What 15 Public SMM Panel Catalogues Listed, Read Three Times in a Month

Between 00:16:39 and 00:16:51 UTC on 2 September 2026 — a twelve-second window — we read the public service catalogue of fifteen SMM panels and recorded every row on every page: 44,064 listed services, each with its name, its rate per 1,000, its order limits, and whatever it said about start time, delivery rate and refill. Nothing was ordered. Nothing was converted or estimated. Every figure below was read off a page anyone can open without an account.

The twelve seconds matter more than they sound. These catalogues move — several panels stamp update dates on their categories — so reading all fifteen inside one window means the prices below are prices at a single instant, not prices gathered across an afternoon of crawling. Each page was stored exactly as fetched, with a per-URL timestamp, HTTP status and byte count, and all fifteen are published below along with the scripts that read them.

Eleven days later we did it again. Between 20:27:59 and 20:28:11 UTC on 13 September 2026 the same fifteen pages were read from the same host with the same script, and every listing was matched to itself by the panel’s own service id. Of the 43,269 listings present both times, 2,811 carried a different price. What moved, and why a single average would describe it badly, is in its own section below. On 2 October 2026 we read the same pages a third time and followed every listing across all three dates; that is in the section after it. Everything before those two sections describes the first capture.

This is the first measurement of our own that this site has published. Our editorial policy ranks four kinds of evidence in descending order of confidence, and direct observation — “what a panel’s service catalogue contains” — is the third of them, below primary platform documentation and published research. It is not the strongest evidence there is; it is the kind we had never produced. That is a fair criticism of everything written here before, and this page is the start of an answer rather than the whole of one.

Conflict of interest, stated before the numbers

Two of the fifteen panels in this survey — smmlaunch.com and adderpanel.com — are operated by the company that publishes this site. They are labelled as such in the underlying data and named as such below. They were included on purpose: dropping them would have removed the one commercial relationship in the sample we can verify from the inside, and would have produced a survey that quietly declines to measure its own publisher. Where they landed is the part that matters, and it is four answers rather than one. On Telegram channel members — the basket item the most panels carry — they are the tenth and eleventh cheapest of the twelve panels listing it, at $0.17 and $0.20 against a median of $0.05585. On the other three they are unremarkable. Telegram post views: smmlaunch.com is seventh of twelve, in a three-way tie at $0.0040, and adderpanel.com is tenth. Instagram followers: they are fifth and sixth of eleven, both at $0.4500 — and because eleven prices have a middle one, the $0.4500 median printed in the table below is their own price. YouTube views: smmlaunch.com does not list it, and adderpanel.com is fifth of ten at $0.3700, one of the two prices the $0.3800 median sits between. So the sample does not flatter them on the item most panels carry; on the other three they sit at or near the middle, and on two of the four rows the median a reader takes off the table is our own number or rests on it. The full commercial position is on our disclosure page.

The same service, fifteen shops, one night

Four services that most panels carry were picked as a basket: Telegram channel members, Telegram post views, Instagram followers and YouTube views. For each panel, the figure below is the cheapest listing that qualifies, after dropping rows that cannot honestly be compared — the exclusions are described further down, and they change the answer a great deal. Two of the fifteen panels are in the corpus but not in this table: growfastsmm.com prices in rupees, and peakerr.com publishes no order limits at all, which the eligibility filter requires.

The interesting number is not any single price. It is that three of the four ratios sit close together — 30.6×, 37.5× and 46.8× — across three unrelated products on three platforms, while the fourth is nothing like them: YouTube views is much tighter, at 11.2×. Where an asking price varies by thirty to forty-seven times, price is not tracking a difference the buyer can see. The cheap listing and the dear listing might genuinely be different things; the listings give a buyer no way to settle it. The reseller structure is the usual explanation for spreads of this shape.

The YouTube row is a caution about this kind of arithmetic as much as a finding. Its ratio is a quarter of the others because no panel in the comparable set lists a cheap YouTube view — and the moment one panel that publishes no order limits is let into the column, the cheapest listing drops to $0.05 and the ratio triples to 38× on the strength of a single row. An earlier version of this page printed that 38×. A spread is a statement about exactly two listings, and two listings are a thin thing to hang a market on.

How often a listing commits to anything

Excluding the one panel that prices in rupees, 43,812 services carry a USD price. Of those:

• 22.1% (9,662) name a refill or guarantee period — “30 Days Refill”, “Lifetime Guarantee”.
• 18.1% (7,932) say explicitly that there is none — “No Refill”.
• 54.9% (24,062) use no refill or guarantee word at all. The buyer is told nothing either way. Updated 8 October 2026: this overstates silence; see the note below.
• 44.1% (19,327) state a start time in any form.
• 33.4% (14,618) quote a delivery rate per day.
• 26.1% (11,441) print a duration in an average-completion-time column — “10 hours 21 minutes”, “0–2 hrs”. That column belongs to the panel and we did not audit how any panel produces it, so this is evidence that a number is published, not evidence that it is right.

The first three lines do not sum to 100%, and it is worth saying why rather than letting them look exhaustive. A further 2,204 services (5.0%) use the word “refill” or “guarantee” without naming a period and without saying there is none — most often a literal Refill: -, or the word sitting in a paragraph of boilerplate. Forty-eight more match the first two patterns at once and are counted twice. Neither group is a disclosure and neither is a refusal: they are the rows where the wording decides nothing.

The average-time line needs its own warning, because we got it wrong first. An earlier count treated every non-empty cell in that column as a measurement and put the figure at 53.8%. It is not one: 5,251 of 1xpanel.com’s cells and 6,769 of crescitaly.com’s read New Service, which is a panel saying it has no statistic yet — the opposite of a disclosure; six of peakerr.com’s eight read Instant, which is a promise; and all 123 of smmfolgen.com’s are a fault in our own parser, which captured the whole table row instead of the cell. Counting only cells that contain an actual duration halves the figure. That test is one function in the published script, and anyone can change it and re-run.

Those averages hide the real finding, which is that disclosure is close to a per-panel policy rather than a per-service one:

The range is the point. 1xpanel.com prints a start time on 99.7% of its 8,477 services; smmkings.com prints one on 0.4% of its 3,274 and prints a duration in the average-time column on none of them. Our own smmlaunch.com sits near the bottom of the refill column at 1.8%. Of the forty-five rows behind the price table above, 22 carried refill wording and 11 stated a start time. All forty-five state a maximum order, but that is not a finding — stating one is a condition of being in that table.

One bias runs through this table in a known direction. Five of the fifteen panels — smmlaunch.com, adderpanel.com, crescitaly.com, smmcost.com and smmkings.com — hide their service descriptions behind a JavaScript button that fetches over AJAX, which we did not click. For those five a claim counts only if it is written into the service name itself, so their true disclosure rates are higher than shown. The direction of that error is certain; its size is not. It affects no price.

What survives the bias is the advice in the buyer’s guide: separating a listing that commits to something from one that commits to nothing is free, takes seconds, and on most catalogues removes more than half the shelf.

The same name at four addresses, at four prices

Normalised for case, punctuation and emoji, the 44,064 listings reduce to 33,981 distinct names. 2,175 of them (6.4%) appear on more than one panel; 113 appear on three; 12 appear on four. Twelve of the fifteen panels share at least one verbatim name with another. Shared catalogue text is less surprising once you know that most panels are not separate products: in our survey of 76 readable panels, 54 ran the same vendor’s software.

Some are specific enough that coincidence is a weak explanation. Telegram Online Members [Uzbekistan] [Random Message] appears word for word on 1xpanel.com, easytopromo.com and both of our own panels, and those four also carry the Turkey, Israel, Iran, India, Germany, English, China and Arabic variants. The largest single overlap is between crescitaly.com and smmkings.com: 1,361 identical name strings — 41.6% of the 3,274 services smmkings.com lists, and 44.7% of the 3,046 distinct names those services reduce to.

Shared names do not mean shared prices. Of the 2,087 name strings priced by two or more panels, the median ratio between the highest and lowest price for the identical string is 1.89×, and only 20.2% carry the same price everywhere they appear. “Telegram BOT Start” is $0.13 on global-smm.com and $20.00 on crescitaly.com — 153.8× for the same words, in the copy of the first catalogue we stored and the copy of the second. On the crescitaly × smmkings pair, none of the 1,361 shared names carry the same price.

There is one exception in the sample, and it is ours. 93.3% of smmlaunch.com’s catalogue names also appear on adderpanel.com, and 97.1% of those shared names carry an identical price — 430 of 443. That is what a genuinely shared catalogue looks like in this data. It works as a control in both directions: the measurement can detect a real relationship, and it does not detect one between crescitaly and smmkings, whose 1,361 shared strings are priced entirely differently.

What the overlap cannot show is who supplies whom. Identical strings are consistent with a shared upstream provider and equally consistent with one panel copying another’s catalogue text. The survey measures the overlap and stops there.

A null result

We expected larger catalogues to disclose more, on the theory that a bigger operation has more process behind it. It is not so. The two largest catalogues sit at opposite ends: smmcost.com lists 9,476 services and states a start time on 51.9% of them, while crescitaly.com lists 8,477 and states one on 5.0%. The smallest real catalogue in the sample, smmfolgen.com with 123 services, discloses more than either — 47.2% start times, 48.0% delivery rates, 46.3% refill terms. Across the fifteen there is no visible relationship between how much a panel sells and how much it tells you. Disclosure looks like a house style, set per panel.

What a live price list contains that is not a price

159 listed services across six panels are priced at $100,000 or more per 1,000, and another 650 sit between $1,000 and $100,000. Inspected, they are mostly not products at all. morethanpanel.com lists a section divider, 👇👇👇👇👇👇👇👇👇 DİSCORD SERVICES 👇👇👇👇👇👇👇👇👇 — nine of that arrow on each side, and a Turkish dotted İ — as a priced row at $1,111,111,111,111.00, and carries 142 such rows; smmcost.com lists a row named only with a line of dashes at $1,000,000,000,000,000.00; 1xpanel.com prices a Telegram members row at $14,509,675,690,176.00; and one of the six panels is ours. smmlaunch.com publicly lists BAN Channel/Group/Account/Bot at $9,999,999.00, under a category it has itself named Ban(Delete!) Channel/Group/Account/Bot — a listing for getting somebody else’s channel deleted, on the publisher’s own panel, found by the publisher’s own survey. Every one of those rows is in the parsed service list, and the pages they came off are stored below.

A further 2,844 of the 44,064 rows (6.5%) quote a rate “per 1,000” while capping the maximum order below 1,000 — a price for a quantity the panel will not sell. And smmlaunch.com, our own, publicly lists a $0.01 Telegram Premium Members service under a category heading it has itself named “Dont use!” — one of 27 rows sitting in that category, priced from $0.0001 to $22.50, all of them visible to any visitor.

These are housekeeping artefacts rather than deception, and they are reported for one reason: they are why a naive reading of these catalogues produces nonsense. Without the filters described below, this survey’s own YouTube column would have shown a 19,000× spread — $0.0001 to $1.90 — instead of 11.2×.

Eleven days later: the same fifteen catalogues, read again

On 13 September 2026, between 20:27:59 and 20:28:11 UTC, we fetched the same fifteen services pages again — same host, same user-agent, same order, one request each — eleven days and twenty hours after the first reading. All fifteen answered with HTTP 200, and peakerr.com once more redirected to its homepage. The parser’s reading rules needed no change: all fifteen tables carried the same column headers as eleven days earlier.

The two captures were joined listing by listing on the panel’s own service id — the number an order is placed against, and so the panel’s own definition of “this service” rather than ours. The figures below cover the thirteen panels that price in dollars and publish ids; growfastsmm.com and peakerr.com stay out of every price figure, for the same reasons as before.

• 43,269 listings were present both times. Of the 43,804 listed on 2 September, 535 (1.2%) had gone, and 826 listings were new.
• 2,811 of the 43,269 (6.5%) carried a different price: 1,486 higher and 1,325 lower.
• The two directions are not mirror images. The median rise was 25% and the median cut 2%, but both medians lean on one panel: 816 of the 1,325 cuts are 1xpanel.com’s, most of them from a single 2% rule. Leave 1xpanel.com out and the median rise is 17.9% and the median cut 13.5%.
• Measured by factor, which treats rises and cuts alike, 276 prices at least doubled and 116 at least halved.
• Twelve of the thirteen panels changed at least one price; global-smm.com changed none. One panel, 1xpanel.com, accounts for 1,471 of the 2,811.

So a price read on 2 September was far more likely than not to be the same on 13 September: 93.5% were. Where one had changed, it was more likely to have doubled than to have halved. Two readings eleven days apart cannot say whether that holds for any other eleven days.

About a third of it looks like a rule, not a decision

When hundreds of listings on one panel move by the same factor, that is unlikely to be hundreds of pricing decisions. It is more probably one rule applied to a catalogue — a supplier’s rate passed through, or a margin changed — though the data shows the pattern, not the cause. Counting every group of ten or more listings on a single panel whose price moved by the same ratio, to three decimal places, 892 of the 2,811 changes (31.7%) fall into such groups.

The largest is 527 listings on 1xpanel.com, mostly TikTok, Facebook and Instagram services, starting from 159 different prices and all cut by 2% to the rounding of their prices. That is the clearest case of a proportional rule in the data. The second largest is a different kind of group: 163 YouTube likes listings on easytopromo.com, 151 of which went from $0.715 to $0.832 — one product line repriced together, not a rule across a catalogue. The threshold of ten is ours, and rounding splits one rule across neighbouring ratios (1xpanel.com has further groups at 0.981 and 0.982), so 31.7% is a lower bound on the share of changes that moved in step with ten or more others.

“Stable price”

1xpanel.com writes the words “Stable price” into the name of almost every service it lists: 8,271 of the 8,425 listings it carried on both dates. 1,460 of those 8,271 (17.7%) changed price in the eleven days. The largest move was an Instagram saves service that went from $0.0032 to $0.0539 per 1,000 — a rise of 1,584% — under a name that read “Stable price” on both dates. We do not know what the panel means by the label; it may refer to something other than the listed rate, such as what an order costs once it has been placed. What it does not describe is what the listed rate did. A further 134 listings on autosmo.com carry the same words, and eight of them changed price.

What the basket table did

The cheapest price in all four rows is exactly where it was on 2 September. Of the 46 panel-and-service cells behind the table, 32 kept the same listing at the same price, though not always under the same name or with the same order limits. The movement is concentrated, and the medians show how little it takes to move a median.

The Telegram members median nearly doubled on the strength of three of its twelve cells. mysmm.co raised the same listing, under the same name, from $0.053 to $0.362. autosmo.com withdrew both of its $0.03 listings, which left a $0.16 listing it already carried as its cheapest. And smmpanel.one’s service 92 went from $0.09 to $0.15 — but in the same eleven days its name changed from ♻️R3 Telegram Members ~20k/day to Telegram HQ Members ~100k/day and its order limits changed with it, so it is fairer to call that a different product behind an old number. Nine of the twelve cells did not move, three did, and the figure a reader would take off the table went from $0.05585 to $0.105.

The Instagram spread nearly doubled at the other end of the row. smmpanel.one’s cheapest qualifying listing went from $0.64 to $2.14, because all five listings it priced below $2.27 on 2 September, from $0.64 to $2.25, are now listed at between $2.43 and $3.18. That leaves a service that was $2.27 then, and is $2.14 now, as its cheapest. The same thing happened less dramatically on 1xpanel.com, crescitaly.com and mysmm.co: in each, the old cheapest Instagram followers listing was repriced upward, past another one. The YouTube spread widened for a reason that is about the definition as much as the market: smmkings.com, which had no listing matching the basket’s YouTube definition on 2 September, now lists a high-retention views service at $3.08, and it became the dearest in the row.

Two same-id moves in the Instagram row are worth reading on their own. smmkings.com raised an unchanged listing, Instagram Followers [Mixed - Max 50K] [No Refill], from $0.26 to $0.48. morethanpanel.com cut its service 5751 from $1.08 to $0.52 — and in the same eleven days the delivery speed printed in its name went from 5-10K/Day to 500-2K/Day. The price roughly halved and the promised speed fell by as much as a factor of ten, which is a reminder that a cheaper listing is not necessarily a cheaper version of the same thing.

Names that changed under an unchanged id

135 listings kept their id and changed their name, and 104 of those also changed price. Some renames only add an emoji; others change the product description. On mysmm.co, service 9CHI0W said [No Refill] on 2 September and [30 Days Refill] on 13 September, and its price went from $0.602 to $0.81. That is why the comparison joins on the id but labels a same-id move “renamed and repriced” whenever the name moved too, and why all 135 are published in renamed-listings.csv: whether a given id is still the same product is a judgement a reader can check for themselves.

Our own two panels, again

adderpanel.com is operated by the company that publishes this site, and two of its numbers stand out. It withdrew 211 of its 3,941 listings (5.4%), the largest share withdrawn by any panel in the sample. And the median change across the 207 listings it repriced was +31.25%, the largest median change of any panel that repriced more than one listing; 146 of those went up, 61 down, and 24 at least doubled. Its share of listings repriced, 5.5%, is close to the sample’s 6.5%, and none of its four basket prices moved. We did not ask its operator why, just as we asked no other panel. smmlaunch.com, the other, changed one price, on a listing named Id test 555, and added one listing, named Test.

That second row is why this survey now has a fourth eligibility filter. Filed under a Telegram members category at $0.15, Test would have become smmlaunch.com’s cheapest Telegram members listing and made our own panel look cheaper than anything it actually sells. Rows whose name begins with “test” are now dropped. One basket definition was tightened on the same day: autosmo.com files a service called Instagram Channel Members under an Instagram followers heading, and it had become that panel’s cheapest “followers” listing. Applied to the first capture, the two changes alter two rows of the published basket CSV — a count and a panel median in each — and no figure on this page; the method note lists both.

On 13 September the two are ninth and tenth cheapest of twelve on Telegram channel members, ninth and tenth of twelve on Telegram post views, and joint fourth of eleven on Instagram followers, both at $0.45; adderpanel.com is fifth of eleven on YouTube views. Neither the Instagram median nor the YouTube median rests on their prices any more.

What eleven days cannot tell you

Two readings give a difference, not a trend. Nothing here says whether 6.5% of listings repricing in eleven days is normal for these catalogues or unusual, whether the rises are seasonal, or whether any price change came with a change in what is delivered. A withdrawn listing may have been disabled, moved behind a login or renumbered, and the data cannot tell those apart. We said here that a third capture was the smallest thing that could begin to say whether any of this is typical, and we took one on 2 October; it follows.

Eighteen days later: a third reading, on 2 October 2026

On 2 October 2026, between 17:19:05 and 17:19:19 UTC, we read the same fifteen services pages a third time — same host, same script, same order, one request each — eighteen days and twenty hours after the second reading and a month after the first. This window was fourteen seconds rather than twelve. All fifteen answered with HTTP 200, and peakerr.com once more redirected to its homepage. One reading rule did have to change. morethanpanel.com no longer prints each service’s description in a pop-up; it carries the same text in an attribute of the row, so the parser now reads it there when the pop-up is missing. Without that, every description on that panel would have read as empty and its start-time disclosure would have fallen from 91.9% to 0.9% overnight. Run on the first two captures, the change produces exactly the rows it produced before, and it touches no price.

Followed across all three dates, nine listings in ten showed the same price on all three dates, and of those that changed in both intervals, almost half went back the other way. On the thirteen panels that price in dollars and publish ids, 40,870 listings were on all three pages. 36,593 of them (89.5%) carried the same price on 2 September, 13 September and 2 October. 1,534 changed only in the first interval, 1,775 only in the second, and 968 in both. Of those 968, 509 moved the same way twice and 459 reversed — 336 cut and then raised, 123 raised and then cut — and 19 ended on exactly their 2 September price. The reversals are not spread evenly: 293 of the 459 are on 1xpanel.com.

More than half of those 293 are the 2% rule from the first interval, and what came after it. Of the 527 listings 1xpanel.com cut by 2% between the first two readings, 497 were still listed on 2 October. 337 kept the cut price. The other 160 moved again, every one of them upward and every one to above its 2 September price; the most common move was a rise of 50%, on 48 of them. In the second interval only six 1xpanel.com listings were cut by 2%, counting the 0.980 to 0.982 ratios together — too few to count as a group. Small proportional rises did appear there instead: groups of 19, 18 and 12 listings went up by 2.3%, 1.9% and 2.2%. So the clearest proportional rule in the first comparison, a 2% cut, was not repeated, and for a third of the listings it touched it was a small dip before a larger rise. As before, the data shows the pattern, not the reason for it.

Taken on its own, the second interval looks like this, over the same thirteen panels and with the same definitions as the first:

• 41,646 listings were present on both 13 September and 2 October. Of the 44,095 listed on 13 September, 2,449 (5.6%) had gone, and 1,053 were new. Two panels account for most of the withdrawals: 1xpanel.com, 1,124, and our own adderpanel.com, 1,045.
• 2,845 of the 41,646 (6.8%) carried a different price: 1,991 higher and 854 lower. In the first interval rises and cuts were close to even, 1,486 against 1,325. In the second, seven changes in ten were rises.
• The median rise was 35.7% and the median cut 22.9%. 1xpanel.com again repriced more listings than any other panel, 1,174; leave it out and they are 28.1% and 26.6%.
• By factor, 142 prices at least doubled and 201 at least halved — the other way round from the first interval, when doublings outnumbered halvings 276 to 116. 106 of the 201 halvings are on morethanpanel.com, and 62 of those are one group of listings that all started at the same price and were all cut to about a sixth of it.
• Eleven of the thirteen panels changed at least one price. global-smm.com again changed none, and neither did our own smmlaunch.com.
• Groups of ten or more listings on one panel moving by the same ratio account for 633 of the 2,845 changes (22.2%), against 31.7% in the first interval. The same caveats apply: the threshold is ours and the share is a lower bound.

The two intervals are not the same length — 11.8 days, then 18.9 — so their totals should not be set side by side as if they were. Over the 40,870 listings present all three times, 6.1% changed price in the first interval and 6.7% in the second: 0.52% of listings a day against 0.36%. Even that is not a like-for-like rate. A listing counts once per interval however many times it moved inside it, and a longer interval hides more of those moves, so both figures are floors and the second is the lower one. What the data does show is that an interval more than half as long again did not produce proportionally more changed prices.

“Stable price”, a second time

1xpanel.com’s label behaved as it did before. Of the 7,218 listings carrying “Stable price” that it listed on both 13 September and 2 October, 1,168 (16.2%) changed price, against 17.7% in the first interval. Followed across all three dates, 1,919 of the 7,283 listings that carried the label on 13 September — on 1xpanel.com and autosmo.com together — changed price at least once in the month. On autosmo.com, 4 of its 134 changed in the second interval.

The basket on 2 October

Of the 46 panel-and-service cells behind the table, 31 kept the same listing at the same price. Two single cells did most of the visible work. The Telegram post views ratio quadrupled because smmfolgen.com raised its cheapest qualifying listing, under an unchanged name, from $0.02 to $0.12; a spread is a statement about two listings, and this one moved on one. The Telegram members median moved from $0.105 to $0.15 because smmkings.com raised its cheapest listing, also under an unchanged name, from $0.06 to $0.15, which put a $0.15 price on both sides of the middle of the row.

morethanpanel.com’s service 5751 is worth following through all three readings. On 2 September it read Speed: 5-10K/Day | Max 100K at $1.08. On 13 September it read Speed: 500-2K/Day | Max 100K at $0.52. On 2 October it reads Speed: 5-10K/Day | Max 500K at $0.24. One id, three names and three prices in a month, and the cheapest of the three promises the same speed as the dearest. Whether that is one product being repriced or three products sharing a number is not something a catalogue can settle.

Our own two panels, a third time

adderpanel.com, operated by the company that publishes this site, withdrew 1,045 of the 3,732 listings it carried on 13 September (28.0%) and added two. That is the largest share withdrawn by any panel in the sample for the second interval running; the next is smmfolgen.com at 17.2%. The withdrawn listings come from many headings, the largest being Facebook reactions (27), WhatsApp channel reactions (20) and, tied at 15 each, Instagram views and impressions and Twitter tweet views. Of the listings it kept, 131 (4.9%) changed price, 60 up and 71 down, with a median change of −3.2%. Among the withdrawals was its $0.45 Instagram followers listing, so its cheapest qualifying Instagram followers listing is now $1.65, ninth of ten. As before, we did not ask its operator why.

smmlaunch.com changed no price. It withdrew eight listings, all of them under its Instagram heading — on 2 October it lists no Instagram services at all, so it drops out of the Instagram followers row. And it renamed one listing, which bears on the filter described above. Service 1, the $0.15 row named only Test that prompted the fourth eligibility filter on 13 September, now reads Euro🇪🇺 [Fast] Fake Members [90 days non drop], at the same $0.15, with a minimum order of 100 and a maximum of 10,000. It now passes every filter and is smmlaunch.com’s cheapest Telegram members listing. The filter kept a test row out of the 13 September table; it does not keep a product out because it was once a test.

On 2 October, on Telegram channel members, smmlaunch.com is one of three panels at $0.15, joint sixth to eighth of twelve, and the $0.15 median in the table above is that shared price; adderpanel.com is tenth at $0.20. On Telegram post views smmlaunch.com is seventh of twelve at $0.004, one of the two prices the $0.0035 median sits between, and adderpanel.com is joint ninth at $0.01. On Instagram followers adderpanel.com is ninth of ten; on YouTube views it is fifth of eleven at $0.37. On 13 September neither the Instagram nor the YouTube median rested on their prices; on 2 October smmlaunch.com’s $0.004 is one of the two prices the Telegram post views median sits between, and its $0.15 equals the Telegram members median — though that median would be $0.15 without it, since smmkings.com and smmpanel.one are also at $0.15.

What three readings cannot tell you

Three readings are three points, not a series. A price that changed and changed back between two of them reads as unchanged, so every count of “changed” here is a floor. Nothing was ordered: these are listed prices on three dates, and nothing here says whether a price change came with a change in what is delivered. A withdrawn listing may have been disabled, hidden behind a login or renumbered, and the data cannot tell those apart — which applies to our own panels’ withdrawals as much as to anyone’s.

What this supports, and what it does not

It supports statements about what fifteen public catalogues listed, in USD, at 00:16Z on 2 September 2026: how far apart their asking prices were, how often a listing named a start time, a delivery rate or a refill period, and which catalogues shared verbatim service-name strings with which others. With the second capture it also supports statements about what those same catalogues changed in the eleven days to 13 September — which is a difference between two dates, not a trend — and with the third, what they changed in the eighteen days to 2 October and which listings moved in both intervals: three points, still not a series.

It does not support anything about delivery. Not one order was placed. A listed price is an asking price; a listed “30 Days Refill” is a promise, not an outcome; a listed average time is the panel’s own unaudited statistic. Nothing here shows that the cheap panels are worse or the dear ones better. It is not a ranking and it names no panel as a good or bad place to spend money — including the two we own.

Two further limits are structural. This is a sample of panels that choose to publish a catalogue publicly; those keeping prices behind a login — including justanotherpanel.com, one of the largest — are absent by construction, and their absence may bias the spread. And prices at 00:16Z on one night are not prices in general; the second capture measures how far they had drifted eleven days on, and two readings still cannot say whether that drift is typical.

How the sample was built

One hard rule governed selection: the full catalogue had to be readable at a public URL with no account, no login and no form submission. Candidates came from our own two panels, two Google queries, and the outbound links of a public panel directory; each candidate’s services page was fetched once and tested for a machine-readable price table.

Panels were dropped for stated reasons: justanotherpanel.com renders a login form; smmfollows.com, yoyomedia.in, n1panel.com and bulkfollows.com return marketing copy with the catalogue loading after authentication; smmbin.com and b2bsmmpanel.com returned 403; top4smm.com and smmheaven.net returned 404 or nothing; socpanel.com is a directory that would double-count other panels’ services; bharatsmmpanel.com published nine services, too thin to characterise. growfastsmm.com is kept in the corpus but excluded from every USD comparison because it prices in rupees and converting would need an exchange rate we did not observe. peakerr.com is kept in the corpus but excluded from every price comparison. Its /services URL redirected to the site’s homepage — the fetch log records both the URL requested and the one that answered — and what came back is a marketing summary of eight headline services with no service IDs, no categories and no order limits. Those are advertised prices, not catalogue entries, and in the first version of this page one of them was the cheapest YouTube listing in the table above. It is still counted in the claim percentages, where it is a catalogue of eight like any other.

Three eligibility filters were applied before choosing each panel’s cheapest qualifying listing, and each removes a way of producing a fake bargain. Rows that do not publish a maximum order of at least 1,000 were dropped, because a rate “per 1,000” is not comparable if 1,000 cannot be bought, and because a missing maximum is not a large maximum. The first version of this filter tested only whether a stated maximum was too small, so it admitted every row on the one panel that states no limits at all — precisely the case the rule exists for. Rows filed under a category named for disuse — “don’t use”, “deprecated”, “disabled”, “test only” — were dropped, which is what removes our own $0.01 row before it became this survey’s headline low price. Rows whose name is an instruction rather than a product (“ask me”, “make ticket”) were dropped, which removes a $0.0001 listing with a minimum and maximum order of one unit.

Claim fields were quoted from the page, not paraphrased, and anything not found was recorded as “not listed” rather than guessed. The order in which refill patterns are matched is deliberate, because the qualifier carries the meaning: “No Refill” and “30 Days Refill” are read before the bare word “refill”.

The data

Everything this page is built from is published under novamya.com/data/panel-price-survey/ — the fifteen catalogues as fetched, the fetch log, the parsed service list, the basket table and the three scripts. The README lists the lot; the table below says which figure comes out of which file. The first version of this article made the same promise while the files sat in a temporary directory on one laptop, which left the one claim the page exists to support as the only claim on it a reader could not test. Several figures above changed when somebody finally tested them.

We do not link to the panels themselves. This is a survey and not a directory; the pages we read have already moved, and the version we measured is the one published here.

Licence and reuse

Everything under novamya.com/data/panel-price-survey/ — the raw catalogues from all three captures, the parsed rows, the aggregates, the comparison, the fetch logs and the four scripts — is released under Creative Commons Attribution 4.0. Reuse it commercially, republish it, build on it, correct it; the one condition is that you say where it came from. Cite it as:

Novamya (2026). What 15 public SMM panel catalogues listed on 2 and 13 September and 2 October 2026. Dataset. https://novamya.com/data/panel-price-survey/

The standing rules this survey was collected under — public pages only, robots.txt respected, raw captures retained, our own panels in the sample and labelled — are set out on the research and data page, along with everything else we have measured. A licence is stated here because data without one is data nobody responsible can reuse, and a survey published so that others can check it should not then make checking it a legal question. If you are a researcher and need this in a different shape — a different serialisation, an earlier or later capture, the intermediate parse — write to editor@novamya.com and we will produce it.

If any of it is wrong we want to hear it: corrections to this page will be dated and logged on the corrections page, and the wider argument these figures feed into is in the social media growth guide.