Even allowing for the fact some Belarusians have multiple mobile phone numbers, as do some people who aren't eligible to vote for perfectly normal reasons, it's quite telling that somebody polled twice as many votes on a communication channel compared with the official 'count'.

What about those who are underage? There's a surely a large internet-savy base with phone numbers who would otherwise be unable to vote.

Underage people and foreigners are in the 'unable to vote for perfectly normal reasons' bracket. Extrapolating from census figures, you'd need pretty much every single person in the 13-17 age range to vote Tikhanovskaya to make up the disparity between Telegram and official tallies though. And if she got literally every teenager in the country to register a protest vote on a web app, there's a sneaking suspicion she might actually have got more than her 10% official tally with the adults too...

Forget percentages, even the raw numbers show a mismatch between official and Telegram results.