Voice Order Entry: How Zivni Lets Sales Reps Place Orders in Urdu With Just Their Voice

Last month, I was in Faisalabad sitting with a distributor who manages about 300 outlets across the city. His top sales rep — a guy who's been doing this for 15 years — pulled out his phone and showed me how he takes orders. He had a crumpled paper list in one hand, a calculator app open on his phone, and was mentally converting dozens (darjan) to units while trying to remember which SKUs the retailer actually wanted.

I asked him, "What if you could just say it out loud in Urdu and the order was done?"

He laughed. Then I showed him Zivni's voice order entry. He stopped laughing.

The real problem nobody talks about

Here's something most SaaS companies building for field sales don't understand: a huge percentage of sales reps in Pakistan, and across emerging markets in the Middle East and Africa, are not comfortable typing on apps. It's not that they can't use a smartphone — they all have one. It's that typing product names, navigating dropdowns, selecting quantities for 30-40 SKUs per outlet, doing this 25+ times a day... it's exhausting. And slow.

I've watched sales reps in Lahore's Anarkali market, in Karachi's Saddar, in small towns in interior Sindh. The pattern is the same. The retailer rattles off what he needs — usually in Urdu or a mix of Urdu and Punjabi — and the rep either scribbles it on paper or struggles to punch it into whatever app the company forced him to use.

Most order-taking apps were designed by people who've never stood in a 4-foot-wide shop in Ichra with three customers waiting behind them.

That's the gap we built for.

How voice order entry actually works in Zivni

The idea is simple but the execution took us months to get right. When a sales rep opens Zivni at a retail outlet, they can tap the mic button and just... talk. In Urdu.

"Surf Excel chota packet, das. Lux sabun, bees. Tapal Danedar, paanch."

Zivni's voice engine picks that up, matches it against the distributor's actual SKU catalog, maps quantities, and creates an order draft. The rep reviews it on screen, confirms, and it's done. The whole thing takes maybe 20 seconds for a 10-line order.

A few things we had to figure out that weren't obvious:

Local language nuances. Nobody says "Surf Excel 500g washing powder" out loud. They say "Surf chota" or "bari Surf" or "neeli Surf." We built a layer where distributors can configure local nicknames and aliases for their products. So when a rep says "neeli Surf," the system knows exactly which SKU that maps to.

Numbers in Urdu vs English. Some reps say "das" (ten), some say "ten," some mix both in the same sentence. Our voice engine handles all of it.

Noisy environments. Markets in Pakistan are loud. Rickshaws, people shouting, TVs blaring. We spent a lot of time on noise filtering specifically for bazaar-type environments. It's not perfect — I won't pretend it is — but it's gotten good enough that reps trust it.

Confirmation before submission. This was non-negotiable for us. Voice order entry doesn't mean blind automation. The rep always sees the parsed order on screen before confirming. If the system misheard something, one tap fixes it. But honestly, after a week of use, most reps tell us the accuracy is above 90%, and it keeps improving as the system learns their voice patterns.

Why this matters more than you think

Let me throw some numbers at you from our own data.

A typical FMCG sales rep visiting 25 outlets a day spends an average of 3-4 minutes per outlet on order entry when using a traditional app. With voice order entry, that drops to about 60-90 seconds. That's roughly 50-60 minutes saved per day. Per rep.

Now multiply that across a team of 50 reps. You're looking at nearly 50 hours of productive time recovered every single day. That's time reps can spend actually selling — having conversations with retailers, checking shelf placement, pushing new products, visiting those 3-4 extra outlets they could never get to before.

But here's what I think is the bigger deal: adoption.

I've talked to dozens of distribution company owners in Pakistan and the UAE who tried other field sales apps and failed. Not because the apps were bad, but because reps wouldn't use them. The apps felt like extra work. And when a 45-year-old sales rep who's been taking orders on paper his whole career is told to start typing on a phone app, there's resistance. Naturally.

A hands-free order taking app changes that equation. Speaking is natural. Everyone does it already — the rep is already verbally discussing the order with the retailer. Now instead of converting that conversation into typed text, they just let Zivni listen.

We've seen adoption rates jump from around 40-50% (with traditional apps) to over 85% within the first month when voice is enabled. That number honestly surprised even us.

What's coming next

We're working on a few things I'm excited about. Urdu voice commerce isn't just about order entry — we're expanding it to let reps do stock checks, report market feedback, and log competitor activity, all through voice in Urdu. Imagine a rep walking out of a shop and saying, "Competitor ne yahan buy-two-get-one chalaya hua hai on Olper's" and having that logged as structured competitive intelligence in your dashboard.

We're also testing Pashto and Sindhi support. Pakistan isn't monolingual, and neither are the markets in the Gulf where Pakistani and Indian sales reps work for distribution companies.

The way I see it, the future of field sales tech in our part of the world isn't about building more complex dashboards for managers. It's about making the rep's life simpler. If the rep is happy, uses the app willingly, and gets accurate data in — everything else downstream just works.

If you're running a distribution operation and your reps are still punching orders manually, try Zivni's voice order entry for a couple of weeks. I genuinely think you'll see the difference on day one. And at $5 per user per month, the math kind of speaks for itself.