I've rebuilt aside more than once, and the last rewrite is the one that changed the most. Before I say what I changed, the first thing: I tried to sell aside and I couldn't. I'm not saying it with drama, it's what happened.
What I tried first
Between March and May I was posting about aside on LinkedIn. I saw no results and I stopped.
How it was before
The first version "saw" the screen and decided on the spot where to click or what to highlight. It failed too much. When something fails because of how it's set up, tuning doesn't fix it.
What changed
Now it's your app that tells aside what it can do: create an invoice, change a setting, open a screen. aside chooses what to do and with what data, and your own code runs it.
You tell it "invoice Orbit Labs €2,400 for the month of September" and it leaves the invoice filled in, ready for you to send. What changes is who decides what is possible: your app, not aside guessing.
What I removed
Before, you could also talk to it by voice. I realised it caused problems, it was very expensive and it wasn't essential. I removed it to focus on making what it does work well every time.
The price
Before, it started at €249, too much for solo founders and small startups. Now it starts at €49.
Before and now
| Before | Now | |
|---|---|---|
| How it decides what to do | It saw the screen and decided where to click | The app tells it what it can do, aside picks and the SaaS's own code runs it |
| Voice | You could talk to it | I removed it |
| Price | From €249 | From €49 |
What it does today
Two things. The first is onboarding: it takes your users through setup step by step, in a guided and natural way, and fills in each field while they talk. The second is support: it answers with your documentation and, when that isn't enough, it acts.
aside's own dashboard runs on aside: you tell it in one sentence how you want your agent and it sets it up for you.
All of this came from talking to customers and seeing where it failed. I don't know if it's the final version, but it is the most polished one.
So far, three conversations
I've talked to three leads. That's few, and I'm not going to draw market conclusions from three people. What I'm telling you is what happened.
One told me the price was expensive. The other two never replied.
About the one who said it was expensive I know something concrete: for that person, the price is an objection. It's a data point, even if it's only one.
About the two who disappeared I know nothing. A silence is not a no, and it's not a yes either. It doesn't tell me whether they weren't interested in the product, whether the timing was bad or whether my message didn't manage to say anything to them.
With three conversations there is no pattern yet.
Top comments (0)