How to Set Up Perplexity Source Tracking (Manual or With a Tool)

Perplexity source tracking means asking a fixed set of questions on a regular schedule and recording which web pages are cited in each answer. Over time this shows which sites dominate your market, whether your own pages are being used, and how that changes after you publish or update content. You can do it in a spreadsheet or with a tracking tool.

Step 1: Build your prompt list

Choose 15 to 30 questions that real buyers ask. Mix types:

• Category: “best bookkeeping software for freelancers in Spain”
• Problem: “how to fix damp walls in an old stone house”
• Comparison: “Brand A vs Brand B for small shops”
• Local: “car mechanic in Ghent open Saturday”
• Brand: “is [your brand] good for small teams”

Keep the wording fixed so results are comparable over time.

Step 2: Decide what to record

For each run, capture these fields in a spreadsheet:

1. Date and prompt.
2. Each cited URL and its domain.
3. Whether your domain was cited (yes or no).
4. Whether your brand was named in the text (yes or no).
5. Which competitor brands were named.
6. Any wrong facts about you.

Saving the full answer text is helpful too, so you can check how your description changes.

Step 3: Choose a schedule

Monthly is enough for most small businesses. Weekly makes sense during a launch or after major content changes. Run each prompt at least twice per check, because answers can vary from one run to the next.

Step 4: Turn rows into insights

After two or three rounds, build a few simple views:

• Top cited domains across all prompts. These are the sites that shape answers in your market.
• Your citation rate: the share of prompts where your domain appears.
• Competitor share: how often each rival is named.
• Source types: are cited pages mostly review sites, news, forums, brand sites or directories?

If review sites dominate, work on your presence there. If brand sites dominate, improve your own pages.

When a tool is worth it

Manual tracking becomes slow once you pass about 20 prompts or want to track more than one assistant. A tool helps when you need:

• Automatic, scheduled runs.
• History and trend charts.
• The same prompts across Perplexity, ChatGPT, Gemini, Claude and Copilot.
• Several brands, products or locations.

When comparing tools, check that they record cited URLs, not just brand mentions. Our page on choosing a Perplexity citation tool has a full scorecard.

Mistakes to avoid

• Changing prompt wording every month, which breaks comparisons.
• Tracking only your brand name, which misses category questions where buyers discover new options.
• Reading too much into one run.
• Collecting data and never acting on it.

Tracking is the first step; the next is fixing what it reveals. For a cross-assistant approach, see tools for LLM citation tracking.

A sample tracking row

Here is what one row might look like for a bike shop in Utrecht: date “1 March”, prompt “e-bike repair shop in Utrecht”, cited domains “a local city guide, a review platform, a competitor site”, own domain cited “no”, brand named “no”, competitors named “two shops”, wrong facts “none”. Three months later, after publishing a clear repair services page and fixing its listing, the shop can compare the same row and see whether anything moved. Simple rows like this, kept consistently, are enough to guide real decisions.

How Keboost helps

Keboost automates this whole process. It runs your prompts across Perplexity and other AI assistants, records which brands and sources appear, and shows trends over time. It also creates focused content to help close the gaps it finds. Run a free AI visibility check, or see plans from €29 per month.