From 5% to 23% Conversion Rate: How AI Changed Our Clients’ Sales

The Dashboard That Didn't Add Up

It was 11:40 p.m. in Kuwait, and I was staring at a client's Lojain dashboard the night before a quarterly review I'd been dreading all week. The client was a home renovation contractor in Toronto — good reputation, steady lead flow, but stuck. The number on screen: 1,140 WhatsApp conversations that quarter, 57 closed jobs. A 5.0% close rate, exactly. Three months earlier I'd told their owner on a call we'd get this into double digits. The review was nine hours away and I still didn't have a real answer for why it was stuck at 5%.

I'm Omar Sokar. I founded Kira Agency in 2018, and Lojain AI, our WhatsApp automation product, is the thing clients bring up first on calls like this one — because it's usually where the real problem is hiding, not in the ad account. That night the ad numbers were fine. Cost per lead was around $34, healthy for renovation work in that market. The leads were real. Something after the click was broken, and I needed to find it before sunrise.

Where the Deals Were Actually Dying

I pulled every conversation from the quarter and read the timestamps, not just the outcomes. Average first response time from the client's sales team: 4 hours and 20 minutes. A Friday night lead got a reply Saturday afternoon. Worse, 61% of conversations that opened with a real question — "how much for a kitchen remodel," "do you service Etobicoke" — got a first reply of "Hi, thanks for reaching out, how can we help you today?" Nobody answered the actual question. By the time a human read the message, the lead had usually already messaged two or three competitors.

This is the part clients don't want to hear and most agencies never tell them: your ad spend can be flawless and you'll still lose the sale in the first sixty seconds after the click, because speed and relevance in that first reply matter more than almost anything upstream of it.

What We Rebuilt Before the Number Moved

We put Lojain AI directly on the client's WhatsApp Business number, trained on their actual services, service area, and price ranges — not a generic bot reading from a script. First response time dropped to under twenty seconds. Every lead was asked three qualifying questions before a human ever saw the conversation: project type, budget range, and timeline. Leads that went quiet after the first exchange got pulled into an automated three-touch follow-up over the next 48 hours instead of falling into a dead inbox. Sales reps stopped chasing cold names and started closing warm, qualified ones.

23%, and Where I Got It Wrong First

Ninety days later, on almost the same lead volume — about 1,150 conversations — the close rate was 23%. That's roughly 264 closed jobs against 57 the quarter before, on spend that barely moved. At an average job value of $8,200 for that client, the shift added close to $1.7 million in closed revenue in a single quarter, from the same leads they were already paying for. That's the number I put in the case study, and it's real, but I want to be honest about how we got there, because it wasn't a straight line.

In week one I made a mistake I still think about. I had the AI's opening message try to do too much — acknowledge the lead, then immediately pitch a premium package upsell in the same breath. It read as pushy to someone who'd just asked a simple pricing question, and reply rates actually dipped that first week. I pulled it back: answer the question first, ask the qualifying questions second, sell nothing until a human is on the line. Speed alone had moved the close rate from 5% to about 9% in the first three weeks. Fixing the tone and the qualification logic is what carried it the rest of the way to 23%. Automation without the right first message is just a faster way to lose the same deal.

Why I Think Most Agencies Are Solving the Wrong Problem

Here's my honest opinion, and I know some people in this industry won't like it: most marketing agencies stop at the lead. They obsess over cost-per-lead and ROAS, hand the client a spreadsheet of names and numbers, and call the job done. That's still where most agency reporting decks end — cost, volume, maybe a click-through rate. Almost none of them touch what happens in the first sixty seconds after someone messages, because that's product and engineering work, not media buying, and it's harder to sell on a slide.

I'd argue that fixing response speed and qualification on WhatsApp is now a bigger lever on a client's actual revenue than shaving another 10% off cost-per-click. We've run this across more than a hundred brands now — food and beverage, real estate, retail, healthcare — and the pattern holds more often than not. The ad account gets all the attention because it's easy to measure and easy to sell as a service. The inbox is where the money is actually won or lost, and almost nobody is watching it closely enough.

What I'd Tell You to Do Monday Morning

If your business gets leads through WhatsApp, pull your own numbers before you touch your ad budget. Look at your average first response time and your close rate side by side. If your response time is measured in hours, you already know roughly where your close rate is capped, no matter how good the ads are. Fix the reply before you touch the reach.

That's the whole thesis behind Lojain AI and most of what we build at Kira Agency. If you want to see what your own numbers would look like with a twenty-second response time instead of a four-hour one, message us and we'll walk through it.

Talk to Us on WhatsApp