In this article:
- What problem did LiveAgent have?
- How does the tracking work?
- What did it show us?
- What did we change in our process?
- What is next for the language sites?
This is a first-party account of how LiveAgent uses AmICited for its own tracking, not an independent review.
What problem did LiveAgent have?
Buyers now ask ChatGPT or Google’s AI Overview which help desk to choose before they open a search result. Rankings and traffic reports do not show whether LiveAgent is part of those answers. If an answer names other help desks and not LiveAgent, the shortlist is set before the website gets a visit, and nothing in the usual reports would reveal it.
How does the tracking work?
The setup is a prompt list, the AI platforms that answer it, a record of every citation, and a weekly re-check.
- A prompt list. Questions a buyer might ask: buying questions, brand questions and comparison questions. Examples are below.
- AI platforms. AmICited runs each prompt on the main AI platforms on a schedule, so every answer is stored and can be compared over time.
- A record of every citation. For each answer AmICited keeps which brands are named, which pages and sources are cited, and in what position.
- A weekly re-check. The same prompts run again, so a change in the answers can be tied to a change on the site.
| Type | Example prompt |
|---|---|
| Buying | help desk software with live chat and phone support |
| Buying | Which omnichannel help desk handles email, chat, phone and social media in one place? |
| Brand | How much does LiveAgent cost? |
| Comparison | LiveAgent vs Zendesk: which is better value? |
What did it show us?
AmICited made visible what LiveAgent could not see before.
- Which questions and platforms name LiveAgent. Google AI Overview and ChatGPT are where LiveAgent shows up most. Broad product questions, such as help desk software with live chat and phone support, are where it is cited most reliably.
- Where it is missing. Many questions do not mention LiveAgent at all. Each one is a gap, and the list of gaps is specific enough to act on.
- Which sources the AI platforms like to cite. Competitors’ own sites and YouTube dominate the cited sources. LiveAgent’s own pages and videos are the part of that picture the team controls.
What did we change in our process?
- The gap list is a content to-do list. A question without LiveAgent in the answer points to a page to write, fix or strengthen.
- Fix, then re-check. After a page changes, the same question runs again on the next weekly pass, which closes the loop that traffic reports leave open.
- Share of voice instead of raw citation counts. A count rises whenever the prompt list grows. Share of voice, the portion of brand mentions that are LiveAgent’s, stays comparable over time.
Join our newsletter
The latest tips and deals delivered straight to your inbox.
What is next for the language sites?
LiveAgent runs separate sites in many languages, and they are the unfinished part of this work. English prompts say nothing about how AI platforms answer in German, Polish or Japanese.
- Local prompt sets. Each language site needs its own prompts, written in its own language for its own country, because buyers phrase questions differently and the competitors differ by market.
- The same loop. Once a language has data, the same steps apply: find the gaps, fix the page, re-check.
Once AI sends buyers your way, they still need fast answers. LiveAgent’s help desk puts every channel in one inbox.
