How to Define and Evolve Your Company's Messaging
The lessons from Lovable, Granola and ElevenLabs.

Founders spend a lot of time trying to fit their company’s explanation into a single sentence. The most difficult part is figuring out what that sentence should explain.
I’m always interested in seeing how companies evolve their messaging as they find PMF and scale. Sometimes, the message becomes broader. Often, it gets much narrower as you figure out your target audience and hypergrowth starts.
Let’s look at Granola. In 2024, their website called it a “context-aware workspace of the future." The particular problem I see with this definition is it can be applied to many different tools. It’s not surprising that later Granola radically simplified it and called itself “The AI notepad for people in back-to-back meetings”. See the difference? Now it’s concrete and can’t be applied to another product. Back then being AI also meant something, while founders, executives and managers in tech immediately see it’s tailored to how they work, with calendars filled 9-5 (or more).
There’s a subtle difference between a definition and a tagline. A definition is what’s attached to your name. A tagline explains the product but might not work outside of its context. Lovable’s early tagline was “Idea to app in seconds.” It was perfect. Highlights the speed, pushes you to be creative. Notice how it says “app” instead of “website”: creating websites quickly and without coding was possible for years, but it wasn’t true for actual apps. Below that tagline, they wrote “Lovable is your superhuman full stack engineer” which reinforced the message by explaining their role. Unlike ChatGPT or Claude, Lovable handled deployments, databases, auth, and everything else.
You can broadly categorize the way companies describe their products in three groups:
- Action or result
- Tool or format
- Capability or aspiration
Lovable’s “Idea to app in seconds” falls neatly into the first one. Granola clearly describes itself as a particular tool. If you want an example of the third one, what comes to mind is ElevenLabs’s early description as “the most realistic and versatile AI speech software, ever.”
When you combine taglines and descriptions, you realize these categories aren’t mutually exclusive. Granola combines a tool, a user’s condition, and a benefit. Lovable joins a result and speed promise with a role. ElevenLabs names a category, then claims its quality is radically better.
***
Your definition can’t stay the same forever. The product changes and the environment changes as well. When you have an innovative product, you first have to explain it. Then, as the category emerges and grows, competitors emerge, so you often switch from explanation to differentiation.Let’s get back to Granola and the sequence of their descriptions from 2024 to 2026:
- A context-aware workspace of the future (too broad) — Feb 2024
- The AI notepad for people in back-to-back meetings (found their PMF) — Jun 2024
- The AI notepad for teams in back-to-back meetings (expanded to teams) — Jun 2025
- The AI notepad for people in back-to-back meetings (returns to the initial idea) — Oct 2025
- The AI notepad that helps you get the doing done (focuses on why you’re making notes at all) — Mar 2026
- The AI notepad for back-to-back meetings (the current one) — Nov 2026
You can see how Granola found its PMF, figured out the way to express this, then adjusted to accommodate slightly different customers especially as more and more people learned about the product.
Lovable is even more interesting, since it’s been relaunched several times, found the PMF, then scaled to serve vastly different needs of their users while competing with both rivals and AI labs themselves. Let’s look at both lines they had on their website:
- Rapid prototyping – with speed you haven’t seen before (Jan 2024)
Build and iterate on interactive web apps using natural language - Build for the web 10x faster (Sep 2024)
Chat with AI to build web apps. Sync with GitHub. One-click deploy. - The last piece of software (Nov 2024)
We’re building software that builds software. - Idea to app in seconds. (Dec 2024)
Lovable is your superhuman full stack engineer. - Build something Lovable (Jun 2025)
Idea to app in seconds, with your personal full stack engineer - Build something Lovable (Oct 2025)
Create apps and websites by chatting with AI - Build something Lovable (Nov 2026)
Bring a new product, internal tool, or entire company to life.
You can clearly see their evolution here. First, the models weren’t great, so only developers could meaningfully use it and deal with bugs. As they moved from GPT Engineer to Lovable, the definition became broader and more focused on building something instead of software development as it is. Lovable keeps expanding their promised outcome as AI capabilities grow. Now Lovable is meant for everyone to build anything. Such a description wouldn’t work for them in 2024 but might be the best option right now, since it’s already a household name.
This is context. Your definition and tagline exist in the context of the market, your competitors, technology capabilities, and other things. As the context keeps evolving, your definition must be updated as well.
In 2023, ElevenLabs started using “the most realistic and versatile AI speech software, ever.” By 2025, they told you to “create the most realistic speech with our AI audio platform.” And now it’s just “Bringing technology to life” as they offer three very different products.
I see a bit a problem in that it tells you very little about what you get. I understand ElevenLabs now has more products to explain, and their visitors already know something about the company. But this is why copying the homepage of a successful company can lead you in the wrong direction. A line that works for them can leave your potential customers wondering.
Still, I believe there could be better options. Something like “Give your products a voice”. What do you think is the better alternative for them?
***
How do you figure out your context? You can start by answering these 6 questions:1. What happened? The conditions that make someone look for a solution, especially now.
2. What must be improved? What result do they need? It could be a measurable improvement or a change in how the work feels: less anxiety about missing something, more confidence walking into a meeting.
3. What do they do today? Is there a tool synonymous with the workflow you’re replacing? Maybe a big company? Or is it just boring manual work? Maybe people aren’t even doing this because it’s too cumbersome?
4. What is your product? Is there a new category, an object, or maybe a role that explains what you offer?
5. How does it change the work? What gets better? What constraints are removed? How much faster and easier does it get?
6. Why choose it? Why is your approach better than what they do today? What would make someone switch, and what would make them comfortable trusting you?
The answers give you the ideas, but you have to decide what goes first.
Let’s finish with Granola. It wasn’t the first notetaker: tools like Fireflies and Otter existed but their complete transcripts weren’t too useful. Granola focused on the very specific audience of people working in tech companies who often have their entire days filled with meetings: it’s a very familiar situation to them. They already take notes, but doing it well competes with paying attention. Granola starts with the notes you write yourself, then uses the transcript to fill in the gaps, so it learns what you care about.
That was almost a paragraph, so we need to condense it. Granola chose “The AI notepad for people in back-to-back meetings.” Why?
“Notepad” explains what it is. “Back-to-back meetings” lets workers recognize their own situation (immediate familiarity). The next sentence tells you what happens to your notes: “Granola takes your raw meeting notes and makes them awesome.”
So try a few openings using your own answers. Start one with the product, another with the result, and another with the capability that makes it possible. Then ask which one gives your intended customer the best possible reason to sign up for the product.
A good test is putting another product’s name instead of your own. Your definition should fail: if it doesn’t, it was likely too broad.
Once competitors catch up, try putting one of their names above your description. Does it fit? Too bad, now you should explain what makes you truly different.
Get the next edition
Subscribe for new editions of How They Became Known, featuring deep dives into how technology companies gain attention and establish legitimacy online.




