Skip to main content

That one platform for your successful conversion optimization process

The hypotheses wizard

What’s a hypothesis?

A hypothesis is a scientific verification of an assumption or statement. For example, we make an assumption:

„A user’s transaction is made very complex by too many steps in the order process.“

Our hypothesis is now: By reducing the steps in the order process it will lead the users to fewer drop-outs which is measured by a higher conversion rate.

Therefore a hypothesis is specific and the experiments’ result will either refute or validate it.

screenshot-hypotheses wizard

Why are hypotheses the better alternative than ideas?

For optimizer an unsuccessful experiment normally means that the whole test idea is discarded even though the failure may be caused by the test implementation as well. By framing a correct hypothesis that is independent of the test results it is ensured to get to know the behavior or preferences of the users. The tests following can be run based on that hypothesis.

The free text editor

The free text editor is your tool to frame your hypotheses or ideas in your own style. It is ideally used to keep in mind all the good thoughts that come up in a meeting to either score or enrich them later.

The iridion scoring

Decisions are normally based on logic assumptions as:

  • How high is the traffic on the target page?
  • Which pages show the highest abort rate?
  • What is easy to implement?
  • Does the HIPPO (Highest Paid Person’s Opinion) say what’s going on next?

In reality most decisions are based on either this or similar criteria. But it has been proved to use a scoring system that is as objective as possible and includes all relevant influential variables within optimization measures in order to make established decisions.

Screenshot-scoring

Which parameters are considered by iridion Scoring?

Effort

Backend: How much time is needed for the backend implementation of the experiment?

Frontend: How much time is needed for the frontend implementation of the experiment (incl. the dependence on the testing tool and creating goals)?

Concept: How long does it take to create a good test concept for the experiment (including feedback loops, drafts, process descriptions, etc.)?

Additional: Are there other requirements that generate additional costs (e.g. quality assurance of the technical implementation, internal coordination rounds, political influence of stakeholders, etc.)?

Impact

Visual Contrast: Has there been sufficient change to the design/content in the experiment for the user to notice it?

Behavioral Contrast: Is there sufficient contrast in the variation to change the user’s perception or behavior?

Behavior Patterns: Does the experiment use a behavior pattern which changes the behavior of the users?

Traffic: Is the change part of the main conversion funnel? The closer to the experiment the conversion takes place, the more likely it is to generate validate uplifts (far away from the conversion = e.g. homepage / close to the conversion = e.g. in the checkout funnel).

Other scoring models

With iridion, you’re also able to fully customize your scoring depending on your needs. Or use an existing like PIE.

Stay on the top

You should ask yourself the following questions to steadily maintain your information about your CRO process:

  • What is tested at the moment?
  • What’s the current status of ongoing tests?
  • What tests were successful and which not?
  • What hypotheses/ideas or weak spots were tested?
  • On which pages are recently running which tests?
  • How do I quickly create a good result report?

Well-prepared experiments that are open for consultation by everyone involved are important to make the structure, the current status and the results comprehensible.

iridion helps you to stay on the top of things at any time overlooking all relevant information.

screenshot-experimentlist

The „CRO-Wiki“

Not only the quality and quantity of A/B tests is relevant to ensure continuous progress. The use of already acquired knowledge is a relevant factor for a high likelihood of success and helps to gain an advantage over the competitors.

Why do I need that?

Optimization is always subject to a learning curve and optimization measures can undergo several cycles. If you take care of not losing existing knowledge but make good use of it, hypotheses and test concepts can be better tailored to your own target groups.

It doesn’t matter if there are positive or negative results. Every test contains knowledge that should be kept. Build your own „CRO-Wiki“ with iridion!

screenshot-rollout

All information at a central spot

Iridion ensures to document, analyze and discuss every relevant step of the optimization process to make them sustainable. There is no need to revert to several tools.

What can be replaced by iridion?

  • Excel for preparing and saving test results
  • Powerpoint/Keynote for result reports
  • Dropbox or internal servers for wireframes, concepts, designs, additional interpretations and further attachments
  • Current testing tool for inspecting all experiment-based data and visualizations
  • PM-Tool for coordination and assigning the actual test status
  • Google Docs for collaborative working
  • Internal Wikis as a knowledge store for test concepts