Online store response time. How Viber and Telegram inquiries come together in one CRM

You won't find a store named in this breakdown. We pieced the setup together from requests Ukrainian online stores bring to E-chat, and it repeats with almost no variation. Viber lives on the owner's phone, Telegram on a manager's phone, inquiries land with different people, and the customer waits until one of the two glances at a screen.

We present the threefold drop in response time as a sample calculation and show which minutes it is made of. Alongside it comes a measurement method a store can use to calculate its own first response time before and after connecting and judge by its own numbers.

Key takeaways

• According to a 2026 survey published by Ukrainski Novyny, 74% of respondents in Ukraine use Telegram and 59% use Viber, so a store has to keep both channels.

• First response time is measured as a median and as the share of replies that met the target. An average hides inquiries that sat unanswered for half a day.

• The before measurement is done by hand, using two weeks of chats on the phones.

• In our sample calculation, all the gain comes from two pieces. Messages get noticed sooner and nobody hunts for an order across two phones, while managers type their replies just as they always did.

• Inquiry routing and notifications are set up in the CRM, and E-chat brings Viber and Telegram chats into it.

Where the delay comes from when Viber and Telegram live on different phones

Take a store that gets 40 messenger inquiries a day. There are two managers, and in the evening the owner replies himself. The Viber number is printed on the inserts in every parcel and has been on the website since the first year, while Telegram was added later on a different phone because that was quicker.

Two channels are not a whim here. In a 2021 Interfax-Ukraine survey, 73.6% of Ukrainians used Viber and 31.6% used Telegram. A 2026 survey published by Ukrainski Novyny shows a different picture, Telegram at 74% and Viber at 59%. Longtime customers write in Viber out of habit, new ones come through Telegram, and turning off either channel means cutting off part of your orders.

Whoever holds the phone sees the message

A manager is packing orders and calling Nova Poshta, the phone lies face down. A sizing question sits in Viber until someone notices it. The other manager is free at that moment and sitting three feet away, but can't see a colleague's phone and can't help.

The order has to be found across two chats

Yesterday a customer placed an order in Viber, today she writes in Telegram asking where her parcel is. The manager opens Telegram, sees an unfamiliar name, goes to Viber to search by phone number, then to the spreadsheet for the tracking number. The reply itself is one line, but it goes out about ten minutes later.

If Viber is on a colleague's phone, the team starts forwarding screenshots to each other, and those minutes also count toward response time, even though the customer never sees them.

Evenings, weekends and vacations

After seven in the evening, messages pile up. The owner answers from the couch whatever he happens to see, and the rest waits until morning. When a manager goes on vacation, his Telegram stays silent for a week, and when he quits, the number and the entire customer history leave with him.

How to measure first response time before connecting

The CRM doesn't have this data, since the chats live on the phones. So the first measurement has to be done by hand. It takes a couple of evenings, and without it you'll have nothing to compare against a month later.

There is one main rule. Measure the after the same way you measured the before, otherwise the gap in the numbers will only show a gap in methods.

Which message starts the clock and which one stops it

The clock starts at the customer's first message in a new inquiry. If they sent three messages in a row, count from the first one. A new inquiry is either the very first message from that person or a message after a break of more than a day.

The clock stops at the first reply from a real person. A short message like one moment, checking counts too, since the customer sees they've been heard. An auto-reply from a CRM bot never counts, more on that below.

A two-week sample from the phones

Pick two ordinary weeks without sales or holidays. Open every Viber and Telegram chat on each work phone one by one and log in a spreadsheet the channel, date, time of the customer's message, time of the reply, who replied, and whether the message came in during business hours.

At 40 inquiries a day, two weeks add up to 560 rows, which is already a lot. Every other day is enough, as long as you take the days in a fixed sequence rather than picking convenient ones. Be sure to log chats where the customer wrote and never got a reply. They are easy to miss, since at the bottom of the list they get lost among newsletters.

Count business hours and nights separately

A customer writes at twenty to midnight and gets a reply at five past nine, and that single row adds nine and a half hours and wrecks the whole statistic. So we keep two columns. The first holds inquiries during business hours, the second everything that came in during evenings, nights and weekends. For the second column, the clock starts at the beginning of the workday. This shows how fast the team clears the overnight queue in the morning, and that number doesn't get mixed with the daytime one.

Median, tail and lost inquiries

The average response time lies here. Say 90 replies came after 10 minutes and 10 after four hours. The average comes out to 33 minutes, though nobody actually waited that long, some waited ten minutes and others half a day.

So sort the column in ascending order and take three numbers. The median, meaning the value exactly in the middle of the list. The time within which nine replies out of ten were sent, which for 100 rows is row 90. And the share of inquiries with no reply or a reply after more than a day, as a separate percentage. The third number is usually less pleasant than the first two, and you need to calculate it either way.

What changed in the setup

The numbers stayed the same. The Viber number from the inserts and the Telegram number were connected to the store's CRM through E-chat, customers write where they always did, and messages show up in the customer card. Managers can reply from the CRM or the old way from the phone, and all messages from devices are saved to the CRM automatically, so the history stays intact.

Connecting the numbers

In the E-chat dashboard, a number is added and confirmed from a mobile device by scanning a QR code, after which chats sync the same way as in the desktop version. The last step, linking to the CRM, follows a step-by-step guide and takes no more than a minute. Among the CRMs Ukrainian stores use, Viber and Telegram both connect to SalesDrive, OneBox, KeepinCRM and Bitrix24. KeyCRM and Sitniks connect to Viber through E-chat, but not to Telegram.

Viber has one catch, covered in the expert insight below.

Who answers a new message

The integration itself doesn't decide which of the two managers takes a customer. That's a routing rule, and it lives in the CRM. A workable option for a store with two managers looks like this. A customer who already has a card goes to whoever handled their order, no matter which messenger they wrote in. A new customer is assigned in turn. If the assigned manager hasn't replied within 15 minutes, the CRM creates a task for the other one.

Notifications about new messages are also set up in the CRM. Make sure they reach the manager's phone, since an open browser tab won't help while the manager is in the warehouse packing orders.

The overnight auto-reply and why it stays out of the measurement

After seven in the evening, a CRM bot sends the customer a message saying the order will be processed from nine in the morning, and E-chat delivers it to Viber or Telegram. The same bot also creates a morning task for the assigned manager, so the overnight queue doesn't get lost.

For the metric, such an auto-reply means nothing. The customer got a polite acknowledgment, but no actual answer. If you count it, the overnight column drops to zero and the after measurement shows a miracle that never happened.

[Expert insight] In the E-chat help center, we separately cover the things stores trip over after connecting Viber. The phone the number was activated on can be switched off and put in a drawer, and that doesn't affect the service. But you must not delete Viber from it or use the Change Number feature. The second story begins when the entire customer base lands in the CRM for the first time and you want to message everyone. According to our help center, nine out of ten Viber bans are caused by sending first messages too often. We haven't seen a limit with customers who have replied to you at least once, and for contacts with no chat history we recommend no more than 20 new messages a day. On a fresh number, a ban can happen as early as the fifth such message in a day.

Sample calculation. Where the threefold drop comes from

The numbers below are made up for illustration, we did not measure them at a specific client. The logic is easy to apply to your own data. Split first response time into pieces. First the message sits until someone notices it, then the manager looks up the order and past chats, and only after that types the reply.

Median breakdown before and after

Before connecting, the business-hours median is 24 minutes. Of that, the message waits 15 for someone to pick up the phone, 6 go to finding the order across two messengers and a spreadsheet, and 3 to the reply itself.

After connecting, all messages are visible in the CRM, each has an assigned manager, and the notification arrives right away. Noticing a message now takes 4 minutes, the history and order are in the same card and take a minute, and typing still takes 3. That's 8 minutes in total, three times less. Typing didn't get a second faster, all the gain came from the first two pieces.

The tail changes more than the median

In the same example, before connecting nine replies out of ten went out within 2 hours, and after within 25 minutes. The difference here is almost fivefold, because long waits used to happen for one reason, an inquiry sat on a phone nobody was holding. When two people see a message and the CRM sends a reminder after 15 minutes, such cases almost disappear.

In the example, the share of inquiries left without a reply for more than a day falls from 6 in 100 to 1 in 100. What remains usually comes from routing errors and from customers who wrote something like thanks, where no reply is needed. Mark such rows separately in the next measurement.

How to compare before and after without fooling yourself

In the first days after connecting, the team replies faster simply because everyone is watching the new setup. After a couple of weeks the novelty wears off and the numbers return to the working norm. That's why the after measurement starts in week three.

The formula is simple. Mistakes usually happen elsewhere, in which days and which messages end up in the sample.

Comparable periods

Compare two weeks with two weeks, weekdays with weekdays. If one of the weeks had a sale and the other didn't, the comparison loses meaning, since volume went up and managers physically reply more slowly. When volume differs between periods by more than one and a half times, look at time per manager and don't draw conclusions from a single row.

Take after data from the CRM using the same rules

For the second measurement, you don't need to scroll through phones, all messages sit in the cards with their send times. If the CRM has a response time report, check what it counts from. A built-in report may count auto-replies or run around the clock, in which case it's easier to export the messages and calculate them by the same rules as the first time, with the same clock start, the same business hours and no bots.

A one-page table

The measurement result fits into six columns. Period, number of inquiries, business-hours median, time for nine replies out of ten, median time to clear the overnight queue, and share of inquiries without a reply for more than a day. One row - one measurement.

Fill in the first row from the phones before connecting, the second from the CRM in the third week after, and then one row a month on the same weekdays. If a number starts creeping up, open the tail chats in the CRM and see who they got stuck with and at what hours.

FAQ

Question: How do I fit the measurement into the seven-day free trial?

Answer: Take the first row of the table from the phones before the trial starts, while the old setup is still in place. The seven days themselves will go to connecting numbers, setting up routing rules in the CRM and checking that messages reach the right manager. Do the comparison after you've paid.

Question: Does this method work if customers also write in WhatsApp?

Answer: Yes, the measurement rules are the same, and E-chat connects number-based WhatsApp to the CRM too. Keep the limit on first messages in mind. According to the E-chat help center, a new WhatsApp account can open 5-10 new conversations a day, an older active one 20-30, and there are no limits within existing conversations.

Question: Do we need a separate number for each manager?

Answer: No. E-chat charges for connected numbers, and the number of managers is unlimited. Two managers and the owner can easily work with one Viber number and one Telegram number.

Question: How much does a Viber and Telegram setup cost?

Answer: Under E-chat pricing, a Viber number costs $29.99 a month and Telegram $16.99, $46.98 together. When you pay for several numbers for 12 months at once, the discount goes up to 40%.

Question: Our CRM isn't on the integrations list. What should we do?

Answer: Viber connects through ApiX-Drive or Wepster, while Telegram needs your own solution on the E-chat API, the documentation is in the service's help center.

Question: What happens to the chats when a manager quits?

Answer: They stay in the CRM, in the customer cards. An employee can't take the work number, export the database or move chats into personal messengers. A new manager opens the card and sees the full history with the customer.