Response-time audit of 98 aesthetic clinics: where your ad budget goes

Israel is a small market with dense competition. A customer who needs an aesthetic procedure or an implant almost never writes to one place. She opens three or four browser tabs, sends the same WhatsApp message to all of them, and waits. What happens next is decided not only by the quality of your work, your reviews and your portfolio, but also by who replied first.

At WhaleBiz we decided not to theorise about this but to run our own market study in September 2026.

How we measured: the methodology

The sample was 98 aesthetic medicine and cosmetology clinics in Israel. We sent each of them the same WhatsApp message from an ordinary number, exactly the way a person deciding where to go would. Then we simply waited and recorded three things:

• how long it takes to get a first reply during working hours;
• how many enquiries are never answered at all;
• how the speed of handling enquiries changes at weekends and on public holidays.

One important note on the arithmetic, so the numbers can be trusted. Of the 98 enquiries we sent, 92 received a reply. The remaining 6 were still "in flight" when we closed the study, which is to say they were never answered. All the averages below are therefore calculated over the 92 replies received, and the no-reply share is calculated over all 98 enquiries.

We chose WhatsApp for a reason. According to Bezeq's annual Internet Report the app needed a full decade to reach 90% of Israel's population, and today it is effectively the default way to contact a business in the country. A clinic whose WhatsApp is quiet is, for a large share of the market, simply closed.

The results: assumptions versus reality

Few owners of these businesses believe they have a problem handling enquiries. Our audit, however, exposed a critical gap between what management expects and what the customer actually experiences.

Here is what the measurements showed:

• 1.5 hours. That is the average time to a first reply during working hours.
  • What it means for the business. The clinic misses the "golden window" of the first 15 minutes, when the customer's readiness to book is at its peak. The longer the pause, the lower the chance the conversation happens at all: by the time the reply arrives, the person is already talking to whoever responded earlier.
• 6 clinics out of 98 (6.1%) never replied.
  • What it means for the business. A direct drain on the marketing budget. Every sixteenth prospective patient met total silence and almost certainly booked with a competitor.
• 14 hours. That is the average reply delay at weekends and on holidays.
  • What it means for the business. WhatsApp is an instant-communication tool, and a customer who writes at the weekend is not prepared to wait more than half a day. While staff are legitimately resting, enquiries go cold and the patient moves to the few who answer straight away.

How this compares with international research

Here we should be fair: by international standards, Israeli aesthetic clinics do not look bad at all. And that is exactly why the conclusion is less comfortable than it first appears.

The classic audit behind Harvard Business Review's "The Short Life of Online Sales Leads" is built exactly like ours: the researchers sent test enquiries to 2,241 US companies and measured how long a reply took. The result was an average reply of 42 hours, with 23% of companies never replying at all. Against that backdrop, our 1.5 hours and 6.1% silence look almost exemplary.

The problem is that the customer does not compare you to the market average. She compares you to the clinic that answered her first. That is where the real arithmetic begins. The same HBR research found that companies replying within an hour are nearly 7 times more likely to qualify the lead than those who reply later. And the Lead Response Management study by Dr. James Oldroyd (MIT Sloan) together with InsideSales broke the window down further: replying after 30 minutes instead of 5 cuts the odds of reaching the lead at all by roughly 100 times, and the odds of qualifying them by 21 times.

In other words, the meaningful boundary does not sit between "an hour and a half" and "a day". It sits between "minutes" and "everything else". An hour and a half lands squarely in "everything else".

Weekends deserve a separate look. Fourteen hours of waiting on a Saturday or a holiday is not "slightly slower". According to Infobip's data on business messaging in WhatsApp, users treat the channel as instant communication rather than email. Someone who writes on Friday afternoon is not waiting for a reply on Sunday, she is waiting for a reply today. A detailed breakdown of what happens to enquiries over Shabbat and the holidays is in our separate article on weekend and holiday coverage in Israel.

Where the problem actually sits

When a clinic's revenue is flat, the explanation is usually ready in advance: "we don't get enough leads" or "the traffic is off-target, only browsers come in". The blame automatically lands on the marketer who brings in "the wrong people", or the manager who cannot close.

Part of that is true. But the measurements exposed a huge slice of enquiries that nobody in those arguments even mentions. Those leads were neither bad nor good. They simply were not answered, or were answered late.

They came in outside working hours, got lost between channels, or never received a second touch. In the conversion report they were recorded as "did not buy" and filed under "off-target". The argument about who is at fault, marketing or sales, is conducted around the first group of enquiries. The second group sits unhandled in the meantime, even though the ad account has already paid for it.

The same mechanic operates inside the existing customer base: a customer who has already paid once but got no follow-up at the right moment drops out of every report. What that looks like and what it costs is covered in our guide to reviving a dormant customer database.

Why this is more expensive than it looks

In a segment with a high average ticket, the price of a single lost enquiry is calculated very differently than in ordinary retail.

Take an average aesthetic medicine clinic in Israel:

• average ticket: ILS 4,000;
• 200 enquiries per month;
• conversion to appointment: 20%, which is 40 customers and ILS 160,000 of revenue.

Now suppose that slowness or inattention loses a quarter of the enquiries, 50 of them. Even at the standard 20% conversion, that is 10 lost customers and ILS 40,000 per month. Over a year that turns into ILS 480,000 of revenue never realised.

And this is money you already spent at the moment of the ad click. The only question was whether those customers would reach the treatment room or burn up on the way. Even if your clinic is smaller than average and you lose only 15-20 enquiries a month, you still hand the market the price of a good new car every year.

The figures above are a market average. To see your own, we built a calculator: it sits on the pricing page and works out lost revenue from four of your own numbers - how many enquiries come in, what share of them gets no real answer, how many of those would have booked, and what one customer is worth. For some clinics the result will be smaller than our example, for others considerably larger.

Why "one more receptionist" will not fix it

Hiring another person is the obvious but expensive solution, and it closes only one problem out of three. During the day the flow really will clear faster. But it does not solve the main problems:

1. Weekends, holidays and night hours. A customer who writes on Friday afternoon waits, by our measurements, an average of 14 hours. The second and third receptionists legitimately sleep and rest during the same hours as the first, so hiring them does not move that number.
2. The human factor. A receptionist physically cannot hold 40 open conversations in their head. Customers who said "let me think about it" will still be left without a follow-up.
3. Hidden costs and turnover. Hiring, onboarding, sick days, holidays and weeks of training. In the end there is a high risk that the employee you trained at your own expense simply walks across the road to a competitor.

What actually closes the hole in sales

Not "one more person", but an automated AI system that does what a human does worse than a machine:

• Replies within seconds in any channel (WhatsApp, Instagram, Facebook, the website widget), any day, any hour.
• Automatically handles routine questions about prices, address and opening hours, taking load off the staff.
• Screens enquiries and filters out the off-target ones.
• Logs every enquiry in the CRM, so no customer is lost.
• Runs follow-ups on its own with people who went quiet, or with the clinic's existing customer base.
• Hands the receptionist an already warmed-up customer together with the conversation history.

The human being does not disappear in the process. They simply stop working as an answering machine and start doing their actual job: selling and looking after patients in the clinic.

One boundary we hold firmly: the agent does not diagnose, does not collect anamnesis and does not judge from a photo whether a treatment is suitable. Anything that requires clinical judgement, and any report of a complication, goes to a human immediately. The detail on that boundary, and on how the agent is built specifically for an aesthetic clinic, is on the AI agent for aesthetic medicine clinics page and in our breakdown of the WhatsApp bot for a clinic.

Test the technology in one click

Do not take our word for it that multi-agent systems talk like an experienced receptionist. And do not assume a basic agent has to cost five, ten or fifteen thousand shekels.

Test our lead-handling system yourself, right now. Open the chat window in the corner of this page and ask awkward questions in Hebrew, Russian or English, write text or send voice, ask the price of a procedure, send a photo, or try to confuse it. Do not hold back. The same agent answers in WhatsApp too, so if you would rather test it in a real channel, message us from the contact page.

Worth trying to break first:

• switch from Hebrew to Russian mid-sentence;
• ask the price of a treatment we do not offer;
• ask to be booked for tomorrow at 22:00;
• send a voice message instead of text;
• ask something the agent should not answer, and watch how it hands the conversation to a human.

The main conclusion of the study

Clinic managers often make decisions based on internal reports and their own intuition, but rarely see the real picture through the customer's eyes. The statistics from 98 clinics showed that lost revenue sits not only in the plane of poor marketing or staff performance, but also in the banal absence of instant contact at the moment of interest.

Numbers do not lie, but every clinic has its own. If you want dry statistics on your own sales operation instead of guessing from market averages, we are ready to help. We will audit your clinic's response speed: send test enquiries, find the hidden holes in sales, and show you the real numbers.

Leave your details and we will come back to you with the measurement results for your clinic.

Questions and answers

How exactly was the measurement done?

We sent every clinic the same WhatsApp message from an ordinary number, exactly as a real customer would, and measured the time until the first human reply. The sample was 98 aesthetic medicine and cosmetology clinics in Israel, measured in September 2026. 92 clinics replied and 6 had not replied by the time we closed the measurement. All averages are calculated over the 92 replies received; the no-reply share is calculated over all 98 enquiries.

If the global average reply is 42 hours, why is 1.5 hours bad?

Because the customer does not compare you to the market average. She compares you to the clinic that answered her first. She writes to three or four places at once, and the winner is not whoever is fastest on average but whoever is fastest in that particular chat. The MIT and InsideSales research found that replying after 30 minutes instead of 5 cuts the odds of qualifying a lead by roughly 21 times. An hour and a half is far outside that window.

Would it not be simpler to hire another receptionist?

An extra person speeds up the daytime flow but closes only one of the three leaks. Weekends, holidays and night hours stay uncovered, because the second receptionist rests during the same hours as the first. And a person cannot hold several dozen open conversations in their head, so the people who said "let me think about it" still get no second touch. On top of that come hiring, training, sick days, holidays and turnover risk.

Will the AI agent diagnose patients or give medical advice?

No. The agent handles intake, routine questions about prices, address and opening hours, screening and booking only. It does not collect anamnesis, does not diagnose and does not judge from a photo whether a treatment is suitable. Anything that requires clinical judgement, and any report of a complication, goes to a human immediately, at any hour.

How long does setup take and what does it cost?

The timeline depends on the number of channels and calendars and is counted from the moment the clinic hands over its materials: price list, treatment descriptions, opening hours and answering rules. Current plans, what each one includes and the one-time setup fee are published on the pricing page. The same page carries a calculator that works out your own lost revenue from your own numbers.

Sources

• Harvard Business Review, "The Short Life of Online Sales Leads" (James Oldroyd, Kristina McElheran, David Elkington, 2011) - response-time audit of 2,241 companies.
• Lead Response Management Study, Dr. James Oldroyd (MIT Sloan) and InsideSales - how the odds of contact and qualification fall in the 5-to-30-minute window.
• InsideSales, Lead Response Management (infographic) - summary of touch counts and response times.
• Bezeq, Israel's annual Internet Report - WhatsApp and digital channel penetration in Israel.
• Infobip, WhatsApp statistics - business messaging volume and user expectations around reply speed.
• WhaleBiz, in-house response-speed measurement, 98 aesthetic medicine and cosmetology clinics, Israel, September 2026.