In traditional search, losing to a competitor still leaves you on the page. In an answer engine, a competitor who is named while you are not has taken the entire result. There is no second position and no impression to count.
That makes competitive measurement more consequential in AEO than in SEO, and also easier to get wrong, because every competitive number HubSpot reports depends on a competitor list you configured yourself. This article covers how that works, how share of voice is calculated, what to compare, and what to do about the gaps. For the tool overall, see our guide to HubSpot AEO.
Quick Takeaway
- Share of voice is a ratio between your mentions and the mentions of the competitors you listed. It moves when you edit the list, and HubSpot notes competitor changes only affect future data runs. Get the list right at setup, and read the number as relative position rather than market share.
How HubSpot Tracks Competitors
HubSpot AEO compares your brand against named competitors you configure during setup, reporting how often each appears in answers to your tracked prompts.
Competitors are set up with a name, domain and brand name variations, the same fields you complete for your own brand. HubSpot’s documentation states you can add, edit or delete competitors later from AEO settings, with one significant limitation: changes to competitors only affect future data runs, and after updating you should set your dashboard date range filter to start after the date of the change.
That limitation deserves emphasis. Adding a competitor in month four does not backfill months one to three, so your trend line breaks at the point you edited the list.
Two setup errors worth avoiding:
- Aspirational competitors. A market leader your buyers never shortlist will absorb mentions and depress your share of voice while teaching you nothing.
- Missing brand variations for competitors. An unrecognized variation means their mentions go uncounted, which inflates your share of voice. That is the more dangerous error, because it flatters you.
Build the list with sales input, from the names that actually appear in lost deals.
What AI Share of Voice Actually Measures
HubSpot defines share of voice as the proportion of total brand mentions that belong to your brand across your tracked prompts, calculated by dividing your brand’s mentions by the total number of brand mentions from all competitors for those prompts.
HubSpot adds that a higher share of voice means your brand accounts for a larger portion of total mentions across responses. The documentation also lists a separate Competitor Visibility report, which compares your brand identity against competitors.
Three things follow from that definition:
- It is bounded by your list. Adding a strong competitor mechanically reduces your share without anything changing in the market.
- It is not market share. It is a share of mentions within a set of prompts you chose, among a set of companies you chose.
- It can rise for bad reasons. If competitors drop out of answers, your share rises while total category presence falls.
Brand Visibility and Share of Voice Answer Different Questions
| Measure | What it answers | When it misleads |
| Brand visibility | How often are we mentioned across our tracked prompts? | Reads as a market position when it is a function of your prompt set |
| Share of voice | Of all brand mentions in those answers, how many are ours? | Moves when you edit the competitor list, not just when performance changes |
| Competitor visibility | How does each competitor’s visibility compare with ours over time? | Compares only the competitors you configured |
Table 1: Three related but distinct measures.
Read them together. High visibility with low share of voice means you are present but crowded out. Low visibility with high share of voice usually means a sparse category where few brands are named at all, which is an opportunity. Blog 2 covers the visibility measure in detail: HubSpot AI visibility.
Comparing Across ChatGPT, Gemini and Perplexity
HubSpot notes each engine may generate different responses, cite different sources or return different levels of detail for the same prompt, and provides filters to compare by engine.
Competitive differences between engines are diagnostic. If a competitor consistently outperforms you on one engine only, the likely explanation is source coverage: that engine draws on material where they are present and you are not. Filter to the engine, read the actual responses, then look at that prompt’s citations to identify the source.
Prompt-Level Competitive Comparison Is Where the Detail Is
The aggregate view tells you whether you have a problem. The prompt view tells you where.
Opening an individual prompt gives you, among other reports, competitor visibility over time and prompt visibility rate versus competitors, which HubSpot describes as showing how much of the conversation your brand owns versus competitors.
A practical triage:
| Pattern at prompt level | What it means | First action |
| You absent, one competitor named | They own that question | Read the citations for that prompt |
| You absent, several named | The category is crowded and you are outside it | Look for a format gap rather than a content gap |
| You named last or briefly | Present but not the recommendation | Check how you are described, and sentiment |
| You and competitors both absent | Nobody owns it yet | The cheapest opportunity available |
Table 2: Reading prompt-level competitive gaps.
The last row is the one most teams overlook. A commercially relevant prompt where no vendor is consistently named is an open position, and it is far easier to claim than to displace an incumbent.
Analyzing Competitor Citations
The citation reports work competitively too. HubSpot’s owned domain citation rate versus competitors report compares how often engines cite content from your website versus competitor domains, and its documentation notes this can reveal gaps where competitors are cited more often, which may show opportunities to improve the clarity, coverage or relevance of your content.
Three comparisons worth running:
- Which domains cite them and not you. Your outreach list, in priority order.
- Which content types support their mentions. If comparison articles and user reviews dominate and you produce neither, the gap is format rather than volume.
- Which channels they are strong in. A competitor visible mainly through user-generated content has community presence, which is slow to build and slow to lose.
Full detail is in citation analysis.
Order matters here, because the cheapest wins are not the most obvious ones.
Turning Competitive Findings Into Action
- Claim unowned prompts first. Commercially relevant questions where no vendor is consistently named.
- Fix description problems. Where you are named but described inaccurately or weakly, check sentiment at prompt level and the sources behind it.
- Close format gaps. Produce the content types citation data shows being used in your category.
- Target specific sources. Pursue presence on the domains citing competitors, which is PR work rather than publishing.
- Only then contest the hardest prompts. Displacing an entrenched competitor on a heavily cited question is the slowest work available.
HubSpot’s Recommendations tab automates part of this, noting that competitor presence is one of the patterns its analysis includes. See AEO recommendations.
There is also a positioning question underneath all of this. If answer engines describe you and three competitors in near-identical terms, no amount of citation work fixes that; the problem is that your positioning gives a model nothing distinct to report.
Tracking Competitive Visibility Over Time
| Cadence | What to review | What to watch for |
| Monthly | Share of voice trend and competitor visibility | Movement large enough to sit outside normal variance |
| Monthly | Prompt-level wins and losses since last review | Specific questions changing hands |
| Quarterly | Competitor citation sources | New third-party sources they have entered |
| Quarterly | Whether the competitor list still reflects lost deals | Emerging competitors you have not configured |
| On list change | Reset the date filter | Trend breaks, since changes apply only to future runs |
Table 3: A competitive review cadence.
Set the window in months. HubSpot advises reviewing performance across multiple days or weeks before evaluating trends, and competitive movement is noisier still because it depends on several brands’ fortunes at once.
Who Needs This Data
| Leader | The question they ask | What to show them |
| CEO | Are we in the conversation when buyers choose? | Evaluation and Decision prompt performance, with named competitors |
| CMO | Where are we losing, and why? | Prompt-level gaps plus the citation evidence behind them |
| Sales leader | Does this match what we see in deals? | Actual answer text for comparison prompts |
| PR lead | Which sources do competitors have that we do not? | Competitor citation domains |
| Brand lead | Are we described distinctly? | Sentiment and description wording at prompt level |
Table 4: The internal buying-committee lens on competitive AI data.
The sales leader row is a useful sanity check. If the tool reports competitors your sales team never meets, your competitor list is wrong rather than your market.
THP Studio Perspective
- The finding we would expect to be most common, and most useful, is not that a competitor is beating you. It is that answer engines describe you and your closest rivals in almost identical language. When that happens, competitive AEO work has a ceiling: you can earn more citations, but a model cannot report a difference that does not exist in the sources. In our view the tool is then telling you something about your positioning rather than your content, and the fix belongs with the leadership team before it belongs with marketing
Frequently Asked Questions
Key Takeaways
- Build the competitor list with sales input at setup, because later changes apply only to future data runs and break the trend.
- Read share of voice as a bounded ratio, not market share; adding a competitor lowers it without anything changing.
- Prioritize prompts where no vendor is consistently named, since claiming an open position beats displacing an incumbent.
- Use competitor citation sources as an outreach list rather than as evidence to publish more.
- Treat near-identical descriptions of you and your rivals as a positioning signal for leadership, not a content task.
Key Takeaways
- THP’s Brand Consulting & Recall team works on B2B positioning, messaging and differentiation. If answer engines describe you in the same terms as your competitors, that is where the work starts, before any amount of citation building.
Author
