companies should combine CRO and scrum. This allows them to improve the correct product(updates) fast deliver. 

Challenges in combining CRO and scrum

Despite all these similarities, it can be a challenge to combine the two frameworks properly. In fact, there are some differences as well.

First, a scrum team and a CRO team work at different paces. A scrum team works in sprints of one to four weeks, while the duration of experiments by the CRO team depends on the minimal detectable effect (the minimum relative improvement that can be detected in an experiment based on the number of visitors and conversions).

Second, the scrum method focuses on executing and delivering a project. If the necessary knowledge is lacking, running experiments can be seen as extra work that comes at the expense of project speed. Also, the CRO team's experiments will frequently show that work delivered by the scrum team does not produce improvement. As a result, the scrum team will have to accept that some of the work delivered will ultimately not be used.

Third, CRO is a continuous process. The outcome of one experiment is the beginning of the next. On the contrary, the scrum process works on a project basis. When one sprint ends, the next sprint is already planned. The outcome of one sprint does not affect the next sprint.

Fourth, the position of the CRO team within the organization can make it impossible to collaborate with scrum teams. We often see that CRO is part of a marketing silo, while scrum is usually used in development. This makes it difficult to properly align work.

Three solutions

There are three solutions used to ensure that the two methods reinforce each other. The first solution, dual-track agile, is increasingly being used. The second solution: making validation part of the work of scrum teams, is harder to achieve because it requires changing existing processes and mindsets. The third option is to combine both solutions.

1. Dual-track agile

With dual-track agile working product teams cross-functional and multidisciplinary. The work is divided into two tracks: discovery and delivery.

In the discovery track, hypotheses are created based on research and validated through experimentation. The work focuses on rapid learning and generating validated product ideas to fill the development backlog. 

In the delivery track, the scrum team continues to work the same way it is used to, with the same roles and in sprints.

The discovery track can learn from each new product update delivered by the delivery track and validates new ideas. 

The key to success is close cooperation between the two tracks. All team members must be aware of what is happening in both tracks. In most organizations, the product owner will play an important role in this.

In the ideal situation, the CRO team is part of the discovery track, working on key experiments such as product and feature tests. This makes the organization more results-driven rather than output-driven. 

The advantage of dual-track agile, with CRO as part of the discovery track, is that scrum and CRO teams can continue to work at their own pace and work processes. It also increases efficiency because scrum teams will need fewer iterations before the product (updates) are properly delivered. This results in faster release cycles, saving costs.

In practice, we see that CRO teams often work in the delivery track, where they work on relatively minor changes. Discovery work is also performed, but it is focused on existing products, features and Web pages. Decisions about major product and feature changes are made elsewhere in the company.

2. Make validation part of the work of scrum teams

Making validation part of the work of scrum teams requires overcoming a number of challenges. But when successful and properly implemented, this solution can have a huge positive impact on business results.

There are four ways to implement this solution:

1. Non-binding aggregation.

In this situation, the product owner and scrum team have the mandate to conduct experiments within sprints. They must be allocated sufficient time and resources to do so. 

The scrum team will often aim to perform one or more experiments per sprint.

In practice, this way of working rarely solves the aforementioned challenges. In many cases it will slow down rather than speed up CRO activities. Therefore, it is difficult to achieve good results with this way of working, but if everyone is well aligned then it is possible.

2. Making experimentation part of assessments

If experimentation and validation is part of the assessment of product owners and scrum teams, then validating ideas is no longer optional. 

If this applies only to some team members but not to their managers, then a situation can arise where different interests cause the validation process to slow down. This happens, for example, if the manager is judged by output and therefore works more output-driven rather than result-driven. 

When validation is also part of the assessment of (senior) management, successful collaboration will occur because everyone has the same interest in experimenting and validating.

3. Experimentation as part of the ‘definition of done’

Within the agile way of working, the definition of done the acceptance criterion. It is a checklist of things that must be completed by the scrum team before a project may be considered complete. 

If experimentation is part of the definition of done, then the scrum team must conduct experiments. After all, without an experiment, the project cannot be completed. In this situation, everything that is released is validated with one or more experiments.

4. Automated experimentation

The fourth way to implement validation in scrum is to automate it. For scrum teams in organizations where this way is used, it is impossible to put new features live without experimentation. 

Like any other scrum team, they can make product or website updates. Only these cannot be released directly to 100% of users. Instead, each release is automatically put live as an experiment, where some of the users see the original product/website and some see the updated version. 

If the experiment is successful, the update will be visible to 100% of users.

3. Dual-track agile combine with automated experimentation

The third (and presumably best) option is the solution we often see at big tech companies. 

Experimentation is automated for the delivery track, so every update built by the scrum teams is released as an experiment.

At the same time, extensive research is done in a discovery track. As in dual-track agile, the development backlog is supplemented with ideas for new features and product updates based on this research. 

This solution combines the best of both worlds. Organizations that implement this solution achieve rapid growth and better alignment with their customers' needs.

Quickly deliver the right products while meeting customer needs

In summary, while there are some challenges to overcome, scrum and CRO ultimately pursue the same goals. To achieve the best results, companies should combine both approaches.

This does not happen by itself. It is a long-term process that requires patience and perseverance. Together with like-minded colleagues, you can change the culture step by step. By doing this, the organization begins to evolve faster and faster and become more and more responsive to the needs of its customers.

Want to know more? We hosted an event on this topic. Now read the summary.