Nobody in Bulgaria Uses WhatsApp: A Messenger Map for Entering a New Market
A launch plan for a new country usually contains one silent assumption: that messaging means WhatsApp. It is a reasonable default in most of Western Europe, Latin America and India, and it is wrong often enough that it deserves a check rather than a habit. Two markets where it is wrong in different ways are Bulgaria, where the dominant messenger is Viber, and Ukraine, where it is Telegram.
The practical problem is not that a plan names the wrong app. It is that the wrong app changes the cost model, the automation stack, the support workflow and the consent mechanics all at once. This piece is a map: what is actually measurable in each market, what is only claimed, what business messaging costs, and what the sensible first move is.
Start with what can be verified
Messenger market share is one of the worst documented numbers in digital marketing. Platforms with an advertising business publish audience reach because they sell against it. Platforms without one, Viber included, publish very little, and what circulates instead is vendor communication recycled through agency decks until it reads like a census.
The figure that always turns up for Bulgaria is that Viber reaches roughly ninety percent of the online population. It originates with the vendor. We could not locate a primary, dated, methodologically described publication behind it, and we are not going to repeat it as a fact. The direction is almost certainly right: Viber genuinely is the default messenger in Bulgaria, and anyone who has worked in the market will tell you so. The decimal places are not evidence.
What can be checked is platform reach reported by the platforms themselves and compiled with a stated method. DataReportal’s Digital 2026 report for Bulgaria puts the population at 6.70 million with 5.84 million internet users, an 87.1 percent penetration rate, and 4.50 million social media user identities, or 67.1 percent of the population, as of October 2025. Facebook’s advertising reach is given as 3.60 million people, Messenger’s as 2.80 million, which is 41.8 percent of the population. Viber does not appear in that series at all, which is precisely the problem.
Ukraine: Telegram is not a channel, it is infrastructure
Ukraine is better documented, and the numbers are unusually decisive. Ipsos research on Ukraine’s political, social and religious landscape, with fieldwork in November 2025 among 2,000 respondents aged 18 and over in government-controlled territory, found Telegram named as a source of news and information by 62 percent, ahead of television at 49 percent and YouTube at 46 percent. The same study records that no platform enjoys a trust level above 40 percent, which is a useful corrective to anyone about to treat reach as credibility.
DataReportal’s Digital 2026 report for Ukraine counts 35.3 million internet users at the end of 2025, at 89.6 percent penetration, with 23.0 million social media user identities in October 2025.
The consequence for a business is structural rather than cosmetic. In Ukraine a bot is not a novelty channel bolted onto a website. It is often where the customer already reads, asks and buys, which is why the payment rules inside that ecosystem matter so much. If any part of what you sell is digital, read our piece on why Telegram forces digital goods through Stars before you design the funnel, because it constrains the checkout and not just the copy.
What business messaging actually costs
The cost gap between these ecosystems is larger than the reach gap, and it is the part that most entry plans get wrong.
WhatsApp is the only one of the three with a fully metered business API, and the meter changed shape recently. According to Meta’s own pricing documentation for the WhatsApp Business Platform, from 1 July 2025 billing moved from conversations to individual messages: you are charged when a template message is delivered, at a rate that depends on the template category and on the country calling code of the recipient’s number. Volume tiers can lower the rate for utility and authentication templates. Service conversations have been free since 1 November 2024.
That free window is closing. Meta’s page on non-template message pricing states that effective 1 October 2026 service messages will be charged, at the same rates as utility and authentication messages for the market in question. Non-template messages can only be sent inside an open 24-hour customer service window, which resets with each user message, so the change hits exactly the reactive support traffic that most teams assumed would stay free.
Two operational points follow. First, the rate is set by the recipient’s country, not yours, so a rate that looked acceptable in a pilot market can double in the next one. Meta publishes per-country rate cards and revises them; read the current card rather than a blog post about last year’s card. Second, from October 2026 a support conversation has a marginal cost, which changes the economics of encouraging customers to message you at all.
Viber and Telegram price differently. Viber’s business messaging is sold through partners and priced per message by market and message type, which means the number you get depends on who you ask. Telegram bots are free to operate within generous limits, with charges appearing only in specific cases such as very high broadcast rates. That asymmetry is the single biggest planning difference between the three markets.
The map
| Market | Dominant messenger | What is measurable | Business messaging cost model | Practical entry move |
|---|---|---|---|---|
| Bulgaria | Viber, on vendor-reported dominance rather than audited share | 5.84 million internet users, Messenger reach 2.80 million, per DataReportal Digital 2026 | Per message through a Viber business partner, quoted per market and message type | Put a Viber contact route on the site next to the phone number, and do not make WhatsApp the only option |
| Ukraine | Telegram, with 62 percent naming it as a news source in the Ipsos November 2025 survey | 35.3 million internet users, 23.0 million social media identities, per DataReportal Digital 2026 | Bot operation free within platform limits; digital sales must run on Telegram Stars | Build the bot as a working channel, not a widget, and decide where the payment lands first |
| Typical Western European market | Meta publishes advertising reach and per-country rate cards | Per delivered template message since 1 July 2025, service messages chargeable from 1 October 2026 | Model the messaging cost per conversation before you promise a messaging-first support experience |
What actually goes wrong on entry
The contact block gets copied. A site launched in a new market keeps the same three icons it had at home. In Bulgaria that means offering a channel most visitors do not use and omitting the one they do. This is the cheapest possible fix and one of the most frequently skipped.
Localisation stops at the text. Translating the pages and leaving contact routes, working hours, phone formats and currency untouched produces a site that reads correctly and behaves foreign. The mechanics of doing multilingual properly, including the hreflang setup that determines whether the right version even gets shown, are covered in our guide to building a multilingual website.
Discovery is assumed to work the same way. Where a messenger dominates, a large share of recommendation traffic travels through group chats and channels rather than through search or feeds. That is one reason the search-like behaviour of social platforms is worth understanding on its own terms, which we set out in our piece on social media used as a search engine.
The local basics are left for later. For anything with a physical presence, the map listing is usually the highest-yield first week of work in a new market, ahead of the messenger question. Our walkthrough of local search and business profiles covers the order of operations.
A sequence that works
- Establish the real default, not the assumed one. Ask ten local customers or partners which app they would use to contact a company. That sample is unscientific and still more decisive than a vendor statistic.
- Price the channel before you commit to it. Estimate conversations per month, multiply by the current per-country rate, and add the October 2026 service message change if WhatsApp is in scope.
- Decide where money changes hands. In-messenger checkout, website checkout, or offline. This decision constrains everything downstream and is expensive to reverse.
- Instrument the channel like any other. Conversations started, first response time, resolution, conversion. A messenger that is not measured becomes an unowned inbox within a quarter.
- Give it an owner. The most common failure is not the wrong app. It is the right app with nobody responsible for answering it.
Where the messenger genuinely is the primary channel, the build question shifts from website to conversation design. That is the work behind our Telegram bot development practice, and the decision of what to automate and what to keep human is a business question long before it is a technical one.
The short version
Bulgaria runs on Viber and Ukraine runs on Telegram, and neither fact is optional information for a market entry. The Bulgarian dominance figure everyone quotes comes from the vendor and should be treated as a direction rather than a measurement; the auditable numbers are the DataReportal series, which put internet penetration at 87.1 percent and Messenger reach at 41.8 percent of the population. Ukraine is better evidenced, with Ipsos recording Telegram named by 62 percent as a news source in November 2025 and DataReportal counting 35.3 million internet users. WhatsApp is the expensive option and getting more so: per delivered template message since July 2025, and service messages chargeable from October 2026 at utility rates.
The map is worth drawing before the plan, because the messenger decision quietly sets the cost model for everything that follows it.
