The competitor AI pairs you with most probably doesn't pair you back

Parse mapped 153,572 brand pairs across 1.3 million co-mentions in the 30 days ending June 26, 2026. A brand's most frequent partner was also its partner's top match only 23.6% of the time. In 85.2% of top pairings, the partner had more total co-mentions, by a median factor of 2.96. Co-mention frequency measures association, not mutual competition.

The brand AI pairs you with most is usually bigger than you

Take every brand active enough to have a stable competitive set: at least five co-mention partners and at least 50 co-mentions in the window. That is 7,252 brands. For each one, find its single strongest pairing, the brand AI names it alongside most often, then look at how that partner ranks the focal brand in return.

The reciprocity is weak. The top partner names you as its own number one only 23.6% of the time. It places you in its top three only 40.8% of the time, and in its top five only 51.6%. The median brand sits fifth on its closest partner's list. Among the brands whose top pairing is not mutual, 44% land outside their partner's top ten entirely.

The reason is size. In 85.2% of these strongest pairings, the partner is the bigger brand, carrying more total co-mentions across the month, by a median factor of about three. You are tied most tightly to a company for whom you are a minor association. Sort the 7,252 brands by what their top pairing actually is and the shape is stark:

• 74.7% are paired most often with a bigger brand that does not reciprocate. AI keeps naming the smaller brand next to a category leader, but the leader ranks it well down its own list.
• 23.6% have a genuine mutual rivalry, where each brand is the other's number one.
• 1.7% are the reverse case, where the focal brand is the larger partner for a smaller specialist.

Only about a quarter of brands have a mutual top rivalry

The 23.6% that do reciprocate are worth looking at, because they are the cases where AI and the market agree on a real two-way contest. These are the pairs where each brand is the other's most-named companion:

These are familiar category rivalries. When a pairing is mutual, each brand is the other's most frequent co-mention. That is uncommon: three quarters of top pairings are one-sided.

If you want to know when AI changes its answer about your brand, start with a free brand check — it takes a minute.

When the pairing is one-sided

For most brands, the strongest pairing connects them to a larger brand that does not rank them as highly in return. The pattern is clearest in crowded software categories, where a few large brands account for many co-mentions:

Read the Figma row carefully. AI names Notion alongside Figma more than any other brand, yet from Notion's side, Figma is the 30th brand it gets paired with. The link is strong in one direction and faint in the other. The same shape repeats down the table: Loom's top association is Slack, but Slack is ten times larger and files Loom 44th; Wrike's strongest tie is ClickUp, which ranks Wrike eighth.

This is a consequence of concentrated attention. A buyer asking about Notion may see Figma mentioned in passing; a buyer asking about Figma is more likely to see Notion as a primary comparison. The larger brand appears in many smaller brands' answers, while its own list contains many established peers.

How we measured it

Parse maintains a co-mention layer: for each rolling 30-day window, it counts how many times each pair of brands is named together inside the same AI answer, across the surfaces it monitors. We used the latest complete window, May 27 to June 26, 2026, which sits entirely in the current collection era (ChatGPT Search and Google AI Mode, web-search-grounded answers), so there is no model-migration seam inside it. Each pair is stored once in canonical order; we collapsed the handful of duplicate recompute rows by taking the maximum count per pair before building the graph.

From the 153,572 pairs we built a directed view: for every brand, its partners ranked by shared co-mentions. We restricted the analysis to brands with at least five partners and 50 total co-mentions, leaving 7,252 brands. For each, we took the number one partner and checked where the focal brand ranked on that partner's list. "Mutual" means each brand is the other's number one. "Larger partner" compares total co-mention volume across the window. Results are aggregate, with no single customer identifiable. A co-mention shows association, not preference. For preference data, see how AI picks a winner in a head-to-head.

Because the signal is co-mention volume, larger brands naturally accumulate more pairings. Size and one-sidedness are therefore related by construction. If a brand is not a category leader, its strongest AI pairing will often be a larger company rather than a balanced rival. The same size effect appears in Parse's analysis of larger brands in unbranded category answers.

What to do with your real competitive set

Use the co-mention list to see which brands AI associates with yours and whether those relationships are mutual.

First, separate mutual pairings from one-sided ones. A mutual pairing is more likely to represent direct competition. A larger brand that does not rank you highly in return may provide category context but is not necessarily a direct rival.

Second, look past the top partner. The brands ranked two through ten in your set often include companies that are not on your competitor slide at all, and some of them sit in adjacent categories AI has decided are next to yours. That overlap is where surprise comparisons come from, and it is the same boundary-blurring we found when we measured which industries AI treats as neighbors.

Third, track the rankings over time. Growth in a one-sided pairing means AI increasingly associates the smaller brand with a category leader. A new mutual pairing indicates that both brands are appearing together more consistently. Compare these changes with recommendation win rates before treating them as competitive movement.

Compare the co-mention list with the company's formal competitor list. Differences can reveal adjacent-category brands or larger reference brands that buyers see in AI answers.