The problem with brainstormed card lists

Most card sorts begin the same way. A researcher gathers the product team in a room, they brainstorm a list of items, someone cleans up the wording, and the list becomes the card set. The study runs, results come back, and everyone feels good about the "user-validated" information architecture.

The problem is that this process tests the team's assumptions about what items exist and what they should be called. If your brainstormed list is missing items users expect, no amount of sorting will reveal the gap. If your card labels use internal vocabulary, participants will sort them as best they can — but the resulting structure will be built on language users would not have chosen themselves.

Brainstorming is not wrong. But it is insufficient as the sole input. The fix is straightforward: before you write a single card, document how your competitors organize their navigation. This gives you an external, user-facing baseline that exposes your blind spots before the study runs.

What a competitive navigation audit gives you

When you document the information architecture of 3-5 competitors, you get four things that brainstorming alone cannot provide.

Vocabulary grounded in user expectations. Your competitors' navigation labels are user-facing by definition. They have been tested, iterated on, and shaped by the same market forces you operate in. "Brainstorming" might produce "Purchase History." Competitor research might show you that four out of five competitors call it "Orders." That distinction matters — your card sort results are only as good as the language on your cards.

Coverage you did not think of. Competitors serve the same user needs you do, but they may have identified needs you have not considered. A competitive audit surfaces features, content areas, and navigation patterns that are absent from your product but present in the market. These become candidate cards your brainstorm would have missed.

Industry conventions. Some navigation patterns are so widespread that users expect them. If every competitor in your space puts "Pricing" in the main navigation, burying it under "Company" is fighting convention. A competitive audit separates true conventions (patterns that appear across most competitors) from individual choices (patterns unique to one or two).

Points of divergence. Where competitors disagree about how to organize something, you have found a genuine information architecture problem — one that your card sort is uniquely positioned to resolve. These divergence points are the highest-value items to include in your study because they represent real ambiguity in user mental models.

Step by step: from competitor audit to card sort

1. Select 3-5 competitors

Choose direct competitors who serve a similar audience with a similar product. Include one market leader and at least one smaller player — smaller companies sometimes make bolder IA choices that reveal interesting alternative structures. If your product spans multiple categories, include competitors from each.

2. Document each competitor's navigation

For each competitor, record:

Primary navigation (top-level categories visible on every page)
Secondary navigation (subcategories within each primary section)
Footer navigation (often contains items that did not fit elsewhere — these are architecturally interesting)
Utility navigation (account, settings, help — the functional items)

Capture exact labels. Do not paraphrase "My Learning Dashboard" as "Dashboard" — the specificity of their language choice is data.

3. Build a comparison matrix

Create a spreadsheet or table with competitors as columns and navigation items as rows. Align equivalent items across competitors. This matrix will immediately show you:

Universal items: present in every competitor's navigation. These are table stakes — your card sort should include them.
Common items: present in most competitors. Strong candidates for your card set.
Unique items: present in only one competitor. Include selectively — they may represent innovation or mistakes.
Naming variation: where competitors use different labels for the same concept. These variations become alternate card labels to test.

4. Identify patterns and divergence

Look at how competitors group items, not just which items they include. If three competitors place "Returns" under "Orders" and two place it under "Help," you have found a divergence worth testing. Note every structural disagreement — these become the most valuable parts of your card sort.

5. Generate your card set

Pull card labels from the comparison matrix using these rules:

Use the most common user-facing label for each item, not your internal term.
Include items from all frequency tiers — universal, common, and a selection of unique items.
Write neutral labels. If competitors split between "Orders" and "Purchase History," do not pick one — use the more generic term or test both across participant groups.
Keep the count between 30 and 60 cards. If your competitive audit surfaced more items, prioritize by relevance to your research question.

CardSort lets you document competitor navigation structures in the Competitive Analysis phase and generate a card sort directly from those labels in one click. The labels carry over exactly, so you are not manually re-entering data or accidentally rewording items between phases.

What the results tell you when cards come from competitive research

Card sort results are always useful, but they are more interpretable when your cards come from competitive analysis because you have a frame of reference for the output.

High agreement with a competitor's structure. If your participants independently recreate Competitor A's navigation groupings, that is a strong signal. It means Competitor A's IA aligns with user mental models — and you should seriously consider a similar structure. This is not copying; it is evidence that the competitor got something right.

Participants create entirely new categories. When users sort competitor-sourced items into groups that no competitor uses, you have found a differentiation opportunity. These novel categories represent user needs that the market is not serving well structurally. Building your IA around them gives you a genuine advantage.

Items that nobody agrees on. Cards that split evenly across multiple categories in your sort are the same items that competitors cannot agree on either. This convergence of ambiguity tells you the item is genuinely hard to place — consider cross-linking, search prominence, or a dedicated section.

Naming patterns. In open and hybrid sorts, the category names participants create can be compared directly to competitor navigation labels. Where participant language matches competitor labels, the market vocabulary is settled. Where it diverges, you have an opportunity to use language that is closer to how users actually think.

Case example: an e-commerce home goods company

Consider a mid-size e-commerce company selling home goods that is redesigning its product navigation. The team's initial brainstorm produces 35 cards based on their product catalog categories.

Before running the sort, they audit five competitors: two large retailers, two specialty home goods stores, and one direct-to-consumer brand. The audit reveals:

All five competitors separate "Furniture" from "Decor," which the team had combined into "Home Furnishings."
Three competitors have a dedicated "Outdoor" top-level category. The team had distributed outdoor items across other categories.
Competitors use "Bath" consistently, while the team's brainstorm used "Bathroom Essentials."
One competitor groups items by room ("Living Room," "Bedroom," "Kitchen"). The others group by product type ("Furniture," "Lighting," "Textiles"). This is a genuine structural divergence.

The revised card set includes 48 items sourced from the competitive matrix. The open card sort with 40 participants reveals:

Users strongly prefer product-type grouping over room-based grouping (only 4 participants created room-based categories).
"Outdoor" emerges as an independent category in 32 of 40 sorts — confirming the competitor pattern and overriding the team's original plan.
Participants split on whether "Candles" belongs with "Decor" or "Fragrance." No competitor had a "Fragrance" category, but 15 participants created one — a potential differentiation opportunity.

Without the competitive audit, the team would have tested their internal taxonomy ("Home Furnishings," "Bathroom Essentials") and received results shaped by those labels. The competitive-first approach gave them cards grounded in market reality and results they could interpret against industry patterns.

The competitive audit does not replace user research — it makes it better

To be clear: a competitive audit is not a substitute for card sorting. Competitors' navigation reflects their design decisions and their user research (or lack of it). Those decisions may be wrong. The audit is an input to your card sort, not a replacement for it.

What the audit does is raise the floor. Instead of starting with a brainstormed list and hoping you covered the right items in the right language, you start with market-validated items and externally-sourced labels. Your card sort then adds the layer competitors cannot give you: how your specific users think about these items.

The sequence is: competitive analysis to source cards, card sort to understand user mental models, then tree testing to validate the resulting structure. Each phase builds on the last, and the data compounds.

Why your card sort needs competitive analysis first