Sourcing US Members With Shopping Intent for a Deals Group

Your group has 20,000 members. Your deal posts pull 400 views and two clicks. The invoice said the delivery was complete, and technically it was - the members are sitting right there in the member list, doing nothing.

That gap is not a fraud problem. Most of the time the accounts are real people. They are just the wrong real people, in the wrong country, with no reason to care that a Ninja blender dropped to $79.

The number that pays you is clicks per post

Affiliate revenue from a Telegram group decomposes cleanly:

revenue = posts × view rate × CTR × conversion rate × commission per order

Member count appears exactly once, buried inside view rate, and it is the weakest term in the chain. A 5,000-member group where 35% of members open a deal post produces more clicks than a 20,000-member group at 3%. You cannot buy your way out of a bad view rate with volume, because the two are inversely related when the extra volume comes from people who muted you on day one.

The term that actually moves is CTR, and CTR is a function of two things you specify at purchase time: whether the member can buy the thing (geography - Amazon.com links are useless to someone in Jakarta) and whether the member came from somewhere that suggests they buy things (intent). Everything else in your operation - post timing, deal quality, copy - swings CTR by maybe 2x. Geography and intent swing it by 10x.

Telegram has no country field

This trips up everyone briefing a supplier for the first time. There is no country attribute on a Telegram user. Phone numbers are hidden by default and, even when visible, are visible only to contacts. The language_code field exists in the API but is only delivered to bots, and only after the user interacts with your bot - you never see it for a member who simply sits in a group.

So when a supplier says "US members," they are not applying a filter. They cannot. They are making a claim about where the accounts came from. That distinction matters because it tells you what to ask for: not "US targeting," which is unverifiable at the API level, but the specific sourcing pool, which is.

What you can actually verify after delivery

Four proxies, in descending order of usefulness:

1. Activity-hour histogram. Post at 03:00 Eastern and at 19:00 Eastern for a week and compare view accrual in the first 30 minutes. A genuinely US-weighted group shows a sharp evening peak and a dead overnight window. A South-Asian-sourced batch shows the inverse.
2. Bot-captured language_code. Put a low-friction bot interaction in front of something people want - a /deals command, a restock alert opt-in - and log language_code and is_premium for everyone who touches it. Self-selected sample, but it is real data.
3. Display-name script. Export the member list and count non-Latin scripts. A US-shopper cohort is not 100% Latin script, but it is not 40% Cyrillic either.
4. Link-click geography. Route deal links through a short-link service that reports country. Even at a 1% CTR you will have a country distribution inside two weeks.

Run these before your second order, not after your fifth.

Intent is a property of the source, not a checkbox

One line in a supplier brief does more work than all the others: the list of source groups. Naming them yourself moves you from buying a demographic to buying a behaviour.

Sources that carry shopping intent, roughly in order:

• Retailer-specific deal groups (Costco, Target, Woot-style clearance communities)
• Cashback and rebate communities - these people have already installed an extension and linked a card
• Restock and drop-alert groups for a category you sell into
• Coupon-code groups for named brands

Sources that look adjacent and are not:

• Airdrop, faucet and giveaway groups. These members have been trained by every prior group they joined that the correct response to a link is to claim something free and leave. They will inflate your member count and destroy your CTR at the same time.
• Generic "money making" and "side hustle" groups. Intent to earn is not intent to spend.
• Follower-exchange and mutual-sub groups. Every member is there to be counted, by definition.

The trade-off is real and you should price it in: named source groups give you tighter intent but a smaller reachable pool, because the enumerable member list of any single group is capped and a large share of those members have privacy settings that block being added at all. If you name three sources and need 10,000 members, you may be asking for more than those three groups can yield. Naming eight to ten sources across the same vertical solves this without diluting intent.

What a brief that works looks like

Field	Weak brief	Brief that gets what you want
Geography	"US members"	"Sourced from US-only retail deal communities; I will verify with an activity-hour histogram in week one"
Source	"Anything in shopping"	Ten named t.me links, all US retail or cashback
Destination	"My channel"	"My group" - adds only work on groups; channel subscriber lists are not addable at scale
Pace	"ASAP"	"Max 800/day, spread across the day" - a 10k spike in 24 hours reads as an attack to Telegram's anti-spam and to every human who opens the group
Retention	Unstated	"Measure at 72 hours and 14 days; replacement on drop below X%"
Exclusions	Unstated	"No airdrop, giveaway, or follower-exchange sources"

That third row is the one people get wrong most often. Telegram lets you add users to a group; a broadcast channel past its early-growth stage effectively only accepts joins through invite links, and its subscriber list is hidden from everyone but admins. If your affiliate operation lives in a channel, the honest version of this purchase is: fill a group, run the group as your engagement layer, and cross-post to the channel. Anyone quoting you a per-1,000 rate for channel subscribers is describing a different, and much less reliable, mechanism.

This is the shape of what we sell, so treat the next two sentences as disclosure rather than a pitch. TeleReach fills a Telegram group either from source groups you name or from our own US online-shopper pool, at $60 per 1,000 with volume rates above that, and we quote against the constraints above rather than around them - including telling you when ten named sources cannot physically produce the quantity you asked for.

Worked example: what 10,000 sourced members earn

Assume a US electronics-and-home deals group, three posts a day, Amazon Associates in a 3% commission category (rates vary by category and change; check your current schedule), $42 average order value, and Amazon's standard 24-hour cookie window.

Targeted cohort - 10,000 members at $56/1,000 = $560

Stage	Rate	Result
Post views	28% of members	2,800
Click-through	1.6%	45 clicks
Conversion (24h window)	3.5%	1.57 orders
Commission	3% of $42	$1.26/order
Per post		$1.98
Per month (3/day)		$178

Payback lands around month three, and everything after that is margin on an asset you already own.

Untargeted cohort - same 10,000 members, same price

View rate 9%, CTR 0.2%. That is 900 views, 1.8 clicks, 0.06 orders, roughly $0.08 per post - $7 a month. Payback: never, in any timeframe that matters.

Same headcount. Same invoice. A 25x difference in return, decided entirely by a field in the brief.

One caveat on the Amazon side that catches deal operators repeatedly: the Associates programme requires you to declare where you promote, and the approved social platform list in the US programme does not include Telegram. Read your operating agreement before you build a business on it. Plenty of operators route Telegram traffic through their own site first, or use networks whose terms explicitly permit messaging apps - Impact, CJ and Awin advertisers set their own placement rules and cookie windows individually, and many are fine with it. Check per advertiser; the network does not set that policy.

What usually goes wrong

You measure on day one. Every batch shows some fall-off. Judging at 24 hours tells you nothing except that Telegram's own cleanup hasn't run yet. Measure at 72 hours and again at 14 days, and write both numbers into the order before delivery starts.

You order the whole quantity first. Buy 1,000 from a new supplier, run the full verification pass, then scale. The $60 you spend proving the sourcing is the cheapest money in this entire exercise.

You name sources you have never read. If you have not scrolled a source group's last 200 messages, you do not know whether it is a shopping community or a link-dump farm with a shopping-sounding name. Half of them are the latter.

You post to a room with no floor of native activity. Ten thousand new members landing in a group with four messages in it will mute in bulk. Have 30-50 real posts up first, ideally with a few genuine conversations, and stagger delivery so the group never looks like it doubled overnight.

You treat retention as the supplier's whole job. Sourcing controls who arrives. What they see in the first ten minutes controls who stays. A pinned message that reads like a spec sheet is a retention problem no supplier can fix for you.

Next step

Pick three deal posts you have already published, pull the view counts, and calculate your current view rate against member count. If it is under 15%, your problem is the composition of the audience you already have, not the size of it - and buying more of the same is going to make the ratio worse.

Then write ten source-group links into a brief and price a 1,000-member test against them. Get a quote and name your sources - the test costs $60 and it answers the question your last three months of posting could not.