Wilson Wilson

One of the biggest mistakes I first made when I started building https://senja.io a year ago was immediately building every feature request I got. At the time, I was desperate for customers. So once I got a feature request, I would get started on it immediately and release it in a matter of hours to impress my users. It took me a while to realize it, but by doing this I started building a product that had no real focus. I would build features that didn't align with my product vision. Users would also make feature requests that often conflicted with each other. > To make matters worse, while I did impress the users I got, none of the users I did this for upgraded To regain control, and to build a product that my ideal users would love, I started democratizing feedback collection. I created a feature requests page where users could submit ideas for features they wanted, and other users could upvote on them. I'm using canny.io for this, but there are many feedback tools I know of, like upvoty.com and hellonext.co. I chose Canny because it's completely free to get started with. I added the link to the feedback page to my marketing site's menu and my app's dashboard, and slowly but surely, feedback from genuinely interested users started rolling in. When a user requests a feature in our support, we redirect them to our feedback page so that other users can vote on it. By doing this: - we now know which features our users want the most. - knowing what to build next is way easier - our users can also discuss with each other about how they want to use features.

When we started building https://senja.io, I had the brilliant idea of coding our marketing website from scratch. I've always preferred having control over every element on a site. Building the site myself would mean having unlimited flexibility concerning design, performance, SEO etc. My only limitation would be my skillset. The moment I got a marketing cofounder though, everything went downhill very quickly. He handles marketing, so he needed the power to control the site's structure, copy and design. If he wanted to update a little text on the website, he'd have to message me. Then I'd drop what I'm doing, and make the update. He couldn't create pages or blog posts on his own. He'd need to write the copy, send it to me and I'd build the page. As his requests increased, things became a drag very quickly. I thought, no problem! I was still hell-bent on coding the website myself, so I decided I'd build with the Jamstack > Jamstack is an architectural approach that decouples the web experience layer from data and business logic. To simplify, we have a database with all our website content in a friendly UI, and the marketing website would just pull from that. That way, my cofounder could update the site without touching the code. By integrating a CMS, he would have control over all the site's content and there'd be much less back and forth. Even though it helped for a while, **this didn't solve the problem.** Building a rapidly changing site with the Jamstack has been nothing but a pain. - I have to reinvent the wheel for literally _everything_. Basic SEO, animations, performance, structure, navigation, image optimisation etc. - Building new things is just horrible. I have to constantly worry about backwards compatibility. I also have to create + maintain new components for the smallest additions. Rebuilding our landing (https://senja.io/blog/should-i-c...) page has proven to me beyond a doubt that coding a marketing site from scratch is just a really bad idea. At first I thought it was cool. Now it's just a pain and is keeping me from doing the thing I want to be doing. Actually improving my product. For a scrappy MVP, it might be worth it. But once things start to scale you'll get into trouble _very_ quickly. Just use Webflow/Framer/whatever floats your boat. You'll save so much time in the long run.