How many rounds of formula sample revisions are realistic?
It’s rare to approve a formula right after the first sample, and when revisions drag on, both the timeline and costs grow. I imagine each team has different standards, so I’d like to hear what others think.
Common reasons revisions take longer
- Usage feel criteria are vague
- What you can do in advance
- Be specific about the reference product and what you "dislike"
- Internal opinions keep changing
- What you can do in advance
- Decide on the decision-maker and evaluation criteria in advance
- Ingredients are changed late in the process due to cost issues
- What you can do in advance
- Share the target cost from the beginning
| Reason | What you can do in advance |
|---|---|
| Usage feel criteria are vague | Be specific about the reference product and what you "dislike" |
| Internal opinions keep changing | Decide on the decision-maker and evaluation criteria in advance |
| Ingredients are changed late in the process due to cost issues | Share the target cost from the beginning |
Poll
Which sample round do you usually finalize on?
Sign in to vote
0
If the formula isn’t completely new and the project involves adjusting an existing base, finalizing on rounds 3–4 seems like a pretty realistic benchmark. The flow is to establish the feel and direction with the first sample, adjust key aspects like viscosity, absorption, and residue with the second, then fine-tune with the third. However, if you need to revisit ingredient changes or compatibility with the fragrance, color, or packaging, you may need one more round.
More important than the number of rounds itself is limiting each revision round to 2–3 items and agreeing on pass criteria in advance. If you change viscosity, fragrance, and stickiness all at once, it’s hard to tell what improved. After the third round, it also seems cleaner for both sides to treat new requirements as a change in scope rather than a revision, and revisit the timeline and cost.
Of the poll options, rounds 3–4 seem like the most reasonable benchmark, though it probably varies quite a bit depending on the formula’s complexity and the developer’s sampling policy. At what point do you start renegotiating additional costs or timelines?