Product seeding, done so it is not just free stuff in the post

Product seeding means sending your product to creators as a gift, with no fee, no contract and no obligation to post anything. The creator keeps it either way. Some of them will post because they genuinely liked it, most will not, and the ones who do produce something more convincing than a paid post because the enthusiasm is real. Done properly it is the best use of a few hundred dollars in this entire category, and it is the only tactic here that works when you cannot afford to pay anybody. Done badly it turns into a request for free labour dressed up as a gift, which damages your name among exactly the people you were trying to reach. The difference between the two versions is almost entirely in the message you send and what you expect back. This page covers when seeding works, what response rate to plan for, how to choose who receives a parcel, the wording that keeps it clean, and the disclosure rules that apply the moment a free item changes hands.

What seeding is, and what it quietly is not

The definition is narrow and the narrowness is the point. You send the product. The creator owes you nothing. That is the whole arrangement, and every problem in this tactic comes from somebody blurring it.

The right-hand column is a paid campaign with the payment removed. It is also where the reputational damage happens, because creators talk to each other and the brands that do this get named. If you want the right-hand column, book it and pay for it.

One more boundary worth stating early. Anything a creator posts about a gifted item belongs to them. Sharing it to your own Stories with credit is normal courtesy and generally welcomed. Putting it behind ad spend is a licence you have not bought.

What actually comes back

Set the expectation before you pack the boxes, because a seeding round judged against the wrong number looks like a failure when it was fine.

Assume most parcels produce nothing visible. A useful planning assumption for a well-chosen list of fifteen is three or four posts, several Stories that vanish in a day, and a handful of people who simply keep the product. Anything above half the list posting means either your list was unusually good or your product is unusually photogenic. Neither is the normal case and you should not build a forecast on it.

Cost the round honestly too. The real number is not the retail price of the product, it is what the unit costs you plus postage plus the hour it takes to pack and the hour it takes to write fifteen individual messages. For most small brands a fifteen-parcel round lands somewhere between three hundred and eight hundred dollars all in, which is why this is the tactic that fits under a thousand.

When seeding is the right call

• The product is visual, demonstrable or consumable, so there is something to show.
• It is worth having on its own terms. Nobody makes content about a keychain.
• You want early content and early reviews more than you want a launch-day spike.
• You can afford to give away twenty units and learn something either way.
• You have time. Seeding moves at the pace of other people's lives, not your calendar.

When it is the wrong call

• You have a fixed launch date and need coverage on it.
• The unit cost is high enough that giving away twenty of them hurts.
• The product needs weeks of use before anyone can say anything true about it.
• You need the content for your ad account, in which case buy it properly and own the rights.

Choosing who gets a parcel

The list decides the outcome, and most brands build it backwards by sorting on follower count. Sort on three things in this order: would this person plausibly buy the product, does their audience engage, and only then how large is it.

On the second question, the 2026 Creator Benchmark Report, covering 3,125 creators, gives you the shape of a normal engagement rate by size. On TikTok the median runs about 4.41 per cent for nano accounts and falls to about 1.57 per cent by the hundred thousand to five hundred thousand band. On Instagram it runs about 2.66 per cent at nano and falls to about 1.27 per cent in that same band.

Two practical conclusions. First, judge an engagement rate against the normal band for that account's size rather than against a flat threshold, or you will reject good nano creators and accept weak larger ones. Second, the floor sits in the mid-tier band rather than at the very top, which is a useful corrective to the assumption that engagement simply declines forever as accounts grow.

For a seeding round specifically, weight towards nano and small micro accounts. They are the most likely to post about something they were sent, because a parcel from a brand is still a notable event rather than a Tuesday.

Read the last twenty posts before adding anyone to the list. You are checking two things software will not tell you: whether the comments are real conversation or the same six accounts every time, and whether the rest of their content is something you want your product sitting next to.

What to say, so you are not implying an obligation

The wording matters more than anything else in this process. Every phrase that hints at a trade turns a gift into an unpaid brief in the reader's mind, and they will price it accordingly by ignoring you.

A message that works

Say who you are in one line. Say one specific thing about their content that made you pick them, and make it genuinely specific or leave it out. Say what you would like to send and why you think it suits them. Then say plainly that there is no expectation of a post, that the item is theirs regardless, and that you would only ask them to add the disclosure if they do decide to share it. Ask for a shipping address. Stop there.

Phrases to delete

• "In exchange for" anything. That sentence turns a gift into a contract nobody is being paid for.
• "We would just need one Reel and a few Stories." That is a deliverable list, which means you are booking work.
• "Let us know when you are planning to post." This is the follow-up that ruins an otherwise clean round.
• Anything about a launch date, because it communicates a deadline you have no right to set.
• "Exposure to our audience" as a stated benefit. Creators have heard it and it reads as a substitute for money.
• A discount code presented as the payment. A code is a sales tool, not compensation.

One follow-up is acceptable, and only to confirm the parcel arrived. If someone does post, reply properly, thank them, and put them on the list of people to pay next time. That last step is how a seeding round turns into a roster.

Disclosure applies to gifts, not just to payments

This is the part brands most often get wrong, usually because they assume that no money means no advertising. That is not how either regulator sees it.

In Canada, Ad Standards requires a disclosure that is clear, prominent and close to the claim, not buried at the end of a hashtag block or hidden behind a more link. Separately, the Competition Act treats an undisclosed paid endorsement as misleading advertising, which puts it in territory with real consequences rather than guideline territory.

In the United States, the FTC endorsement guides apply to any content reaching United States audiences, and they extend liability to the agency alongside the brand. So if we run a round for you and the disclosures are wrong, that is our problem too, which is a reasonable reason to trust the check.

• Free product counts as a material connection whether or not a single dollar moved.
• The disclosure belongs at the start of the caption and in the spoken or on-screen content for video, not only in a platform toggle.
• Plain words work best: gifted, or sent to me by the brand. Ambiguous tags like sp or collab do not meet the standard.
• It is your job to tell the creator this when you send the product, not to discover it afterwards.
• A creator who declines to disclose properly is not someone to send a second parcel to.

None of this is legal advice, and none of it is optional either. If a round is going to reach audiences in both countries, write the message so it satisfies the stricter reading and stop worrying about which rule applies.