Digital marketing
Run a marketing experiment you can learn from
An experiment needs a question, comparison and decision rule. Changing everything together makes causes difficult to identify.
Put it into practice
- Write one measurable hypothesis.
- Keep conditions consistent where practical.
- Allow enough observations and record limitations.
Start with the part of the customer experience you can observe. Speak to the people handling enquiries, review recent questions and inspect the destination before commissioning more creative. A modest improvement to a real bottleneck is usually easier to evaluate than a broad campaign with several competing objectives.
A practical example
A training provider can compare two benefit explanations with the same audience and destination.
What to measure
Use qualified outcomes and consider uncertainty.
Stopping when a version first looks better can produce false conclusions.
Your next step
Choose one action from the checklist and assign a person to complete it. Record the current situation, the change you intend to make and the evidence you will review. Set a review date that gives the change enough time to produce useful observations. Keep what works, adjust what does not and note anything the data cannot yet explain.