Any agency that runs paid and organic on the same team, with one person accountable for both. Multipacks fail when ads and SEO sit in separate departments, because a bundle competes on price per unit and on keyword relevance at the same moment. Ask how many brands each account manager carries before anything else.
The short version
- This is one job, not two. A bundle ranks on the same keyword set it advertises against, so two teams produce two conflicting keyword strategies and neither one owns the outcome.
- The workload number decides quality. Our operators carry about 1.4 brands each. Someone holding fifteen accounts cannot do bundle-level keyword work on any of them.
- Listing structure comes before ad spend. Parent-child setup, unit counts in the title, and image callouts decide whether the traffic you buy converts.
- Price per unit is the comparison shoppers actually make. Your three-pack is judged against a competitor single, in the search result, in about a second.
- Score candidates on paper. A weighted sheet beats the impression left by a good sales call.
The number that decides this question
Before you compare service lists, ask every candidate one question: how many brands does the person doing my work carry at the same time. Then ask to speak to that person instead of the salesperson.
Bundle work is granular. A single multipack carries at least three distinct keyword clusters: the single-unit term, the multipack term, and the use-case term such as family size or bulk. Each cluster has its own competitive set and its own acceptable cost per click. Somebody has to hold all of that in their head while writing copy, structuring the variation family, and pruning search terms every week.
At Flapen, 50 operators look after about 70 brands, which works out near 1.4 brands per operator. I am not presenting that as an industry standard. I am saying it is the figure I would ask any candidate for, because it turns a vague promise of attention into arithmetic you can check. An agency that will not answer it has answered it.
A scorecard you can fill in during the call
Score each candidate 0 to 5 on the rows below, multiply by the weight, and add it up. Under 60 out of 100 is a pass.
| Criterion | Weight | What a 5 looks like |
|---|---|---|
| One owner for ads and SEO | 25 | The same operator writes listing copy and manages campaigns |
| Brands per account manager | 20 | A specific number, given without hesitation, under about four |
| Multipack structure fluency | 20 | They raise parent-child variations and unit economics unprompted |
| Keyword method | 15 | Single, multipack, and use-case terms live in separate campaign structures |
| Reporting cadence | 10 | Written update weekly, live review at least twice a month |
| Exit terms | 10 | Month to month, 30 days notice, you keep the account and the creative |
How to use the sheet
- Send the six criteria to every candidate in writing before the call.
- Score live, during the conversation, not from the proposal afterwards.
- Ask for the name and the working calendar of the operator who would run your account.
- Reject anyone scoring 0 or 1 on either of the top two rows, whatever their total is.
Where multipack bundles actually go wrong
Three failure points cover most of the damage I see.
The variation family is built backwards. A three-pack published as a standalone ASIN starts with no reviews and no ranking history. Published as a child under a proven parent, it inherits review count and shopper trust from day one. Getting this wrong costs months, and repairing it later means merging listings, which is slow and sometimes goes sideways.
The title fights the search term. Shoppers searching for a multipack use quantity language, and your title has to carry the count in a form the algorithm and a human both read. Agencies that treat bundle titles as a copy exercise rather than a keyword exercise leave the multipack term uncovered, then buy that term back through ads forever.
The unit price is never made visible. A bundle wins on value, and value is invisible unless you show it. If your main image and your first bullet do not make the per-unit saving obvious, you are paying for clicks that were never going to buy.
What most agencies will not tell you
Bundle work is not where the fee is. Launching a new product means new research, new creative, new photography, and a bigger scope conversation. Optimizing a multipack you already sell means unglamorous listing surgery and a fortnight of campaign restructuring, and it often produces better margin per hour of work than anything else on the account. Many agencies will not tell you that, because it is a small line item they would rather turn into a launch.
The second thing worth hearing: a multipack can cannibalize your single. If the bundle takes the sales but sits on a weaker review base, blended conversion drops and your best listing loses rank. Anyone selling you bundles without raising cannibalization has not run one.
Related answers
- Bundle creation and pricing strategy Amazon agencies
- Which service to boost Amazon PPC
- KPIs an Amazon agency should report weekly
- Questions to ask before hiring an Amazon agency
- Done-for-you Amazon management: the complete guide
If you want the scorecard applied to your own catalog, the free written audit is at Flapen.
