Abhi Vaidyanatha

2 hours ago

The Convex Marketing Journey 2022-2026

We haven't "made it" yet, but it's important to be self-reflective when figuring out how to present yourself. In honor of us being a little better at marketing than we were last year, I'm taking a trip down memory lane going over marketing ideas that we tried. We honestly haven't figured this out perfectly, but we're constantly in pursuit of convergence toward a limit that exists and is actually useful.

A lot of stuff worked and a lot of stuff didn't, so I thought it would be useful to share. Also, every evolution is made with knowledge that we didn't have at the time; just because we made unoptimal decisions in the past doesn't mean we could have done it any differently to get here.

States (of) Matter

Before I joined the team, there were four main primitive messages, all with their various benefits and drawbacks.

The Reactive Backend for your Reactive App

This one is... clear, but maybe too clear! We always want reactivity to be front and center, but this is super narrow.

The Global State Management Platform

James and Jamie will tell you that this is still the perfect summation of Convex, but they know that we can't call ourselves this yet. It's ahead of its time and perfectly encompasses what we do. We'll come back to this some day when we are big enough to call ourselves whatever we want.

The Front end just went full-stack

Writing your backend like you write your frontend is one of our core ideas, and this is a representation of that. Outside of being a little cheesy, I personally think this failed because both frontend and fullstack are phrases that don't know whether they want to be hyphenated or not.

Build products, not backends

While this is accurate, and we do say stuff like this every now and then, it's just too generic. So many people over the years have made various broad brush strokes in this direction and it glazes over the eyes. I do, however, like phrases that indicate that we have made the backend choices and engineering for them.

Naturally, there was less usage and feedback at the start, but the founders and early employees figured out the following important concepts that we needed to focus on:

• Backend architecture is hard and we've made it easier
• Switching between writing Convex code and writing frontend should feel seamless
• Reactivity is really easy to show off, so it should be front and center
• Using a phrase like "State Management" helps contextualize the fact that we solve a large class of problems, not just one problem

Lesson 1: Let your pillars of truth inform your messaging, but don't overindex on them.

Type Shit

I joined two and a half years ago right before our 1.0 launch to work on developer experience and education. My role was to figure out how to message the launch, create a video for it, and figure out what themes were going to stick with people. The team (and me) were convinced that types were the key to it all; Convex helps your app enforce strong typing, a perfect marriage with TypeScript.

It's funny looking back at the video I created and seeing how specifically we leaned into type safety and type propagation, but I'm glad that we tried it out. It's very easy to look back and think you made mistakes with messaging and positioning, rather, you should think of every version as required to get to where you are right now.

The good

Good design plays off of things that people recognize. People recognized that TypeScript was one of the keys to building a maintanable app, so attaching on to that brand was a good way to teach people what we were about.

The bad

It wasn't really the reason that would get someone to use Convex; if someone wanted to start building on TypeScript, the brand association was too strong with T3 Stack to really evaluate using anything else.

Lesson 2: Associating with a product or group only works if the group is ready to receive something.

Lesson 2.5: Think of your journey in marketing messaging as an identity exploration rather than a bunch of mistakes that you made to get to where you're at now.

The Only Decision?

The messaging pendulum oscillates between an abstraction and a feature dump; the marketing brain wants to list every feature to wow potential users and yet also wants to convey everything in a simple phrase. This was our turn to do a simple phrase that conveyed how we have made an opinionated platform that makes your job easier.

Who is your marketing for?

When you write messaging, always ask this question. It's so easy to write "overtrained" messaging that was birthed from individuals that are already immersed in your ecosystem, then refined by more people within your company who also understand your product too deeply.

The statement was quite generic, but it did resonate strongly with some people. I'd get comments from people that already loved us who told me that the message was perfect for them. They understood that using Convex was akin to standing on the shoulders of infrastructure veterans who made strong decisions for them, leading to a performant backend. However, those who were not already in the know didn't understand how to take it.

This is where I made the fundamental realization that your most forward facing messaging is not necessarily what your diehards tell you; the best message is the one that got them to give you a try, which they often forget.

Lesson 3: Always ask who your marketing is for, and don't overtrain on people that already get it.

Interactive Exhibit

Back when humans still wrote code, we had a hypothesis that people needed to tangibly understand what it felt like to write Convex code, which led us to creating an interactive experience on the website. This experience still exists to this day, as we still find it relevant, however the motivation behind it has completely changed.

Additionally, we were given some advice by none other than Theo himself to brand ourselves as the "Missing half of your React app." The cool part about this was identifying that it was totally okay to narrow our scope down to one framework, because React takes up such a large percentage of web app composition. It also elegantly expresses a pillar that we defined earlier in the first branding exercise:

Switching between writing Convex code and writing frontend should feel seamless

Overt subtlety

I think this message rocks, but it was both too subtle and overt at the same time. When people interpret it literally, it's a bit of a nothing-burger and sounds like a throwaway statement. If people are in the know, they get it... but those are the people we don't need to market to!

Lesson 4: Messages can mean multiple things to different people, but needs to actively have meaning to each group.

I omitted one time period from this article where we used the tagline, "The backend for AI." While I would prefer to erase this from history, I will admit it here for posterity because it's important to remember if you've been a trend-follower. There was a weird midpoint where this messaging made sense, but it very quickly became way too obvious. To be clear, we actually do believe that Convex is the best backend to use with AI programming tools, but the message just won't hit with everyone trying to say the same thing.

(i)NSYNC

Sync is the glue word that describes everything that we are. We are realtime, we have a sync engine that reconciles state, and we have some of the energy that local-first development is trying to capture. I'm not too attached to the other words in the tagline, but backend platform suffices right now to describe what we are clearly.

Perhaps the most obvious change was the color palette change, which moved us away from the colors on the Convex "ball." I love the evolution of the brand colors to be warm and the olive green feels inviting. The initial brand colors are difficult to work with and we found it easier to convey the feelings we were targeting with less intense colors.

Work in Progress

There's still a lot we haven't communicated well, as I think our main value proposition now is the ability for people to build apps that can immediately scale and be modularly iterated on with AI tools. Our homepage doesn't do a great job of communicating this, but our meetups, YouTube content, and social presence does communicate it. I look forward to writing the same post three years from now.

Lesson 5: Changing your messaging probably won't change the trajectory of your business, but keep it aligned to who you're selling to.