A recommendation portfolio should not give every content type the same placement quota. Best Of lists, direct comparisons, reviews, buyer guides, alternatives pages, expert roundups, directories, and decision frameworks answer different buyer questions. They also have different editorial limits, publisher supply, maintenance requirements, and levels of repetition before another placement stops adding much value.
The right strategy is weighted. Scale the formats that can credibly cover many distinct buyer contexts. Keep narrower formats selective. And judge every additional placement by the new recommendation territory it opens, not by whether the campaign has hit an arbitrary monthly count.
Placement volume should vary by content type because each format has a different number of legitimate buyer questions it can answer before repetition turns into redundancy.
A broad Best Of strategy may support many category, audience, use-case, geography, and constraint angles. A specific Brand A vs. Brand B comparison has a much narrower job. Equal volume would ignore that difference.
The mistake is treating every placement as interchangeable
Once marketers decide third-party recommendation content matters, the next instinct is often to simplify the program into units: ten placements per month, twenty per quarter, or an even mix across every content type.
That makes budgeting easier. It does not necessarily make the strategy better.
A Best Of list can answer broad category questions, industry-specific questions, audience questions, geography questions, price-sensitive questions, and “best for” use cases. A comparison page usually exists because buyers are actively evaluating two specific competitors. An alternatives page often depends on one established product or incumbent having enough demand that users genuinely seek substitutes.
If all three formats receive the same placement volume, the campaign is optimizing for symmetry instead of demand.
What new buyer decision does the next placement cover?
If a new article cannot answer that question clearly, it may be duplicating recommendation territory the portfolio already owns.
AI citation research already shows that content types play different roles
Wix Studio and Peec's study of 75,000 AI answers and 1,056,727 citations found large differences in citation patterns by prompt intent.
For commercial-intent prompts, listicles represented 40.86% of citations. Category pages followed at 12.42%, discussions at 11.44%, product pages at 7.14%, articles at 6.15%, comparison pages at 4.08%, and alternatives pages at only 0.54%.
Those percentages should not be converted directly into a media allocation model. They describe observed source behavior, not how your brand should divide its budget. What they do establish is that content types are not functionally interchangeable, so placement volume should follow the buyer job each format is best equipped to handle.
Broad formats naturally support more placements
Some formats have a large legitimate angle universe. Best Of content is the obvious example.
A cybersecurity company might reasonably belong in separate articles covering the best providers for healthcare, mid-market businesses, compliance-heavy organizations, companies without internal security teams, or a particular region. Those can represent different buyer problems, competitive sets, and evaluation criteria rather than duplicate articles.
The same is true of use-case recommendations. If a product solves materially different problems for different audiences, multiple placements can establish distinct associations around the brand.
Higher volume becomes defensible when the buyer context changes enough to make the recommendation meaningfully different.
Narrow formats hit diminishing returns faster
Other content types have much tighter boundaries.
A direct Brand A vs. Brand B comparison may justify more than one independent publisher, especially when audiences or evaluation criteria differ. But ten nearly identical comparisons do not create ten times the value. At some point the portfolio stops expanding and starts repeating itself.
Alternatives pages can reach that point even faster. They are valuable when buyers genuinely seek substitutes to an incumbent, not when marketers manufacture an “alternatives to” angle around every minor competitor.
The narrower the buyer question, the faster marginal placement value usually declines. Scale only while each additional source adds credible independence, audience reach, evaluation depth, or a meaningfully different recommendation context.
A practical weighting model for the 10 core content types
The following framework is not a universal benchmark or required placement count. It is a strategic way to think about relative volume inside a diversified recommendation portfolio.
| Content type | Typical relative volume | Why | When another placement adds value | When to stop |
|---|---|---|---|---|
| Best Of List | Higher | Broad angle universe across categories, audiences, use cases, geographies, and constraints. | The new list targets a distinct buyer question or credible publisher audience. | New lists merely repeat the same category and positioning. |
| Comparison Page | Selective to moderate | Useful when buyers actively evaluate two known options. | A new publisher, audience, methodology, or comparison criterion adds independence. | The same pair is being compared repeatedly without new insight. |
| Alternatives Page | Selective | Depends on real switching or substitute demand around an incumbent. | The target competitor has meaningful buyer awareness and the brand is a legitimate substitute. | Pages are being created around weak competitors solely to manufacture volume. |
| Expert Recommendation Roundup | Selective to moderate | Value comes from expert diversity and authority more than repetition. | A distinct expert community or audience brings new recommendation context. | The same voices or generic recommendations are recycled. |
| Buyer Guide | Moderate | One strong guide can cover many related criteria and questions. | A separate market segment has materially different buying criteria. | New guides simply repackage the same selection framework. |
| Use Case Recommendation | Moderate to higher | Distinct use cases can create legitimately different shortlists. | The audience, constraint, workflow, or problem meaningfully changes. | Use cases become artificial variations rather than real buyer needs. |
| Product / Service Review | Selective | Depth, independence, and evidence matter more than raw review count. | A credible publication can add a distinct perspective, test, or audience. | Reviews become templated endorsements with no differentiated evaluation. |
| Educational / Problem-Solution Article | Moderate, but not primarily placement-driven | Supports problem awareness, expertise, and category context more than direct shortlist visibility. | The article contributes new evidence or reaches a relevant external audience. | The program begins mass-producing commodity explainers. |
| Resource / Vendor Directory | Selective to moderate | High-quality directories establish inclusion but can become redundant quickly. | The directory is authoritative for a distinct industry, geography, or audience. | Remaining directories are low-quality or add no new market context. |
| Decision Matrix / Selection Framework | Selective | High-value format, but each matrix should represent a real evaluation model. | Different buyer requirements create a genuinely different framework. | The scoring system becomes cosmetic or repetitive. |
Commercial importance and citation frequency are not the same thing
It is tempting to look at the Wix/Peec citation percentages and conclude that lower-cited formats deserve lower strategic investment. That would be too simplistic.
A direct comparison page accounted for only 4.08% of commercial-intent citations in the study, but a comparison can still be extremely important when your buyer is deciding specifically between you and the category leader. An alternatives page represented just 0.54%, yet “alternatives to [competitor]” may be one of the most valuable commercial questions in your market.
The volume decision should therefore combine frequency with decision value.
How broadly does this content type appear?
Research can show which formats AI systems cite often for a particular intent and which formats are common across the market.
How important is the buyer moment?
A less common format can still deserve attention if it corresponds to a high-value competitive decision close to purchase.
Broad web presence matters more than a content arms race
Ahrefs' cross-platform study of 75,000 brands found branded web mentions correlated roughly 0.66–0.71 with AI visibility across ChatGPT, Google AI Mode, and AI Overviews. The same study found almost no relationship between the number of pages on a brand's own site and AI visibility, with page count around 0.194.
Correlation does not prove causation. But the pattern argues against treating AI visibility as a content-volume competition. Ahrefs' researchers explicitly described the takeaway as more than a content-creation arms race and emphasized being mentioned across a broad scope of sites.
That reinforces the portfolio logic. Five recommendation placements distributed across five meaningful contexts and credible sources may add more strategic range than fifteen repetitive pages built from one template.
The right volume is therefore not just a question of how many assets exist. It is also about how broadly the brand is represented across independent sources and how much new context each source contributes.
Publisher supply should cap volume
Every content type also has a different supply of credible publishing opportunities. Best Of and buyer-guide content may fit many industry publications, while decision matrices, rigorous reviews, and expert roundups require more editorial capability.
Once the campaign exhausts the strong publishers for a format, lowering the quality threshold simply to maintain volume is usually the wrong move.
This is especially important because Google now explicitly includes attempts to manipulate generative AI responses inside its spam-policy framework. Its policies address scaled content abuse, low-value third-party content used to exploit site reputation, and link schemes. Paid or sponsored content can exist legitimately, but it should provide real value to readers rather than exist primarily as a ranking manipulation tactic.
Maintenance burden should also influence the mix
Some recommendation content ages faster than others. Software comparisons, Best Of pages, directories, and reviews can become inaccurate as pricing, features, competitors, and products change.
Higher-maintenance formats should not be scaled beyond the team's ability to keep them credible. Every placement creates a future refresh obligation.
Can the brand and publisher ecosystem keep this format accurate after publication?
If not, the sustainable placement volume for that format is lower than the theoretical opportunity count.
Google's guidance reinforces the danger of equal-volume scaling
Google's current generative Search guidance warns publishers not to create separate content for every possible search variation simply to manipulate rankings or generative responses. Google notes that a high quantity of pages does not make a site higher quality or more relevant.
The principle applies directly to recommendation portfolios. If marketers identify 50 slightly different prompts and respond by commissioning 50 nearly identical placements, the strategy has confused prompt variation with genuine buyer variation.
Scale should follow substantive differences: a different audience, use case, geography, competitor relationship, purchasing constraint, publisher audience, or editorial methodology. If none of those things changes, another placement may simply be duplication.
Build the mix from gaps, not percentages
The best way to determine volume by content type is to map the recommendation market first, then assign formats to the gaps.
Group buyer questions by job
Separate shortlist questions, direct comparisons, alternatives, reviews, buyer-criteria questions, use cases, directories, and decision-framework needs.
Measure the current footprint
Identify which formats already contain the brand, which competitors dominate each area, and which publishers repeatedly appear around the target prompts.
Assign the format that best solves each gap
Do not default every problem to a Best Of placement. A comparison, review, buyer guide, directory, or use-case recommendation may answer the buyer question better.
Set volume from the legitimate angle universe
Estimate how many genuinely distinct questions, publishers, audiences, and methodologies exist for each format. That becomes the upper boundary for useful scale.
Rebalance as marginal value changes
As one content type becomes saturated, shift effort toward uncovered formats and buyer decisions rather than continuing to add redundant placements.
One format should not consume the entire strategy
Listicles receive disproportionate attention because the research around commercial AI citations is compelling. But a recommendation footprint made only of Best Of lists is incomplete.
A buyer may first ask for the best providers, then compare two finalists, look for an alternative to the incumbent, read a review, and finally ask what option fits a particular use case. Each step can pull from different sources and content structures.
This is why the 10 core Mention Engine content types should be treated as a portfolio rather than a checklist with equal quotas:
- Best Of List
- Comparison Page
- Alternatives Page
- Expert Recommendation Roundup
- Buyer Guide
- Use Case Recommendation
- Product / Service Review
- Educational / Problem-Solution Article
- Resource / Vendor Directory
- Decision Matrix / Selection Framework
The right portfolio rarely distributes placements evenly across all ten. It weights them according to the questions the market is actually asking.
Placement volume should rise with the number of legitimate buyer contexts a content type can cover—and fall as repetition, publisher scarcity, or maintenance burden increases.
Do not build ten of everything. Build enough of each format to cover the recommendation gaps it is uniquely qualified to solve.
The portfolio should get more selective as it matures
Early in a recommendation strategy, obvious gaps may exist across Best Of coverage, comparisons, reviews, and use cases, so several formats can expand at once.
As the footprint matures, the strategy should become more selective. One high-authority comparison or specialized use-case recommendation may add more value than five generic placements. That is the natural evolution from volume to portfolio management.
Mention Engine is designed around this kind of content-type planning. Brands can build third-party recommendation coverage across multiple formats instead of treating every publisher opportunity like a generic sponsored post. The important question is not how many placements fit into a package. It is which combination creates the most complete and credible recommendation footprint for the market being targeted.
When placement volume varies by content type, the portfolio starts to resemble actual buyer behavior rather than an arbitrary production calendar. That is the goal.
Weight your recommendation portfolio around real buyer decisions
Mention Engine helps brands build Best Of, comparison, alternatives, review, buyer-guide, use-case, expert-roundup, directory, educational, and decision-framework content across relevant third-party publishers. Start with the gaps, use the format that best answers each buyer question, and scale each content type only while the next placement adds meaningful coverage.
Explore Mention Engine