Guides
The practical guide to AI visibility and source footprints
Define the questions that matter, map mentions and cited sources, identify meaningful source gaps and connect them with credible digital PR work without promising an AI citation.
AI visibility is often reduced to a question such as “Does ChatGPT mention us?” That can be a useful observation, but it is not a complete strategy.
A brand's representation depends on the question, system, location, time and sources available. Improving that representation requires understanding the wider information environment, not simply adding the phrase “best provider” to another page on the company website.
This guide explains how to audit AI visibility, map the sources behind it and turn credible gaps into digital PR work.
Define the questions that matter
Track customer decisions, not only the brand name.
Record the exact wording and intended market. A small change to a prompt can alter the meaning and the answer.
- Category questions
- Best-provider and shortlist questions
- Alternative and comparison questions
- Problem and solution questions
- Industry or location questions
- Trust, safety and compliance questions
- Questions connected to a founder or expert subject
Understand the possible observations
A mention and a citation are not the same.
Do not turn one answer into a permanent ranking. AI outputs can vary, and the reporting should retain the date and system for every observation.
Brand mention
Does the answer name the brand, product or spokesperson?
Description
How is the brand characterised, and is the information accurate?
Prominence
Is the brand central, one item in a list or mentioned only in supporting context?
Competitors
Which alternatives appear, and what evidence or attributes are associated with them?
Sources
Which pages are shown as citations or supporting material?
Variation
Does the answer change across repeated observations, systems or markets?
Map the source footprint
See where information about the brand exists outside its own website.
Classify each source by subject, page type, market, freshness, independence and whether it mentions the brand accurately.
- Company and product pages
- Original research and primary evidence
- Earned media coverage
- Expert commentary
- Listicles and comparison articles
- Independent reviews
- Business and industry directories
- Local citations
- Podcast and conference profiles
- Partner, association and accreditation pages
- Press releases and announcement pages
- Linked and unlinked mentions
Identify meaningful gaps
Prioritise the missing sources that influence the category.
A competitor appearing on a page does not automatically make that page a target. Ask:
- Is the source visible for a relevant question?
- Does the audience matter to the brand?
- Is the page current and credible?
- Does the brand genuinely meet the inclusion criteria?
- Is there an editorial or submission route?
- What evidence would be needed?
- Would inclusion help customers understand the brand?
Quality over count
Reject irrelevant, low-quality or inaccessible targets. The goal is a stronger source environment, not a larger count.
Choose the right route for each gap
Different source types require different work.
Earned editorial coverage
Develop a useful story, evidence set or expert contribution for journalists who cover the subject.
Journalist requests
Respond directly when an approved expert or piece of evidence matches the live need.
Listicles and comparisons
Understand the page's audience and criteria, then present a concise case for consideration without demanding inclusion.
Reviews
Provide honest product information, appropriate access and clear disclosure while respecting the reviewer's independence.
Citations and directories
Submit or correct accurate business information through the source's expected process.
Primary evidence
Publish clear, supportable first-party information when the source gap exists because the brand has not explained or evidenced the subject properly.
Build material worth referencing
Make the information useful before making it visible.
Avoid manufacturing low-value pages solely to target a prompt. Clear structure helps systems understand content, but it does not make unsupported material authoritative.
- Original research with a transparent methodology
- Practical analysis of a market change
- Expert explanation based on direct experience
- A useful dataset or benchmark
- A well-documented product or policy change
- A customer example cleared for publication
- Accurate definitions and primary company facts
Connect digital PR with AEO
Build independent evidence around the questions.
Digital PR supports answer engine optimisation by creating relevant third-party references and making approved expertise available to publishers and audiences.
The process is indirect. Lula can identify an influential source gap, run an appropriate campaign and verify a placement. The AI system still decides whether to retrieve, mention or cite that source.
- Question and baseline observation
- Source gap identified
- Campaign or submission completed
- Placement verified
- Later answer observations
- Attribution note and alternative explanations
Measure responsibly
Use a dashboard that preserves context.
Look for repeated patterns rather than reacting to a single answer. A strong source footprint is valuable to customers even when one system does not cite it on a particular day.
- Exact question
- System and market
- Observation date
- Brand mention and description
- Competitors mentioned
- Cited source URLs
- Source type and freshness
- Whether the source mentions the brand
- Relevant placements earned
- Change since the previous observation
- Search impressions, clicks and referral visits where available
- Attribution limitations
Common mistakes
Avoid strategies that create noise instead of authority.
Tracking only the brand name
This misses the category and problem questions that shape discovery.
Treating every mention as positive
An inaccurate or poorly contextualised mention may need correction.
Chasing every cited domain
Some sources are irrelevant, inaccessible or inappropriate for the brand.
Claiming one placement caused the change
Timing is evidence of sequence, not automatic proof of causation.
Publishing pages only for machines
Useful, verifiable information for people is the stronger foundation.
Promising a guaranteed citation
No external product controls the independent output of an AI system.
Frequently asked questions
- How often should AI visibility be checked?
- Use a frequency that can show patterns without treating normal variation as an incident. The right cadence depends on the question, market and plan.
- Which questions should I track first?
- Start with high-value category, comparison and problem questions connected to real customer decisions.
- Is a citation always better than a mention?
- They answer different questions. A mention shows brand representation. A citation shows a source presented by the system. Both need context.
- Can I ask a publisher to add my brand to a cited page?
- Only where the brand meets the page's purpose and criteria. The request should respect editorial independence and provide a useful, evidence-backed reason for consideration.
- Are directories useful for AI visibility?
- Relevant, accurate directories can contribute to the wider source environment, particularly for business identity, category and location information. More directories are not automatically better.
- Can Lula guarantee a change in AI answers?
- No. Lula can systematise monitoring, gap analysis and source-building work, but AI systems control their own outputs.
Build the source footprint behind the answer
Meet Lula shows where the brand is represented, which sources shape the category and which credible gaps can become digital PR campaigns.