The psychology of the soft launch

Lauren Macpherson

When you go swimming, do you step tentatively into the shallow and or make a big splash diving into the deep end? These are the same choices you have when you launch a new product.

But even if the food was slow in coming to the table, the chances of your guests getting up and leaving are pretty low. Unfortunately, digital products are not given anywhere near as much grace. 

When these are the statistics you’re facing, a beta test or soft launch is always the best way to go for launching any software, website, platform, or app. Google beta-tested Gmail for 5 years before fully releasing it to the public.

The classic 3 step soft launch model

You need to ensure that your product performs well technically, performs well as a product, and performs well in the market.

  1. Technical testing - you only want to put a fully functioning product into a users hand. This means reaching a threshold for crashing in the 98th to 99th percentile. 
  2. Product testing - this is where you test all the performance assumptions (or hypotheses) you have made. Look at early engagement and retention metrics. Then, does the user return? Have you provided enough value and interest to keep them returning after a week? Then, are they willing to commit to a purchase?
  3. Market testing - this part moves from simply performance testing into the emotional side. Your users need to enjoy using the product, which means it should have some character and create an atmosphere. How does your audience feel about the product? How can you fine-tune how you speak to and interact with them? How can you retain their interest so they continue to be engaged and active users in the long term? This might be about making a compelling enough proposition from the outset, holding their hand through the onboarding process, or keeping them informed about new features and updates.

Technical testing

The very brief psychology of technical testing is that you need to keep making the test more independent and more thorough as you progress through each round. First, the person who wrote the item needs to test it themselves. Then, because you tend not to see mistakes in the work you have done, another programmer in the team should review and test it to catch anything missed. Then, a test engineer should take a look. They will run a much more thorough series of automated and manual tests to check for bugs and push the limits of the code. After this, it’s sometimes worth paying an outside agency to complete a full round of tests too. While most developers know the drill, it’s never nice to make mistakes and it can be frustrating to fix a bug in something you’ve written. By following these steps you allow developers to find and fix mistakes themselves, or within the safe space of their team first. Allowing for a nicer experience for all.

Product testing

Product testing is all about taking what you expect to happen when people use your product and finding out what actually happens. While you will still likely find some technical issues, this is more about performance, usability, user behaviour, and user opinion. These are often closed tests - private and limited to selected testers. Often these are people who have engaged with news about your product or signed up for updates pre-launch. They are usually reliable as they have already indicated they have some prior interest and commitment. If you find you need more testers, you can try the psychology of exclusivity by reducing the number of available spaces or simply incentivise them with an offer like free lifetime access to the product if they help.

First, you need to define the scope of your beta test. Remember, the bigger the scope, the more time and the more money you’ll have to spend. What are your goals - to find out how many users can complete the journey from beginning to end, how many abandon after a week, or how many are willing to pay for it? This will help to align everyone from stakeholders and investors, to designers and developers - as well as set realistic expectations.

Then, you need to design the test:

When running the test, remember, these users have scheduled their life around helping to improve your product. Respect their time, try not to change the schedule, and communicate well. Set expectations from the start, reach out personally if their participation drops off, and have a plan for re-recruitment during a test if you do lose some testers along the way. 

Market testing

Every product has a key moment they are pushing each user towards - for example Uber will have a golden number of rides a user has to take with them before they are considered to be a fully onboarded and converted user. Social media platforms like Twitter, Facebook and Instagram encourage you to add friends and follow other users to increase the amount of time you spend on the platform. They also encourage you to explore new things and interact with them through likes, views and comments so they can gather data about you. More time = more chances to see ads. More data = better targeting and conversion rates for those ads. These together mean more money for them. 

What is important is not to become too focused on this. Because there is also a very important moment the user needs to reach - the moment they find value in your product. Using Uber as an example again, for some it would have been simply the ability to order a taxi in a few taps rather than searching for a company to call. For others, it would have been feeling safer since the journey is tracked, or paying online and not having to find an ATM late at night. 

Marketing tests are usually open betas. Anyone can sign up, which means they are better for quantitative data where you will find your average user and patterns in their behaviour you can use to improve your product more reliably. Open tests are also good for seeing how well your infrastructure scales.

The psychology of a beta tester

The whole point of this is to demonstrate how important soft launches and beta tests are to success. And at the core of this are the testers. So make sure you tailor your soft launch to suit their needs. The most common reasons people beta test are:

Whatever the product, a soft launch is something that should be carefully crafted not just technically but also psychologically. As we build our own ventures often, we’re always running tests. So if you have any questions, feel free to reach out.

More reads

Papyr app: A post-launch post-mortem
Lauren Macpherson

What happens to a product once it’s been launched? Well Jorrit Keijzer, founder of the social reading app Papyr, is here to tell us. We look at what went well, what didn’t go so well and where he has changed things (or completely pivoted) with the latest release.