Ep. #2: Kiedi.me is live
Where Kiedi comes from
In Ep. #1 I wrote that Kiedi would be out in a few days. Here it is: kiedi.me.
Kiedi didn’t come from a client request. It came from wanting Nube to be more than a software house building for others: a place where we create our own products and sell them.
I started from what I know best: my clients’ brochure websites. A brochure site gets built, goes live and then just sits there. The client pays once and the relationship ends, more or less. A chatbot that answers visitors raises the average deal size and makes the site a little more self-sufficient.
I built the first prototype over a weekend, partly to get my hands dirty with AI and vector databases.
What it does
Kiedi is an AI chatbot that learns from your website and from the documents you upload to your account. It answers visitors right away and helps them when they can’t find what they’re looking for.
What sets it apart from the other chatbots on the market? Honestly, nothing. Other than the fact that I built it.
And that’s exactly why the hard part isn’t building it. It’s getting it found.
How I’m bringing it to market
Data first, then content
Before writing a single page, I did keyword research on the Italian market to see where it’s worth showing up.
The searches closest to a purchase, like chatbot per sito web (chatbot for websites) or chatbot per hotel, are niche but valuable: on Google Ads a click costs between $18 and $27. The people searching them already know what they want, and that’s where I want to be found.
Traffic, on the other hand, comes from informational searches. Chatbot cos’è (what is a chatbot) and its variants get over 8,000 searches a month, and art 50 ai act is up 2,850% in a year, since the EU AI Act’s transparency obligations kicked in.
The research also told me where not to go: people searching chatbot ai want ChatGPT, and people searching assistente virtuale (virtual assistant) want a job.
Many entry points, all connected
Based on this data, I built 23 landing pages and 26 blog posts.
Each page is a different entry point, tied to a specific search: chatbots for a type of business, for a platform, alternatives to competitors, questions about AI. Every business has its own questions: a dentist has to answer patients, a hotel has to handle availability requests. Its page speaks that language, with the questions its customers actually ask.
The pages don’t live on their own. Landing pages and posts link to each other, so traffic moves from one page to the next and the site’s authority spreads across all of them instead of sitting on the homepage.
Link building
This is the piece that closes the loop, and it starts now. The goal is to build domain authority. Through internal links, that authority then flows down to every subpage.
The first step is SaaS and AI directories like Product Hunt, BetaList and AlternativeTo, along with Italian communities of web agencies and e-commerce businesses. Then there’s a WordPress plugin on wordpress.org, which brings an authoritative link and is a distribution channel in its own right. The product does the rest: on the Starter and Pro plans the widget shows “Powered by Kiedi”, so every site that installs it adds visibility. Web agencies, meanwhile, can resell it as a white-label product.
It’s a slow burn, built to pay off in traffic over the next 6, 9 and 12 months.
Where I want to be
Right now three clients are testing Kiedi. So rather than today’s numbers, it makes more sense to talk about goals.
I’ve set three scenarios for the next 12 months. The conservative one is reaching at least 10 customers. The moderate one is landing between 30 and 40. The exceptional one is getting to around 100.
Which line I end up on depends on three things: how well organic acquisition performs, how well I can get word of mouth going among the people using it, and how well I can work with agencies, which can bring in dozens of sites at once.
Customers over the next 12 months
- Conservative 10
- Moderate 35
- Exceptional 100
| Month | Conservative | Moderate | Exceptional |
|---|---|---|---|
| Oct 2026 | 0 | 0 | 0 |
| Nov 2026 | 1 | 1 | 2 |
| Dec 2026 | 1 | 2 | 4 |
| Jan 2027 | 2 | 4 | 7 |
| Feb 2027 | 3 | 6 | 11 |
| Mar 2027 | 4 | 9 | 16 |
| Apr 2027 | 5 | 12 | 23 |
| May 2027 | 6 | 16 | 32 |
| Jun 2027 | 7 | 20 | 43 |
| Jul 2027 | 8 | 25 | 57 |
| Aug 2027 | 9 | 30 | 76 |
| Sep 2027 | 10 | 35 | 100 |
For revenue I went with the most cautious assumption: every customer on the Starter plan, at €29 a month.
MRR over the next 12 months
- Conservative €290
- Moderate €1,015
- Exceptional €2,900
| Month | Conservative | Moderate | Exceptional |
|---|---|---|---|
| Oct 2026 | €0 | €0 | €0 |
| Nov 2026 | €29 | €29 | €58 |
| Dec 2026 | €29 | €58 | €116 |
| Jan 2027 | €58 | €116 | €203 |
| Feb 2027 | €87 | €174 | €319 |
| Mar 2027 | €116 | €261 | €464 |
| Apr 2027 | €145 | €348 | €667 |
| May 2027 | €174 | €464 | €928 |
| Jun 2027 | €203 | €580 | €1,247 |
| Jul 2027 | €232 | €725 | €1,653 |
| Aug 2027 | €261 | €870 | €2,204 |
| Sep 2027 | €290 | €1,015 | €2,900 |
My biggest worry? Right now Kiedi has no integrations: it’s a chatbot that answers questions. But it has plenty of room to grow. Integrations with other platforms will come based on how customers actually use it. I see it as building the product together with the people who choose to back the project and use Kiedi.
I’ll be back at the end of October with the first real numbers, to see which line we’re tracking.