Somewhere right now, a potential customer is asking an AI platform who they should trust.
They are describing their problem, their budget, their location, and the reasons they are skeptical. They are asking for comparisons. They are asking which company actually understands their industry.
Your sales team is not in that conversation.
Your carefully rehearsed elevator pitch is not in that conversation.
And your homepage screaming “innovative solutions” for the seventeenth time probably is not helping.
The information available about your business is doing the talking.
AI search is turning your public footprint into a sales conversation you do not get to personally control.
If you are a CMO, that should get your attention faster than another software vendor promising to “revolutionize your ecosystem.”
Because the real question is bigger than whether your website ranks.
When a buyer asks for help making a decision, does your brand belong in the answer?
And if it shows up, is the answer even accurate?
1. Understand the machine before you buy the bullshit
AI search is an umbrella term. Google’s AI Overviews and AI Mode, ChatGPT search, and other AI discovery experiences do not operate as one identical system.
For a CMO, the useful distinction is between an answer generated from information learned during model training and an answer grounded in information retrieved during the current interaction.
Those are different paths to an answer. Publishing a page does not instantly rewrite a model’s memory.
In a search-grounded experience, a system can retrieve information, evaluate what is relevant, and synthesize a response. Google documents that its AI search features can also issue multiple related searches across subtopics, a process called query fan-out. [1]
Consider an illustrative B2B question:
“Which industrial supplier can help us replace an obsolete component, verify compatibility, and support ordering through our existing purchasing workflow?”
That question contains several needs: product availability, technical expertise, compatibility, integration, and service.
A page that says “We sell industrial parts” leaves most of that work unanswered.
Our strategic takeaway: build a public footprint that explains the actual decisions your customers are trying to make.
Your business needs to be understandable at the level of the problem it solves.
A keyword spreadsheet alone will not get you there.
2. SEO is still relevant. Lazy SEO deserves the funeral.
Every technology shift attracts the same parade.
Someone announces that everything you know is dead. Someone invents an acronym. Someone sells a dashboard. Someone raises their retainer.
Let’s skip the theater.
Google explicitly says the fundamentals of SEO remain relevant to its AI search features. There is no special technical requirement or magic AI schema that buys your way into an AI Overview. [1]
Technical accessibility, useful information, and a website people can navigate still matter.
What needs to change is the scope of your thinking.
A traditional ranking report asks where a page appears for a keyword.
An AI visibility program should also ask:
- Does the brand appear in relevant answers?
- Is it cited as a source, named as an option, or explicitly recommended?
- What does the system say the company does?
- Does that description match the business?
- Which competitors appear, and what evidence supports their inclusion?
- Which buyer questions expose gaps in our public information?
Those are separate observations. A source citation is not automatically an endorsement. A brand mention is not automatically a recommendation.
If your reporting collapses all three into one giant “AI wins” number, you have accounting problems before you have marketing insights.
At Lion + Panda, we use AISO, AI Search Optimization, and GSO, Generative Search Optimization, to describe work focused on brand presence and discoverability across these experiences.
The acronyms describe the work. They do not excuse us from proving what happened.
3. Your brand can be visible and still be misunderstood
Imagine an AI answer correctly names your company but describes an old service, puts you in the wrong market, or misses the capability that makes you valuable.
Congratulations. You have achieved confidently incorrect visibility.
For a CMO, that is a positioning issue.
Start with the facts:
What does the company do? Who does it serve? Where does it operate? Which products and services are current? What limitations should a buyer understand? Who has the expertise to support the claims?
Then look for contradictions across your website, business profiles, partner listings, biographies, product documents, and published content.
Our approach is to reduce ambiguity wherever we can control the information.
If your website describes you as a national specialist, your profile describes you as a local generalist, and your strongest technical capability lives in a PDF nobody can find, your marketing team has homework.
Consistent branding includes consistent facts.
4. Content needs to contain something worth knowing
AI has made generic content cheap.
Very cheap.
The internet did not need another 900-word article explaining that customer service is important. Now it can get 10,000 of them before lunch.
Your advantage is the knowledge your organization actually possesses.
The sales objections. The engineering explanation. The installation constraint. The failure pattern. The customer question that takes fifteen minutes to answer properly.
For a manufacturer, useful content might explain compatibility limits and maintenance requirements.
For a home services company, it might explain when a treatment is appropriate, what affects cost, and when a customer needs a different solution.
For a professional services firm, it might explain the process, the tradeoffs, and the circumstances that change the recommendation.
If a competent competitor could publish your article unchanged, ask what your company contributed.
At Lion + Panda, the content strategy should draw from the people who do the work. We organize their expertise into pages, comparisons, technical explanations, case studies, videos, and articles that help buyers understand their options.
A strong page answers the question directly, supports the answer, and explains the conditions under which it applies.
That structure helps the reader. We do not need to pretend a particular heading format secretly controls an AI model.
The editorial standard matters just as much when automation is involved. Google warns that generating many pages without adding user value can violate its scaled content abuse policy. [2]
An automated pile of mediocre pages is still a pile of mediocre pages.
Now it just arrives faster.
5. Your website is part of the evidence. The rest of the market gets a vote.
Your company calling itself “the trusted leader” is a claim.
An identifiable customer explaining a specific result is evidence.
A detailed case study with a timeframe, methodology, and limitations is evidence.
A legitimate industry publication discussing your work is another piece of evidence.
Our strategic view is that CMOs should manage both the company’s own information and the credible public material around it.
That includes accurate business profiles, authentic reviews, partner relationships, useful expert commentary, and content distributed where buyers actually look.
Google’s guidance supports keeping business and product information current and rejects inauthentic mentions as an AI search shortcut. [3]
The lesson is straightforward: earn a public footprint that can withstand scrutiny.
Do not manufacture a fake consensus and call it authority development.
Also, stop treating every platform as if it owes your brand the same result. What appears in Google can differ from what appears in ChatGPT. Track the environments relevant to your audience separately.
6. Check whether your technology is locking the door
Before buying another visibility tool, ask whether the systems you want to reach can access the information you already have.
Review crawl permissions, indexing controls, canonical URLs, internal links, rendering, and security rules. Confirm that important content is accessible and that your structured data reflects the visible page.
Google’s guidance makes accessibility and accurate implementation part of the foundation. [1]
There is also a distinction between search access and model training access.
OpenAI documents separate controls for OAI-SearchBot and GPTBot. A business can allow the search crawler while disallowing the training crawler. [4]
That is why “we blocked AI bots” is an incomplete answer.
Which ones? For what purpose? What business decision informed that setting?
Your CMO does not need to hand-edit robots.txt. Your CMO does need to make sure marketing and technology are not working toward opposite outcomes.
7. How Lion + Panda approaches the work
We approach AI search as part of a connected marketing system.
The job is to improve the information available about the business, make it accessible, strengthen its credibility, and measure whether the brand becomes more discoverable for the right questions.
That is where our proprietary tools support the strategy.
Boiler Plate: the website foundation
Boiler Plate is Lion + Panda’s standardized website system.
Its role is to support consistent implementation, ongoing updates, content expansion, and integration with the broader marketing operation.
A website becomes more useful when the team can improve it without turning every change into a construction project.
For AI search work, that foundation supports the practical implementation of page architecture, navigation, metadata, structured data, and connected content.
Boiler Plate does not purchase trust from a search engine. It gives us a controlled foundation on which to do the work.
The distinction matters.
A proprietary platform should make execution better. It should not require a supernatural explanation.
Our proprietary SERP infrastructure: coverage with a purpose
Lion + Panda’s proprietary SERP tool supports geographic, service, and industry page development, including AI-managed content, metadata, structured data, and internal linking.
The strategic purpose is to give important parts of the business an appropriate place on the website.
A company serving several markets and customer types should be able to explain those relationships clearly.
A service page can explain the work. A geographic page can explain relevant local availability and experience. An industry page can address that industry’s requirements.
But every page needs a reason to exist.
Swapping a city name fifty times is not local expertise.
Our standard should be substantive differences: real service coverage, relevant projects, distinct customer needs, meaningful technical information, or other facts that improve the page.
Automation supports production and maintenance. Human judgment owns the accuracy, usefulness, and business relevance.
The target is discoverability across meaningful needs. Page count is an operational number, not a strategy.
Pandalytics™: visibility needs an honest downstream picture
Pandalytics™ is part of Lion + Panda’s proprietary analytics and lead management stack.
Its AI lead-scrubbing capabilities help identify spam, bots, and low-quality submissions, with validation of contact information and service relevance.
That matters because a marketing report can look spectacular while the sales team is drowning in garbage.
For AI search, the first objective is brand presence and discoverability. Pandalytics supports the downstream evaluation when a person engages with the business.
We want to distinguish increased visibility from actual visits, inquiries, qualified opportunities, and revenue.
Those are connected stages. They are not interchangeable.
Pandalytics is not a window into an AI platform’s internal ranking system. It helps us evaluate activity within the marketing and lead process we can observe.
A dashboard should make reality clearer.
AISO + GSO: the strategy around the stack
The tools support the program. The program still needs direction.
Our approach brings together technical optimization, semantic architecture, structured content, authority development, and ongoing evaluation.
In practical terms, that means identifying buyer questions, reviewing how the brand appears, finding content and credibility gaps, improving the underlying information, and measuring again.
The proprietary advantage is how we execute and connect the work.
Nobody gets to guarantee that an independent AI platform will recommend a particular brand.
We should be able to explain our method without making that promise.
8. Measure visibility without inventing certainty
The dumbest possible AI search report is one screenshot.
A single answer is an observation. It is not a market study.
A useful measurement program starts with a defined set of questions based on customer research, sales conversations, and commercially relevant decisions.
Include unbranded discovery questions, comparisons, service and location questions, technical questions, and questions about your company.
Keep branded and unbranded results separate. Asking an AI platform about your business by name does not establish that it would discover you independently.
Record the platform, date, prompt, location context where applicable, search mode, and whether the test starts in a fresh conversation. Repeat the tests on a defined schedule.
Then distinguish:
| Measure | What it tells the CMO |
|---|---|
| Brand inclusion | How often the company appears within the monitored questions |
| Source citation | Whether company content is used as a supporting source |
| Recommendation | Whether the company is presented as an option for the buyer |
| Description accuracy | Whether the answer represents the business correctly |
| Competitive presence | Which other companies appear within the same monitored set |
| Referral activity | What identifiable AI-origin visits do on the website |
| Qualified outcomes | Whether observed inquiries become valid opportunities |
These measures describe the sample you monitor. They do not reveal your share of every AI conversation.
Referral tracking also cannot capture every influence. A buyer may discover you in an AI answer, then return through a branded search, a direct visit, or a phone call.
Our recommendation is to combine observable referral data, website analytics, CRM source information, and customer self-reporting. Treat the gaps honestly.
If you need to establish incremental business impact, plan an evaluation with appropriate comparisons and account for other marketing changes. A rising revenue chart beside an AI visibility chart does not prove causation.
Measurement should reduce uncertainty, not decorate it.
9. Give the first 90 days an actual job
“Let’s do some AI stuff” is how budgets disappear.
Here is how I would structure the first 90 days.
Days 1–30: establish the baseline. Define the priority audience and buyer questions. Review brand accuracy and competitive presence. Audit accessibility and the current content. Identify the business facts that need correction. Agree on what will be measured and who owns it.
Days 31–60: close the important gaps. Fix access and architecture issues. Improve the pages closest to priority buyer decisions. Publish original expertise and relevant proof. Use Boiler Plate and our SERP infrastructure to support implementation. Set clear review responsibilities.
Days 61–90: evaluate and refine. Repeat the visibility checks. Look at accuracy, inclusion, citations, and referrals separately. Use Pandalytics and the connected lead process to examine downstream engagement where observable. Prioritize the next changes based on evidence.
Ninety days is a working period for establishing the system and evaluating early signals.
It is not a guarantee of results.
Assign an executive sponsor, a technical owner, a content owner, and a measurement owner. Give them a shared backlog.
The agency, developer, sales team, and subject matter experts need to work from the same business priorities.
Otherwise, everyone is busy and nothing connects.
10. Ask better questions of whoever is selling you this
Before approving an AI search investment, ask:
- Which buyer questions and platforms are you prioritizing, and why?
- What exactly will you change on our website and across our public footprint?
- How will our actual expertise enter the content?
- How will you distinguish mentions, citations, and recommendations?
- What is proprietary about your system, and what does that improve?
- How will you measure progress without claiming access to private platform data?
- Who reviews accuracy and owns ongoing maintenance?
- What would make you change the strategy?
Listen carefully to the answers.
A capable partner can describe the work, the evidence, and the limits.
A weak partner will hand you a new acronym and hope you do not ask a second question.
Your brand needs to deserve a place in the conversation
AI search does not eliminate the CMO’s responsibility to understand customers, build a credible position, and connect marketing to the business.
It makes the consequences of weak information harder to ignore.
The brand needs clear facts. The website needs a usable foundation. The content needs actual expertise. The public footprint needs credible support. The measurement needs intellectual honesty.
At Lion + Panda, Boiler Plate, our proprietary SERP infrastructure, Pandalytics™, and our AISO + GSO approach support that connected work.
Our aim is to help brands become easier to discover, understand, and evaluate when buyers ask questions that matter.
Because being mentioned means very little if the answer gets your business wrong.
And getting traffic means very little if nobody can explain what it accomplished.
The only metric that matters is movement.
Make sure your AI search strategy can show you where it is moving the business.
Sources
[1] Google Search Central, AI features and your website.
[2] Google Search Central, Guidance on using generative AI content.
[3] Google Search Central, Optimizing for generative AI features.
[4] OpenAI, Overview of OpenAI crawlers.
