Cold Telegram Templates That Don't Read as Templates
Short answer: three lines, one specific detail proving you looked, one question answerable with yes. Five working shapes below - adapt the brackets, keep the skeleton.
Why templates fail
Not because templates are evil, but because recipients recognize them in three words. "Dear Sir, I hope this message finds you well" announces automation before the pitch starts. The fix is not abandoning templates - it is templates with mandatory customization slots that force a human glance per send.
Five shapes that work
- The observer: "Saw your post in [group] about [specific topic]. We help [niche] with [problem] - worth a 10-min look this week?" Specificity does the heavy lifting.
- The filter: "Quick check so I don't waste your time: [dealbreaker question]? If yes, I have exactly one option worth seeing." Respects their time visibly.
- The referral ask: "Not a pitch - do you know who handles [problem] at [company]? Two-minute pointer would help." Low pressure, high response.
- The follow-up (day 4): "Still [problem]? A [new relevant thing] came up yesterday. Want the short version?" New information, not "bumping".
- The breakup (day 9): "Should I close your file on this?" Short, honest, no guilt. A surprising share answers here.
Personalization at scale without lying
Real personalization costs two minutes per lead: open their group post history, find one specific detail, reference it. At fifty leads a day that is over two hours - so budget it explicitly or it silently becomes fake ("loved your recent post!" with no post named). Middle path that holds up: segment-level personalization. One sharp opener per niche-segment, with a single custom slot (their post topic, their city, their tool stack). Reads personal, costs thirty seconds. Anything claiming full personalization at bulk volume is either lying or burning operator souls.
More patterns with commentary: https://www.24go.site/blog/shablony-kholodnykh-soobschenij-v-telegram?locale=en - and how leads get collected before you write: https://www.24go.site/blog/kak-sobirat-lidy-v-telegram?locale=en
Verify with reply quality, not rate
Track which template produces calls, not just replies. A template with 30% reply rate and zero calls loses to one with 10% replies and three calls. Curiosity replies ("what is this?") flatter metrics and waste mornings - optimize for intent signals, not politeness.
Micro-test new shapes in one evening
Never rewrite the whole sequence on a hunch. Take twenty fresh contacts, split ten-ten old versus new opener, send both, compare replies next morning. Small samples lie, but directionally they kill obviously dead ideas before they infect the main sequence. Promote winners only after a full week at volume - Tuesday luck has ended more good templates than bad copy ever did.
Honest limits
Templates decay as audiences learn them. Today's fresh opener is next quarter's recognized spam - rotate quarterly and retire anything dropping two months straight. And templates never fix targeting: the best shape sent to the wrong group converts at zero. Half your time belongs to group selection, half to copy. Most teams spend 5% and 95% and blame the words.
Which template shape earned your last real call - and how old is it?
Top comments (0)