Η σελίδα υπάρχει και στα ελληνικά. Διαβάστε τη στα ελληνικά
AI & Custom Solutions
Four ways to get hours back, a written standard for how we do it, and a diagnosis that tells you where to start.
What AI does inside a business: the overviewSectors
Not sure where to start?
Workflow X-Ray
In one to two weeks: a map of hours and euros, five opportunities with the payback worked out, and a demo running on your own data.
From €890 · credited 100% against the pilot
Performance and CRO
We run Google Ads and Meta Ads together with the landing pages, the tests and the tracking. One system, one account, one number we look at every week.
Free · a few slots a month
Tool
Enter your own numbers, or keep the example. Nothing is saved and nothing leaves this page.
Ad3,750
Landing page3,750
Dropping off here: 3,525 people a month
Each +1 point here: +€446 a month
Basket225
Dropping off here: 124 people a month
Each +1 point here: +€59 a month
Checkout101
Dropping off here: 46 people a month
Each +1 point here: +€49 a month
Order56
In the example, it's the landing page: each point is worth ≈€446 a month, against ≈€59 at the basket and ≈€49 at checkout.
In a funnel that multiplies, the same percentage point is worth most where the rate is lowest.
With the example: 3,750 clicks, ≈56 orders, ≈€2,673 revenue, ROAS 1.78, CPA ≈€26.94.
The improvement you pick is an assumption, not a promise. Only a test shows whether it's realistic.
Clicks = spend ÷ CPC. Orders = clicks × landing rate × basket rate × checkout rate. Revenue = orders × average order value. ROAS = revenue ÷ spend. CPA = spend ÷ orders. Value of one point at a step = orders ÷ that step's rate × 0.01 × average order value.
The system
A good ad on a bad page is an expensive way to fail. So we build it all together, and measure it all with the same ruler.
A reason for someone to stop scrolling. From the Strategy Lab, not a stock library.
we measure: watch time, sends
Google Ads and Meta Ads, with budget that moves to wherever cost per result falls.
we measure: cost per lead or order
One page for one message, fast on mobile, with one action. We build it ourselves, in the same codebase as your site.
we measure: conversion rate, LCP
One test at a time, with the hypothesis written down first and a control group. We only call a winner when the sample allows it.
we measure: lift vs control
Conversions are sent from the server, with consent, so the platforms see what the browser misses.
we measure: conversions recorded
Anyone who writes in your DMs, on Viber or on WhatsApp gets an answer straight away, and a person when they need one.
we measure: response time, conversations that close
From ad to conversation
For customers in Greece, Viber comes first. For tourists, the diaspora and buyers abroad, WhatsApp.
Since 1 July 2025 WhatsApp has charged per message. But when someone writes to you from a click-to-WhatsApp ad and you reply within 24 hours, a free 72-hour window opens. We design the conversation to close inside it. source: Meta for Developers
We use ‘comment a word, get the link in your DMs’ as a delivery layer, not as the idea. Vendors of these tools claim 15–25% conversion against 2–5% by hand. Nobody independent has audited that, so we measure it ourselves on every campaign. source: SmartReply · vendor claim
The automated reply is the same work as our AI Front Desk, with the same rule: it says it's AI in its first sentence. See the AI Front Desk
Viber
Hi, is the air-con offer still on?
It runs until Sunday. Shall I hold one for you to collect tomorrow?
Hi, do you ship olive oil to Munich?
Yes, 4–6 working days. Here's the 3-litre tin with shipping included.
Free window: 72 hours
Example conversation
How we test
Every test starts with a hypothesis written in the log, one primary KPI and a control. If the sample isn't big enough, we say so: ‘we don't know yet’ is a result, not a failure.
Every month
The shipping note above ‘Continue’, on 50% of checkout traffic.
Nothing yet: the sample hasn't reached the size we set in advance.
The same test, until the sample is complete. Then the decision.
We only sell ad management together with a landing page and tracking. Otherwise you'd be paying us to send people into a leak.
We price it after the teardown, because it depends on the channels, the number of landing pages and the pace of testing. The 30-minute teardown is free: we look at your ads, page and tracking together, and you leave with three specific fixes, whether or not we work together.
Enough to learn something. A test needs enough conversions to show a difference, so the budget is set by the goal and the cost per result, not by a round number. In the teardown we work out together the minimum that makes sense for your case.
Tracking and page fixes show up in the first months. Tests take as long as the sample needs, and we state that up front for each test. We don't promise percentages before seeing your numbers; after the teardown we give you a range, not a promise.
No. Social content is part of the system: the mechanism that makes someone stop, the ad that brings them in, the page that convinces them. We don't take on account management with posts alone, because there nobody measures what each post brought in.
It means sending conversions to the ad platforms from your own server, with the visitor's consent, rather than only from the browser. Browsers and ad blockers lose some conversions; when the platforms can't see them, they optimise blind.
Yes, with WooCommerce and most platforms that allow page changes and a tracking connection. If we built your e-shop, landing pages and tests go into the same codebase. If not, we check in the teardown what we're able to change.
We go through your ads, landing page and tracking together. You leave with three specific fixes, in writing, whether or not we work together.