Companies are constantly needing to experiment to determine how to get the best results for the business. Growth and Product teams have experimentation as key parts of their roles and there are a host of solutions on the market to help with things like A/B testing.
Experimentation helps companies optimize an outcome, generally an outcome that supports revenue growth in some fashion. With experimentation, you’re able to fine tune your product, sign-up flow, and on-boarding.
Most companies are running between 15 and 30 experiments each year, and well staffed companies are potentially in the hundreds of experiments each year. Experimentation is expensive, but the outsized impact it may have can outweigh the cost.
What experiments are being run?
The most common experiments being run relate to:
- Getting more people signing up
- Increasing product usage
- Decreasing time to value (TTV)
- Improving retention/lowering churn
- Improving paid conversion rates
To determine which experiments to run, organizations rely on:
- Statistical correlation, which never tells the whole story
- Some are building their own models, but that requires significant investment and expertise
This chart from Baremetrics shows the typical TTV experience and provides an impact visual on why getting to TTV sooner is important, and why teams are running those experiments.
What does experimentation cost?
It can be hard to pinpoint an exact cost of experimentation because it can change by industry, like B2B SaaS vs eCommerce, and the size of an organization. At a base level, it’s fair to assume you have at least three roles involved in the experiments, including:
- Growth Manager (the exact role here can change based on the organization)
- Product Manager
- Developer/Front-end Engineer
Some organizations will have additional roles involved, including:
- Data Scientist
- UI/UX Designer
- Project Manager
Beyond that, you have the cost of the SaaS products that help manage the experiments, which can range from $500-$10,000/m.
This doesn’t factor in the cost of lost revenue (or lost opportunity) as you perform experiments that don’t lead to the desired results.
What if you could run fewer experiments to reach the desired result?
Imagine a world where you get some real direction on the experiments you should run so that you get to your valued outcome faster. What if you had the power of prediction to help you with that?
- You’d know what the most impactful predictors are. These are the levers that you should pull to see change.
- You could model scenarios to predict the impact of proposed changes before actually making the changes.
- You’d run more impactful and informed experiments.
- You’d need to run fewer experiments.
- You’d need fewer engineering resources
- You wouldn’t need dedicated data scientists to help you
- You could recognize revenue sooner, driving faster growth, and wider adoption
With Dacture, this could be your reality. You can save money and see better growth using the data you’re already collecting.