The "forever costs" of product-led growth
Unpopular opinion: Product-led growth (PLG) costs more than advertising, and it’s not even close. Hear me out… Ads cost money, but you can turn them off at any time and only have to pay once, it’s a non-recurring expense. Product changes require engineering, which is your most scarce resource. Plus they bring “forever costs” like ongoing regression testing, support, documentation and code maintenance. Each change adds complexity to your codebase and developer workflows, and complicates your product making it harder for users. The larger your codebase, the longer each subsequent feature takes to build, so it slows you down. How to decide PLG is hot and investors will encourage it. But here are a few question to see if it really makes sense for your startup:
Is your product naturally multi-player? Will customers benefit from inviting others to use it?
Is it a self-serve product? Will new users get to the “a-ha” moment easily?
Does your team have the skills to do it well? Can you do great UX design, front-end dev, and rapid in-product experimentation?
PLG requires tenacity. You may need to run 100 experiments before it really starts clicking. So start lean, use a lot of prototype user-testing and false-door tests to optimize flows before you commit to building them. I’m not opposed to PLG under the right circumstances. However, it’s not a one-size-fits-all magic bullet, and it's not always worth the mythical status it's given. Bottom line: PLG is a very expensive decision, and I hope this email helps you make the right call.
Hat tip to Daphne Tideman (her newsletter) for pushing my thinking on PLG.