When to re-book a creator vs. find a new one

Almost every founder I talk to over-indexes on creator discovery and under-indexes on creator retention. They run a campaign, get a result, and immediately go back to the top of the funnel, new shortlist, new outreach, new creators, new briefs. Six months later they have 40 first-time collaborations and zero creators who actually know their brand.

The brands that compound on creator spend do the opposite. They treat their first cohort of creators as a tryout, and their roster as the thing that produces leverage. The decision they make well, and most brands make badly, is whether to re-book a creator who delivered, drop them, or wait and watch.

Here’s the framework I walk founders through. It’s a 2x2 you can run on a deal in under 10 minutes, and the answer it produces is usually clearer than founders expect. There’s a tracker template at the bottom of the post, copy it into Sheets or Notion and run your whole roster through it.

The two axes that matter

Every creator deal produces two outputs worth measuring, and the right re-book decision falls out of where those two outputs land.

The first axis is performance against the conversion target you actually care about. For most small brands that’s CAC, sometimes conversion rate, occasionally direct revenue. Pick the one your business runs on and stick with it across all deals in the cohort, otherwise you’re comparing apples and oranges. Engagement rate is not on this axis; it’s a vanity proxy. A 14% engagement rate that produced $0 in attributable revenue is a “below target” deal.

The second axis is how the working relationship felt. This sounds soft and isn’t. It rolls up four concrete things, scored 0/1 each:

Did the creator respond within 24 hours throughout the deal?
Did the first draft hit the brief on the first pass?
Did they handle revisions without drama?
Did they hit the deadline?

Score 3–4 = “smooth.” Score 0–2 = “rough.”

Score each axis as “above target” or “below target” + “smooth” or “rough” and you get four cells. The actions for each are very different.

Cell 1: Above-target performance, smooth relationship → re-book immediately

This is your highest-value creator and the decision is not actually hard, but most brands still fumble it. The right move is to re-book within 30 days with an iterated brief. Same creator, new angle. Not a new product, not a totally different positioning, just a fresh creative angle on the same product.

The reason for the 30-day timer: momentum. The creator’s audience just got exposed to your brand. The optimal time to land the second message is while the first is still fresh in the recommendation graph, not six months later when you’re effectively introducing yourself again. Brands that re-book their winners within 30 days see the second post outperform the first by 20–40% in my own data. Brands that wait 4+ months see roughly parity or slight underperformance.

When you re-book, raise the rate by 10–15%. This sounds like overpaying. It isn’t. You are buying continued exclusivity in their content calendar, you’re shortening their ramp-up time (which they will pass back to you in faster delivery), and you’re signaling that you treat the relationship as a relationship. Creators talk to each other about which brands re-pay and how. You want to be in the right column of that conversation.

The brands that build moats here pay the +15% and then re-book the same creator a third time within 60 days at the same rate, with a third creative angle. By deal #3, the creator has internalized your brand’s positioning at the level a freelance ECD would charge you $5K to learn, and is producing content for $1,500 that converts at a rate you can’t get from any new shortlist. That’s the compound.

Cell 2: Above-target performance, rough relationship → re-book once, with new terms

This is the cell people get wrong in both directions. Some founders re-book anyway because the numbers were good and ignore the friction, then are surprised when the second deal goes even worse. Others drop the creator entirely and lose the highest-ROI relationship they have.

The right move is to re-book once, but tighten the operational contract: shorter revision window, stricter response-time SLA, a clearer kill-fee trigger if anything slides. Have a one-paragraph conversation with the creator about what went sideways the first time. Most creators (the good ones, anyway) will thank you for the feedback and adjust, because they want the re-book too and they know they were rocky.

What the conversation actually sounds like:

“Loved the content, the numbers were strong. The part that didn’t work for us was the response time on the revision round, we were waiting 4 days for the second draft. For the next deal I want to lock in a 48-hour turn on revisions, and if we can’t hit that we invoke the kill fee. Are you good with that?”

That’s it. Specific, non-personal, names the operational issue, asks for explicit agreement. If they push back or get defensive, that’s your signal, drop and don’t re-book. If they say “yeah that’s fair, I had a personal thing the first time, won’t happen again,” re-book under tightened terms.

If the second deal is also rocky, then drop. But give them one structured chance. The performance signal is real and rare; an operational issue is usually fixable. About two-thirds of the time the second deal under tightened terms is the cleanest one in your roster.

Cell 3: Below-target performance, smooth relationship → don’t re-book, keep the door open

This is the cell where founders most often make a false re-book decision. The creator was lovely, the team enjoyed working with them, and the impulse is to give them another shot “because the first one was probably just a bad creative.”

Resist. The math doesn’t support it. Brands that re-book under-performing-but-friendly creators see the second deal outperform the first roughly 25% of the time, which means 75% of the time you’ve now spent twice the budget for a sub-target result, with the opportunity cost of a creator from your shortlist you haven’t tested yet.

The right move: send a warm, specific “thanks, the timing didn’t land but I really enjoyed working with you” message, and put them on a 6-month watch list. Audiences shift. The same creator who didn’t convert in March can absolutely convert in September if their content evolves, their audience composition changes, or your product offering changes. Re-evaluate them with fresh eyes at that point, not by re-running the same deal on a hunch.

The watch list is not a polite fiction. Actually check on those creators every six months. Look at their new content. Look at the comments. Look at what other brands they’ve worked with since. About 10–15% of Cell 3 creators become Cell 1 creators at the 6-month or 12-month re-evaluation if the broader context has changed. That’s a real source of high-ROI deals that the brands rotating their shortlist never see.

Cell 4: Below-target performance, rough relationship → drop, no second deal, no drama

This is the easiest cell and the one founders agonize over the most. The numbers were bad, the working relationship was bad. There is no second deal. Don’t re-book, don’t “give them one more chance,” don’t try to fix it.

The right move is a clean, short, professional message:

“Thanks for the work, we’re going to take a different direction for the next cycle. Wishing you the best.”

Then mark them on your internal list and move on. No public callouts, no detailed feedback (it won’t be received well in this cell), no leaving the door open for a follow-up pitch.

The compounding cost most founders miss: every hour spent agonizing over a Cell 4 creator is an hour not spent finding a Cell 1 creator. The portfolio matters more than any individual relationship.

How often to re-evaluate

The right cadence depends on your deal volume. Three rough heuristics:

Fewer than 5 deals/month: Re-evaluate after every deal. Your roster is small enough that each data point matters disproportionately, and you can run the 2x2 on a single creator in 10 minutes.
5–15 deals/month: Re-evaluate monthly, in a single 60-minute review where you walk through every active creator and place them in the grid. Make the re-book / drop calls in batch.
15+ deals/month: Re-evaluate quarterly per cohort, with monthly performance check-ins that flag obvious Cell 1s for immediate re-booking and obvious Cell 4s for immediate drop. The middle two cells get the deeper quarterly look.

The trap to avoid: indefinitely deferring the decision. Creators who aren’t actively being re-booked or actively being dropped end up in a purgatory of “we might do something with them next quarter”, which slowly erodes the relationship without ever producing another deal. Decide. The cost of a wrong call is one deal; the cost of no call is the whole roster going stale.

What the roster should look like after a year

If you’ve been running this loop for 12 months on a steady deal volume, the math works out to roughly this shape:

5–8 “core” creators in Cell 1, re-booked 3–5 times each over the year. These do the heavy lifting on your CAC. Average rate paid is roughly 1.5× what you paid them on deal #1.
3–5 “tested” creators in Cell 2 or upgraded into Cell 1, re-booked once or twice. These are your second tier.
10–15 “tested out” creators in Cells 3 and 4, no longer being booked, periodically re-evaluated. These are the cost of having a roster at all.
An ongoing trickle of new candidates, typically 2–3 per month, coming in to refresh the top of the funnel using the discovery loop.

The brands that hit this shape stop spending in unpredictable bursts and start spending in a predictable, compounding pattern. The brands that don’t are still running discovery campaigns in month 13 with no creator they’ve worked with more than once.

The math on why re-booking compounds

A worked example with round numbers. Imagine a brand running 5 deals a month at $1,500 each (= $7,500/month) and an average target CAC of $40.

Discovery-only loop (no re-booking):

5 new creators per month, each producing one deal.
Average CAC across the cohort: $52 (some hit, most miss the target).
Cost per acquired customer: $52.
Annual budget: $90K, total customers acquired ~1,730.

Discovery + re-booking loop:

Months 1–3: 5 new creators per month, same as above.
Month 4 onward: 2 core re-books per month at +15% rate ($1,725), 3 new candidates per month.
Re-book deals hit average CAC of $30 (above target by ~25% on average, momentum + brand knowledge).
New-candidate deals continue at $52 average CAC.
Blended cost per acquired customer drops to roughly $38.
Annual budget: ~$96K (slightly more), total customers acquired ~2,530.

The re-booking loop costs ~7% more and acquires ~46% more customers. That gap is the compound. It widens every month you run it. The brands that never re-book never get there, and they don’t know what they’re missing because they don’t have the comparison.

If you remember three things

Re-book your Cell 1s within 30 days at +15%. Momentum is real, and creators who get re-booked fast become your CAC moat by month 3.
Cell 2 gets one structured second chance under tightened terms. Two-thirds of the time the rough deal becomes the cleanest deal.
Cell 3 is the trap. Friendly + flat is the cell brands re-book on hope. Don’t. Watch list, 6 months.

The decision is not “should I keep working with this person.” The decision is “where on the 2x2 are they, and what does each cell tell me to do this week.” Make the call cleanly, document it, and run the loop again next month.

If you want the dashboard that tracks creators on this 2x2 automatically, performance against your target, relationship score, re-book cadence, join the CollabBook beta. We built it because we got tired of watching founders run discovery campaigns forever instead of compounding on the creators they’d already paid to find. Either way: stop treating every deal as a one-off. The roster is the asset.

— Dorcas Faleti, CollabBook