Speed to Lead Test: Time How Fast Your Business Really Replies

Tonight at around 10:15, borrow a phone your staff will not recognise, maybe your cousin's or your driver's, and send your own business a WhatsApp message: "Salam, what is the price for this, and can you deliver to Gulberg by Saturday?" Write down the time. Then go to sleep.

Tomorrow you will know your real speed to lead, which is simply how long a new customer waits before your business gives them a useful answer. Not the number your team mentions in a meeting. The number a stranger actually lives through. Most owners guess low, because they remember the chats they answered quickly and forget the ones that sat until morning.

What follows is a self-test. You will play mystery shopper with your own business across WhatsApp, phone, your website form and your ads, fill in a simple scorecard, and then fix each leak you found, one at a time. It takes about two days of light effort, and you do not need to buy anything to run it.

Set up your mystery shopper kit before you send anything

The test only works if your team does not know it is happening. Announce it and replies will be lightning fast for exactly two days. Keep it between you and one person you trust.

1. Get two or three numbers your staff have never seen. Family members, a friend, a spare SIM. You need more than one, because the same unknown number contacting you four times starts to look odd.
2. Make a scorecard. A page in a notebook or a simple sheet with six columns: channel, day and time sent, time of first reply, time of first useful reply, who replied, and what they actually said.
3. Write realistic questions. Use what real customers ask you: a price, delivery to a specific area, whether something is in stock, the next free appointment. Write one in English and one in Roman Urdu, and if your customers do this, record one short voice note in Urdu.
4. Decide what "useful" means before you start. A useful reply answers the question or moves it forward with a clear next step. "Ok", a thumbs up sticker, "please call" or "our team will get back to you" does not count.
5. Pick your times. Spread the tests over at least one weekday and one weekend day. Include one after closing time and one during your busiest hour.

That fourth column, the gap between the first reply and the first useful reply, is the one owners skip. It is often the biggest number on the whole sheet.

Test 1: the WhatsApp message that arrives after closing

Send the late night message, then a second one from another number during a busy afternoon. For each, note whether an automatic greeting arrived, when a person first replied, and whether that reply answered the price and delivery question or just asked you to call or visit the shop.

Also notice whether anyone asked you something back. A good salesperson replies with the answer plus one question, such as which size you need or which area you are in. That question is what keeps the conversation alive. A reply that only states a price and stops is easy for the customer to compare and forget.

Then send the voice note. Plenty of customers in Pakistan would rather speak than type, especially older customers and anyone messaging while driving. Watch whether the voice note gets answered at all, or quietly skipped because the person at the counter could not play it with customers standing in front of them.

Test 2: the call that rings out at lunch time

Call your main number when you know the line is busy or the person who answers is away: lunch, Friday prayers, the last hour before closing. Let it ring out. Do not leave a voicemail, because most real customers will not either.

Now wait, and write down what happens. Did anyone call back? Did a WhatsApp arrive saying "Sorry we missed your call"? After how long? A missed call with no call back has a response time of never.

In many businesses nobody owns missed calls. The phone shows a red number, whoever sees it assumes someone else has dealt with it, and the list is cleared before closing without a single call made. You will not see this leak in any report, which is exactly why it survives for years.

Test 3: your website form and your Facebook lead ads

Fill in your own contact form with one of the test numbers and a believable request. Then ask your team, casually and without warning, "Where do our website enquiries end up?" If the honest answer is an email inbox that one person checks twice a day, you already know roughly what your response time will be.

Lead ads on Facebook and Instagram need a different check, because you cannot easily mystery shop your own ad. Ask whoever manages them to pull the last ten leads and put two dates next to each: when the lead came in, and when someone from your team first contacted them. Leads often sit in Meta's leads center or in a downloaded sheet until someone remembers to export them, while you keep paying for the ads that bring them in.

Test 4: ask a second question, then ask for a person

The first answer is only half the test. Reply to whatever you received with a natural follow-up: "Is there any discount for cash?" or "Can you send a photo of the blue one?" Time that reply too.

Then, in a different chat, write plainly: "Can I talk to someone?" See what happens next. This is where badly set up automation fails. The auto-reply answers instantly, the second message falls into a queue nobody watches, and the customer who asked for a human gets the same menu again. Fast and useless is not much better than slow.

Reading your scorecard without fooling yourself

Once the tests are done, look at the sheet as a customer would, not as an owner. Three things matter more than any average.

• Your worst wait. Five quick replies and one overnight gap is not a good score. The overnight one is the customer who booked with someone else.
• Time to a useful reply. Measure from the customer's first message to the first answer that actually helped, not to the first message of any kind.
• The nevers. Count every test that got no reply at all: the unreturned call, the form nobody answered, the ignored voice note.

Then check the "who replied" column. If every good answer came from the same person, your speed to lead is really that one person's free time. The day they are on leave, at a family wedding or busy with walk-in customers, your response time falls apart.

Honestly, almost every business that runs this test finds at least one channel with a never in it. That is normal. It is also the easiest customer you will ever win back, because they already chose to contact you.

Fixing what you found, one leak at a time

Go down your scorecard and fix the worst row first. Do not try to fix all five channels in one week; you will end up with half-finished replies everywhere.

If the WhatsApp message sat overnight

An AI agent connected to your WhatsApp can answer prices, timings, location and stock questions at 10 pm in the language the customer used: Urdu, Roman Urdu or English. It can also ask one qualifying question and book a call for the morning. With WhatsApp Coexistence, the number stays on the WhatsApp Business App on your phone, so your staff can still jump into any chat. When we set this up for clients, we build the answers from the questions in their own chat history rather than a generic script, because the AI is only as good as the answers it is given. For longer conversations, such as qualifying a property buyer, AI agents can handle several steps and pass a short summary to your salesperson.

If the voice note was ignored

Voice notes can be handled too, but test this with real recordings before you trust it. We learned that the hard way: speech-to-text sometimes wrote our customers' Urdu in Hindi script, and the AI started replying in Hindi until we forced Urdu transcription. Send ten real voice notes from your own customers through any system before it goes live.

If the missed call never came back

A missed call can trigger two things: an alert to a named team member, and a WhatsApp to the caller. Know the rule before you plan this. The caller has not written to you on WhatsApp, so that first message must be an approved template, which Meta charges for per message at a price Meta sets and changes. Replies to customers who message you first are free. Give the call back task a time limit, and if nobody acts, move it to a second person. Do not use an unofficial bulk sender app to cut corners here; those tools put your number at real risk.

If the form went to an inbox nobody checks

Send every website form and ad lead straight into one place with the time and source recorded, and send the customer an instant confirmation of what they asked for. Add a clear line on your form asking permission to contact them on WhatsApp. Our lead automation page shows how these pieces connect.

If the first reply was fast but said nothing

Rewrite it. Compare these two:

"Thank you for contacting us. Our team will get back to you shortly."

"Walaikum salam! Yes, this one is in stock and we can deliver to Gulberg by Saturday. Do you want it in grey or blue?"

The second one answers, then asks one simple question. Write your auto-replies the way your best salesperson talks.

If everything depended on one phone

Move enquiries into a shared inbox and an automated CRM that assigns each lead to a person and flags any lead that has waited longer than your target. The manager can then see who is waiting without scrolling through anyone's personal phone.

Run the same test again in two weeks

Once a fix is live, repeat the test with fresh numbers and put the new scorecard next to the old one. Then repeat it every couple of months, and any time you change staff, timings or prices. These setups drift: an offer ends, a new salesperson joins, and the auto-reply still promises last month's deal.

If you would rather have someone else play the customer, our audit starts by checking how fast each of your channels really replies, much like this test. Or simply send that 10:15 pm message tonight. By tomorrow morning you will know more about your speed to lead than any report has told you.

Further reading

See how this looks in practice across our AI automation, custom CRM and lead generation services, or tell us the outcome you need.