Forty Minutes, Four Cuts, Twenty Names
The DAMI creator library holds more than eight million profiles. That number is useful for exactly one thing — guaranteeing that you will never run out of candidates — and completely useless for deciding whom to contact. What turns eight million into twenty is not a better filter setting. It is the order of the cuts, and the judgment applied at each one.
This is the session Elena runs every Monday morning, and it takes about forty minutes. She manages creator partnerships across three beauty storefronts — a US shop, a UK shop running a reduced catalog, and a Canadian shop — out of Irvine, California. The catalog runs roughly 140 SKUs: serum and moisturizer, a hair treatment line, two sunscreen variants, and a rotating set of bundled kits. Her job is not to find creators. It is to produce, once a week, twenty names she can defend.
This article is for the operator in that position: US or multi-market, beauty or adjacent, somewhere between 50 and 200 SKUs, recruiting continuously rather than in campaigns. If you have ever opened a creator database, applied three filters, and still been looking at 4,000 names, the problem is not the filters — it is that you applied them in the wrong order and never made the judgment calls.
It also assumes you are recruiting continuously. A one-off campaign does not need a repeatable session; it needs one good list, and spending forty minutes a week on process is wasted if you are only running this twice a year. The DAMI creator library supports both, and this article only describes the weekly version.
Minute 0–12: Set Boundaries, Then Cut on Recency
Elena opens the DAMI creator library and sets three boundary filters before she looks at a single profile: country, category, and follower band.
Country first, because it removes the largest volume for the least thought. She runs the US shop and the UK shop as separate sessions rather than combining them, since a combined list makes every downstream comparison meaningless — a UK creator and a US creator with identical metrics are not the same opportunity.
Category second, and she keeps it tighter than feels comfortable. Beauty, personal care, and occasionally home. She does not add “lifestyle,” even though it would triple the list, because lifestyle is where the entertainment-audience problem lives.
Follower band third, split into two pools rather than one range: 8,000 to 50,000 for volume, and 50,000-plus for anchor partnerships. She screens the two pools separately because the acceptance criteria genuinely differ — smaller creators take product, larger ones take rate.
Then comes the cut that does most of the work, and it is not a category filter at all: shoppable content in the last 30 days. Elena applies this immediately, and it typically removes 40 to 60% of what survived the boundary filters. What it removes is the single largest category of waste in any creator database — accounts that look alive in a profile view, have healthy follower counts and decent engagement, and stopped producing shoppable content six months ago.
The reason recency goes before everything else is arithmetic. Every later judgment call costs a minute or two. Spending those minutes on a creator who is not currently selling is the most expensive mistake available, so it gets eliminated first, cheaply, by a filter.
Minute 12–22: Cut on Commercial Behavior
With the list down to a few hundred, Elena switches from filtering to reading. This is the pass where most sellers stop, and it is the pass that decides whether the program works.
| What she checks | Passing signal | Why it predicts orders |
|---|---|---|
| Shoppable posts, last 90 days | 4 or more | Occasional sellers do not build volume |
| Price band of products posted | Within 2x her AOV | Audiences have a budget, not just an interest |
| Categories actually sold | Beauty or personal care | Entertainment reach does not convert |
| Gaps in posting history | Nothing over 14 days | Dormant accounts dominate filtered lists |
| Evidence of past brand work | Samples or tagged products | Shows she follows through on commitments |
| Comment language | Matches the market | A local flag does not mean a local audience |
Price band is the one Elena wishes someone had told her about in her first year. A creator who reliably moves $19 lip balms has an audience with a $19 budget. Handing her a $68 serum does not expand that budget; it produces a video that converts at a fraction of the creator’s normal rate. Elena lost most of her second quarter to this before she started reading it, and she now treats a price-band mismatch as a hard cut rather than a maybe.
Comment language catches a subtler version of the same problem. An account with a US flag and a comment section in another language is frequently an audience built elsewhere, and the shipping economics do not work for it regardless of how good the content is.
The output of this pass is usually 40 to 60 names. Elena does not try to get below that here, because the next pass is the one that requires judgment rather than rules, and judgment needs candidates.
Minute 22–32: Read Post History for Category Fit
This is the ten minutes that cannot be filtered, and it is the reason the session is forty minutes long rather than four.
Elena opens post history for each surviving creator and asks one question: would a viewer of this content plausibly buy what I sell? Not “is this creator in my category” — that question was answered by a filter. The real question is whether the audience that watches this specific person is in a buying posture for this specific product.
A creator who posts skincare routines, product empties, and dermatologist reactions passes. A creator who posts travel, comedy, and the occasional skincare video fails, even though a category filter would keep her. Elena’s language for this is adjacent versus incidental: adjacent audiences convert, incidental audiences browse.
She is also reading for timing. A creator whose strongest shoppable content ran in January is a different partner in June than a creator whose strongest content ran last week. The DAMI creator library shows post history alongside the metrics precisely so this judgment happens in one screen — Elena does not open TikTok, and opening TikTok is what turns a forty-minute session into a two-hour one.
The second thing she does in this pass is cross-check against DAMI competitor mining. It pulls the creator list behind any storefront, and Elena runs it on four or five shops selling adjacent beauty products. A creator who appears on that list has already sold in her category, which removes the hardest part of the pitch — she does not need to be convinced that beauty sells on TikTok, she needs a better offer. Those names get marked and move to the top of the shortlist.
Minute 32–40: Build the Shortlist and Assign Tiers
The last ten minutes produce the deliverable: twenty names, each with a tier and a reason.
| Tier | Count | Creator profile | Opening offer |
|---|---|---|---|
| Anchor | 2–3 | 50k+ followers, proven in category | Commission rate, negotiable |
| Core | 8–10 | 15k–50k, adjacent audience, active | Sample plus standard rate |
| Test | 7–10 | 8k–15k, or adjacent but unproven | Sample only, no commitment |
Twenty is not arbitrary. Elena’s team can genuinely manage about twenty active conversations per storefront per week — enough follow-up passes, enough sample decisions, enough patience for the ones who ask questions. A shortlist of fifty would be abandoned by Wednesday, and an abandoned shortlist is worse than a small one because the good names get lost in it.
Each name gets one line written next to it: the reason. “Posted three sunscreen videos in May.” “Sells $40–60 range, matches our serum.” “Already on two competitor storefronts.” Those one-liners are what make the outreach message writable in ninety seconds each, and they are the only part of the session Elena keeps from week to week.
What the DAMI Database Cannot Tell You
Three honest limits, because a screening process that appears to answer everything is not trustworthy.
It cannot tell you whether a creator will post. Every field above is backward-looking. Past brand work is the best available proxy and it is still a proxy. Elena’s no-post rate on a vetted shortlist is about one in six, and no additional filter has reduced it.
It cannot tell you what a creator is currently being offered. Rates move, and a creator who accepted 12% last quarter may be holding a better offer this month. The database shows behaviour, not the competitive situation.
It cannot tell you whether your own product is the problem. If a well-vetted shortlist converts poorly across the board, the failure is in the offer or the product-market fit, and no amount of better screening fixes that. Elena learned this the expensive way with a sunscreen variant that no creator could sell, and the temptation was to blame the shortlist.
There is also a structural caveat: the DAMI creator library is a screening tool, not a sourcing strategy. If you have not decided which storefronts matter in your category, DAMI competitor mining will happily return creators from shops that sell to a completely different customer, and nothing downstream re-examines that choice.
Where the Twenty Names Go Next
A shortlist is worthless if it sits in a tab. Elena’s twenty names move into outreach the same morning, and the handoff is where most screening work dies.
The names go into DAMI bulk invitations with the daily cap set per storefront — 50 for the US shop, 20 for the UK, 15 for Canada. Those caps are below what DAMI permits, and they are deliberately below, because sending capacity has to be earned. The one-line reason written at minute 38 becomes the opening sentence of each message, with the body generated by DAMI from the tier: commission-forward for anchors, sample-forward for core and test.
Replies land in DAMI AI outreach management as a single thread view across all three storefronts, which is what lets the two-pass follow-up happen within hours instead of overnight. Accepted samples move to DAMI sample management, where shipment and publish date sit on the same record — the only way Elena learns her own sample-to-post rate rather than assuming it.
And the loop closes in the DAMI store data view. Every Monday, before the new session starts, she checks which of last week’s twenty produced orders. That number is what tells her whether any of the cuts above are still working, and it is the reason the session is worth repeating rather than a one-off exercise.
Frequently Asked Questions
How is this different from just applying filters? The filters get you from eight million to a few hundred. The judgments at minutes 12 through 32 get you from a few hundred to twenty, and they are the part that determines whether the twenty produce orders. Our filter guide covers the mechanics; this session is the sequence around them. Once the twenty are picked, the batch invitation guide covers what happens to them next.
Can I run this for more than one market in a single session? Not in one pass. Run each market separately — the boundary filters differ, the price bands differ, and the comment-language check is market-specific. Elena runs three sessions, ten minutes shorter each once the rhythm is established.
What if my category has very few active creators? Widen the category rather than lowering the recency cut. An adjacent category with active creators beats your exact category with dormant ones, which is why Elena occasionally screens home rather than forcing beauty.
Should the shortlist size change as the program matures? Yes, but slowly. Add capacity only when your follow-up pass is still happening within twenty-four hours at the current size. Most teams overestimate how many conversations they can hold, and DAMI will not tell you that you have exceeded it — a backlog of unanswered threads looks identical to a quiet week unless you check.
How often should the filters themselves be revisited? Quarterly. Elena reviews which cut removed the most names and whether any of them is now removing good ones — a filter that is too tight is as expensive as one that is too loose, and it is easier to miss.
Twenty Names, Every Week
The value of a creator database is not its size; it is whether you can reliably produce a shortlist you can defend. See how DAMI screens eight million profiles down to a weekly twenty on the product page, or create your account to run your first Monday session this week.
