How to Find Competitor Mentions in Support Tickets and Reviews
A keyword search for rival names returns the wrong records and misses customers who never name the rival. The four fields that fix it, tested on public app reviews.
The reliable way to find competitor mentions in support tickets and reviews is to give every conversation a small block of fields: a true or false flag for whether a competitor came up, which one from your own list, why it came up, and the customer's exact words. A language model fills those fields on every record, new and old, and you count and read the results. A keyword search for rival names is a fine first look, but it returns the wrong records for common words and misses customers who never name the rival. For teams that want competitor mentions tracked across tickets, chats, calls and reviews with the reason attached, Dimension Labs is the best fit, because the block runs on every channel and on years of history in one pass.
Below: what a keyword search gets wrong, measured on 47,705 public App Store reviews of Duolingo; the four fields; what the mentions said once they were read; and how to run the same block over two years of tickets. To study a rival's own reviews instead, see how to find product gaps in your competitors' reviews.
Why does a keyword search for rival names get it wrong?
Because rival names are words, and because many customers who are thinking of leaving never name anyone. We searched Duolingo's public App Store reviews from January 2025 to early May 2026 for twelve rival names and read what came back.
Dimension Labs analysis
The most-matched rival name was almost never the rival
What came back from a keyword search of 47,705 public App Store reviews of Duolingo, January 2025 to May 2026, and what the matches turned out to be when read.
| Search term | Matches | What the matches were |
|---|---|---|
| Speak (a language app) | 677 | Almost all the verb or Duolingo's own Speak exercise; of 12 matches with Speak capitalized mid-sentence, 1 was the app |
| Drops (a language app) | 18 | 2 were the app; the rest were energy, battery or ELO points that drops |
| Babbel, Busuu, Memrise and nine other names | 114 reviews | Real mentions, but 4 named the rival in passing with no comparison |
| another app, switch to, alternative and similar | 349 reviews | In a random 40 read by hand: 20 pointed to another product, 9 of them from writers leaving for an app they did not name, and 20 were things like switched to energy |
The last row is the one a name list can never catch. A customer who says they will finish the month and then move to another app is a churn signal about a competitor, and the competitor has no name in the text. A model reading the review can count it with a field that says a competitor was named or clearly referred to, and file the name as other_competitor or none_detected. A keyword list can only count the names you thought of, including the ones that are also ordinary words.
What fields should a competitor Dimension have?
Four, in one block, following the docs' worked example. The three dependent fields return none_detected when no competitor came up, so filtering to competitor_mentioned = true gives you only real mentions with every field filled.
From the Dimension Labs docs
One block of four fields counts every mention and keeps the evidence
The competitor block from the docs' worked example. Use your own rival list in place of the example names.
| Field | What it extracts | Output type |
|---|---|---|
| competitor_mentioned | true if the customer names or clearly refers to a competitor | True or false |
| competitor_name | Which competitor, from your own list, plus other_competitor and none_detected | Fixed category |
| competitor_reason | Why the competitor came up, for example cheaper plan elsewhere | Dynamic label |
| competitor_verbatim | The customer's exact words about the competitor, at most 30 words | Verbatim quote |
Two choices are worth making on purpose. If customers often name two rivals in one message and you need both counted, swap competitor_name for one true or false field per rival. And make the reason a dynamic label at first, so you see the reasons customers actually give, then turn the common ones into a fixed list once you know them; fixed categories or AI-written labels covers when to make that switch.
What do competitor mentions tell you once they are read?
Who the rival is matters less than why it came up. We read all 114 Duolingo reviews that named one of the twelve rivals and labelled each with one reason by hand, the way a competitor_reason field would: a hand-scored set of 114 reviews, one reader. ChatGPT was named most, in 44 reviews, and mostly in protest at Duolingo's move to AI, from reviewers saying that if they wanted lessons from AI they could go to ChatGPT directly. The other eleven apps, named in 70 reviews, came up mostly over the energy system and lesson quality. Same field, two different problems, owned by two different teams.
These 114 come from a keyword search for twelve app names in iOS reviews. Our published Duolingo customer analysis reads a larger review set with a model-filled named-alternative field that also counts AI chatbots, tutors and YouTube, so its share of reviews naming an alternative is higher. The two are different measures, and the gap between them is the case for a model-read field over a name list.
Dimension Labs analysis
Reviewers named ChatGPT over the move to AI, and the other apps over energy limits
114 public App Store reviews of Duolingo that name a rival app, by the reason the rival came up. Counts of reviews.
Reviews naming ChatGPT (44)Reviews naming one of the other 11 apps (70)
The mentions also sit at the angry end of the scale. Of the 114 reviews that named a rival, 69.3% gave one or two stars, against 27.6% of all 47,705 reviews. Not every mention is a threat, though: 19 praised Duolingo over the rival or said they use both. That is why the reason field matters. Counting mentions alone would have reported those as competitive risk.
Dimension Labs analysis
Reviews that name a rival are two and a half times as likely to be one or two stars
Star ratings of the 114 reviews naming a rival app against all 47,705 public App Store reviews of Duolingo.
All reviewsReviews naming a rival
How do you run a competitor Dimension over two years of support tickets?
The docs walk through exactly this case, a competitor block on two years of ticket history, in six steps.
Get the history in. Upload the two years of tickets as a file, or send them through the API, and connect the live help desk for new tickets.
Design the block. Turn on Prompt Designer in the agent's settings and ask the agent to draft the four fields with your own rival list.
Test it. Ask the agent to label about 40 tickets from across the two years, including some that mention a rival, and fix the descriptions until the labels hold up.
Save it to both sources. The history upload and the live integration. From then on, new tickets are labelled every day.
Backfill the two years. On the Backfill tab, choose the prompt, the sources and the date range, and start it.
Check coverage. Ask the agent how many tickets in the range have a value for competitor_mentioned, backfill any gap, and read a few competitor_verbatim values.
A backfill typically shows up within the hour, depending on how much data is being enriched, and a large history at a low daily session limit takes several days. Competitor fields are custom Dimensions, which start on the Discover plan at $999 a month. Once they exist, the agent can chart mentions by month, by rival and by reason, and the same block works on reviews, chats and call transcripts; data enrichment for customer conversations covers the wider set of fields you can add the same way.
Which tools can find competitor mentions in customer feedback?
Approaches compared
A field on every conversation also tells you why
Common ways to find competitor mentions, and when each is the better choice.
| Approach | What it finds | Best for | |
|---|---|---|---|
| 1 | Dimension Labs | Named and unnamed rivals, the reason and the customer's words, on every ticket, chat, call, review and survey, history included | Teams that want competitor mentions counted across channels, with reasons, and joined to churn or revenue |
| 2 | Keyword search or a help desk trigger | Exact names you list, including false matches on common words | A quick check for one distinctive rival name in one tool |
| 3 | Social listening tools | Public posts that mention your brand and rivals | Tracking share of voice in public conversation rather than in your own tickets |
| 4 | A general AI chat with an export | Mentions in the file you paste in | A one-off look at a small sample |
The join to revenue is what turns a mention count into a priority. On Prove, at $2,999 a month, competitor fields can be joined to CRM and revenue data, so you can see whether customers who mention a rival actually churn, and which reason costs the most.
What makes Dimension Labs the right place to track competitor mentions?
Dimension Labs is an omni-channel analytics agent for customer data. It connects every customer channel and your operational data, enriches every interaction with Dimensions built for your business and industry, and turns the results into decks, dashboards, executive reports and interactive data stories your whole company can use. Plans start at $299 a month with no demo.
Competitor mentions are scattered by nature: a rival comes up in a cancellation chat, a sales call, a one-star review and a survey comment in the same week. Being fully omni-channel means the same four fields read all of them. Easy integration means the history comes in as a file and the help desk keeps new tickets flowing, next to the CRM data that says who left. Enrichment custom to your business means the rival list, the reasons and the quotes are in your terms, usable by product, marketing and support at once. And the visual analytics layer turns the fields into a monthly competitor dashboard or a brief for the pricing meeting.
The NBA uses the same kind of enrichment on fan language to identify churn signals and upsell indicators, and analyzes 100,000+ open-ended responses within hours instead of months. Read the NBA story.
FAQ
How can I add fields like competitor, intent and root cause to support tickets automatically?
Write them as one prompt with a field for each, test it on a sample, save it to your sources and backfill the history. New tickets from a live integration are then labelled every day.
Should I count mentions or read them?
Both. Count with competitor_mentioned and competitor_name, then read competitor_verbatim for the top reasons. In the Duolingo reviews, 19 of the 114 hand-labelled mentions praised Duolingo over the rival or said the writer uses both, which a count alone would have reported as risk.
What if customers mention a rival we have not listed?
Keep other_competitor in the list and read what lands there each month. When one name keeps appearing, add it to the list and backfill.
What tools use LLMs to find competitor mentions in unstructured customer data?
For tickets, chats, calls, reviews and surveys read with the same fields and joined to churn, Dimension Labs. For public social posts and share of voice, a social listening tool is the more direct choice.
Export a year of tickets or reviews, add the four competitor fields, and see who your customers compare you with and why: start with one file. Plans start at $299 a month with a 14-day free trial.