ai

You Don’t Know Everything - Two Tactics to Support Any Product Launch

Don't just release a build - it's equally important to measure and test.


Don't just release a build - it's equally important to measure and test.

Even though we have done a ton of work to customize the way our product interfaces with users - learned the best way to reach users to initiate the conversations, convinced users to care about the conversation once engaged - we still don’t know everything that will work (and not work) for our target users. It is easy to think you can push the big red “release” button and go get a beer once you learn this information – but you’re really just getting started.  You must test, test, test.  There are so many things that have to go right for you to gather the optimal data, that it is likely a few things will go wrong. Here’s how to avoid common missteps: 

  • Establish success metrics and collect the data.
    We instrument our systems to know if users are reading our marketing and clicking on the link to get to our conversational interface, and once they do, we measure how far into the conversation they get.  This lets us know where in the user experience journey we are failing so we can fine tune for larger audiences.

  • Deploy iteratively. 
    We deploy the interface with a small, trusted group of users first, and we read the responses from every user to highlight questions that are confusing, gathering the wrong data, or taking too long to answer. 
    We interview each of these users to ensure the entire UX meets the mark. This may seem like a redundant step - but it is necessary. Through these interviews, we usually uncover a goldmine of information regarding we get the added benefit of happy users given the white-glove service they received. 
    We then ask all of these users to share the link to the conversational interface to their friends and post it wherever they can. This starts a grassroots campaign that leads to the ever-important viral growth.
    After we iterate with what we learned from these conversations, we deploy again with 100-200 users and focus more heavily on the analytics to ensure we’re getting rich contextual data. This is answered by running the data gathered through our algorithms to see what sense is made of it.  The changes at this stage of testing tend to be minor tweaks related to the wording of a question or the marketing messaging or channels we’ve selected. 

Once you’re confident that your UX and analytics are aligned, you’re ready to gather great data with a full-scale launch.

Similar posts

Get notified on new marketing insights

Be the first to know about new B2B SaaS Marketing insights to build or refine your marketing function with the tools and knowledge of today’s industry.