We were busy building something else entirely: elEco, an all-in-one system for running a business. Accounting, billing, inventory, CRM, HR, asset care, support tickets, catalogues, all talking to each other. A big, serious, sensible ecosystem.
And like every big, serious piece of software, it had a regular chatbot.
It was a perfectly nice chatbot. It sat there. It waited. If you typed something, it typed back. That was the whole personality.
Down the corridor, we were building analytics. It did what analytics does: a postmortem. It told you, in exquisite detail, all about the people who had already left.
So far, so ordinary.
Meanwhile, we were out meeting clients. Lots of them. And we kept hearing the same story.
They were spending good money on marketing, and it was working. People were arriving on their websites. And then… nothing. Visitors wandered around, found no one to talk to, and quietly slipped out. Enquiries came in and sat unanswered until the buyer had long since gone elsewhere.
For small businesses, it was even harder. Comments and replies on their social media posts piled up with no one to answer them. And they had no way to ask a website visitor for a name or a phone number before that visitor disappeared for good.
All that effort to bring people through the door, and nobody there to say hello.
That, we realised, was the most expensive failure of all.
Then it struck us. We had a regular chatbot that couldn’t see, and a postmortem that came too late to save anyone. What if they worked together, properly, in real time?
What if the system knew what you were reading, where you paused, what you clicked, what caught your eye, and then actually said something? Not after you’d gone. While you were still there.
So we built it.
And then, in the great tradition of software teams everywhere, we put it on a shelf.
Until one day we thought: let’s just try it. We installed it for a client, mostly out of curiosity.
It engaged more than 10% of their visitors.
People who would have browsed and left without a word were suddenly talking to the business.
That’s when it clicked. This wasn’t a regular chatbot anymore. It wasn’t a postmortem either. The two had become something new.
elNudge reads a visitor’s intent from what they do: the pages they visit, the path they take, where they linger, what they click. Then it speaks to them about exactly that. It answers the questions, picks up the conversations nobody had time for, and politely asks for contact details at the moment a visitor is most interested.
That something is elNudge.
An AI concierge.
There Are Two Kinds of Shopkeeper
The first stands behind the counter and lets you get on with it. Browse all you like. Leave whenever.
The second notices what you’re looking at. Shows you something you’ll love. Helps you choose. Remembers you next time.
You know which one you’d rather buy from.
elNudge is the second shopkeeper.
