Comments are the most expensive engagement you can buy and the easiest to get badly wrong. A like is anonymous and generic. A comment is a piece of text sitting permanently under your post, visible to everyone, and if it does not fit it damages the thing it was meant to help.
Some perspective from our catalogue. Instagram likes start at $0.09 per 1000. Comments start at $12.50 per 1000 and run up to $125.
That is a factor of roughly 140 at the low end. The reason is that a comment cannot be produced by a bot pool at scale without being obviously a bot pool at scale. Comments need either an account that will type something, or a text list you supply. Both are expensive compared with an automated tap.
The provider supplies the text from a generic pool. You pay per thousand and take what arrives. In our catalogue these include emoji-only comments at $12.50 per 1000, and language-specific pools (English, Russian, Ukrainian) from $55 to $125.
The generic pool is the problem. Random comments are the ones that read "nice pic 😍", "amazing!", "🔥🔥🔥" under a post about an office relocation. They are contextually blank, and on any post with actual subject matter the mismatch is visible.
Where they work: posts where a generic reaction is the appropriate reaction. A portrait, a product shot, a scenic photo. Emoji-only comments are the safest of the category precisely because an emoji carries no claim about content, so there is nothing to mismatch.
You supply the text. The service posts what you wrote, one comment per line. In our catalogue these run $25.03 per 1000 for the standard service and $37.50 for gender-targeted male or female variants.
Note what that means: custom comments are cheaper than most random-comment services here. The $25.03 custom service costs a fifth of the $125 English random pool. You are paying for delivery, not for someone to invent the words.
Custom is better in almost every case where you are willing to spend twenty minutes writing. You control relevance, tone, language and length, and you can write comments that ask questions or mention specifics that only someone who read the post would mention.
The failure mode is writing thirty comments that are all the same shape. Same length, same enthusiasm, same structure. That is a different tell from generic text but it is still a tell.
What helps:
The male and female custom comment services exist because the account posting matters as much as the text. A women's fitness brand whose comment section is entirely male accounts has a mismatch that no amount of good writing fixes, and vice versa.
If your audience skews strongly one way, matching it is worth the extra $12.50 per 1000 over the untargeted service.
Look at the maximums: 40, 50, 150, 200, 300, 1000, 1200 depending on the service. Compare with follower services that cap at 10 million.
That is not a limitation, it is the nature of the product. It is also a useful constraint, because the correct number of purchased comments is small. A post with 400 likes and 8 comments is normal. A post with 400 likes and 200 comments is not. The ratio is wrong by an order of magnitude, and comment-to-like ratio is one of the easiest things in the world to eyeball.
As a rule of thumb, comments should land somewhere between 1% and 4% of the like count on that post. Above that you are drawing attention rather than deflecting it.
An empty comment section is not a failure. It is the normal state of most posts on most accounts. The question worth asking before ordering is whether a comment section is the specific thing holding this post back, and usually it is not.
Comments are expensive enough that a failed order hurts. Check the measured average completion time and the 24-hour delivery chart on the service first. A comment service that has not delivered in a day should not be tested with your money. Minimums are 1 on every comment service in our catalogue, so a single-comment test costs almost nothing.
All eleven comment services, with live prices and delivery data, are on the services page. For how to read that data, see drop rate, average time and delivery charts explained.