Ghostart: Helping businesses become the ones AI recommends
As customers increasingly turn to AI for recommendations, we designed and built Ghostart to help businesses understand how they're represented, identify opportunities to improve and take practical action.
Ghostart
Challenge
Businesses invest in websites, SEO and marketing, but often don't know whether AI recommends them, why competitors might appear instead or what to do about it.
What we did
Designed and built an AI visibility platform connecting monitoring, diagnosis, recommendations, execution and measurement, grounded in genuine business knowledge.
Outcome
A platform designed to help businesses improve their chances of being recommended without becoming AI visibility specialists.
The broader pattern this reveals
A potential customer asks ChatGPT to recommend a supplier. Three competitors appear. Your business doesn't.
Despite years of investment in SEO, websites and marketing, you might never make the shortlist.
AI visibility monitoring tools address part of this problem. They show where businesses appear and which competitors are recommended, but often leave owners to work out why and what to do next.
For smaller businesses without specialist marketing teams, another dashboard can simply create more work.
The challenge isn't just understanding AI visibility. It's knowing how to improve it.
Why a monitoring dashboard wasn't enough
With nearly 30 years' experience in digital marketing, SEO and content, we recognised the significance of this change in customer discovery.
The obvious solution was a dashboard showing where a business appeared in AI recommendations and how it compared with competitors.
But we challenged that assumption.
Knowing you're missing from a recommendation doesn't explain why. And identifying a possible reason still leaves someone to decide what to change and do the work.
We realised the opportunity wasn't helping businesses monitor AI visibility. It was helping them improve it.
The real problem
Information wasn't the same as understanding
AI recommendations aren't fully transparent. Businesses needed evidence to investigate potential visibility gaps without pretending to know exactly why an AI model produced a particular answer.
Insights didn't automatically lead to action
Identifying missing information or weak positioning didn't mean businesses had the time or expertise to address it.
Improvements required genuine business knowledge
Useful changes depended on real experience, customer evidence and expertise, not generic AI-generated content.
Progress needed to be measurable
Businesses needed to monitor changes over time, rather than rely on a single visibility assessment.
The challenge was to connect understanding, decisions and practical action.
How we approached it
We applied The Foundry's development approach: define the desired business outcome, map the process needed to achieve it, then identify where AI could make a meaningful difference.
1. Started with the outcome, not the technology
Business owners don't ultimately want an AI visibility score. They want more opportunities to be considered when customers choose suppliers.
We focused on commercially relevant customer questions, which businesses AI recommended and what the available evidence suggested about those responses.
That provided a more useful starting point than simply counting company mentions.
2. Mapped the process before building features
We worked backwards from the desired outcome.
What would need to happen after a business discovered it wasn't being recommended?
This became Ghostart's five-stage process:
Monitor → Diagnose → Recommend → Execute → Measure
Mapping the complete journey exposed important dependencies. Recommendations needed business knowledge. Improvements required evidence. Measuring progress required a record of what had changed.
These requirements shaped the system before we designed individual features.
3. Designed where AI should do the work
We examined which activities AI could handle and where people remained essential.
AI could research public information, analyse responses, compare evidence, identify potential gaps and help prepare improvements.
But it couldn't invent customer results, claim expertise a business didn't possess or make every publishing decision independently.
We designed the experience to do as much useful work as possible, asking customers for knowledge, evidence or approval when needed.
4. Built around what makes each business different
One risk was that AI optimisation could encourage businesses to produce more generic content, making them harder to distinguish.
We took a different approach.
Ghostart was designed to understand what each business does, who it serves, what it has achieved and what its people know.
This informed the Company Brain, Proof Stack and Second Brains, connecting business information with genuine experience and evidence.
It also drew on our Beige-ometer and The Beige Code thinking about preserving distinctive content.
Use AI to do the work, but not to invent the reason a business should win.
What it looks like in practice
From AI recommendations to understanding visibility gaps
Ghostart examines relevant customer questions, showing which businesses appear in AI recommendations and highlighting supporting sources.
Rather than presenting a score alone, it helps users investigate where they're missing and what the available evidence suggests.
From visibility findings to practical improvements
The system connects potential gaps to recommended actions, drawing on information about the business and available evidence.
AI helps research and prepare improvements, while people contribute expertise and approve important decisions.
Subsequent monitoring helps businesses observe changes in visibility.
What changed
One connected improvement process
Monitoring, diagnosis, recommendations, execution and measurement designed to work together.
Evidence-led recommendations
Findings connected to observable sources and genuine business knowledge, rather than unsupported assumptions.
Less work for business owners
AI helps prepare practical improvements instead of simply creating another list of marketing tasks.
The platform has been developed. Improvements in customer visibility cannot be guaranteed and are not presented here as measured outcomes.
Why this worked
The initial opportunity appeared to be monitoring a new marketing channel.
But a dashboard alone wouldn't address the work needed to understand and improve visibility.
By starting with the desired outcome, we identified why business knowledge, supporting evidence and practical execution needed to be part of the system.
The bigger achievement was designing a connected process rather than a collection of AI features.
What this proves
The value of AI isn't always in providing more information. Sometimes it's in connecting that information to the work required to make it useful.
Facing a similar challenge?
If your business has information or insights that aren't translating into meaningful action, the problem may be what happens after the analysis.
At The Foundry, we identify those gaps and build practical AI systems that connect insight to action.
Let's talk about turning your business insights into progress.