Our Meta Ads have stopped performing. Can you help us figure out what's wrong?

Absolutely. But we rarely start inside Ads Manager. When Meta Ads stop performing, it's often a symptom rather than the problem itself. Businesses begin optimising for cheaper clicks, higher ROAS or more traffic while losing sight of the customers they actually want to attract and grow. We start by understanding who your most valuable customers are, what they're trying to achieve and whether your advertising, website and customer experience are helping them make progress. Only then do we optimise campaigns. Better ads don't build better businesses. Better customer journeys do. Meta Ads don't stop working overnight. Businesses gradually lose alignment between who they're trying to attract, what customers need and the experience they deliver. That's what we look for first.

If we stop advertising, our sales will disappear. How do we reduce our dependence on paid media?

Paid media should be one part of a healthy growth system—not the system itself. We help businesses build stronger acquisition, customer experience, retention and referral engines so paid media can focus on what it does best: attracting and growing your most valuable customers. The result is a business that compounds growth instead of constantly buying it. The goal isn't to spend less on advertising. It's to build a business where every customer becomes more valuable over time.

We have traffic, but conversions are flat. What's stopping people from buying?

Most businesses treat all traffic as if it's the same. It isn't. Every visitor arrives with different intent, different context and a different level of readiness. Yet most websites deliver the same generic experience to everyone. The result is that businesses optimise for more clicks instead of more meaningful customer progress. Conversion isn't a single moment—it's a journey. We listen to the signals customers leave behind, understand what they're trying to achieve and design experiences that help the right customers progress towards meaningful action. The question isn't 'Why aren't people converting?' It's 'What are our best customers trying to achieve, and are we helping them get there?'

How do we move past plateaued growth?

Plateaued growth is rarely caused by one broken channel. More often, it's a sign that the systems which got you here weren't designed to take you further. We identify the constraint that's limiting progress, prioritise what matters next and build better systems that compound growth over time. Businesses don't outgrow markets. They outgrow their systems.

Everyone says we should be customer-centric. What does that actually mean?

Being customer-centric isn't about trying to serve everyone equally. It's about understanding which customers create the greatest long-term value, what they're trying to achieve and designing every interaction to help them make progress. The goal isn't simply to attract more customers—it's to attract, convert, retain and grow the customers you're best equipped to serve. Customer-centric businesses don't optimise for transactions. They optimise for customer progress—and profitable long-term relationships.

We have customer data everywhere. How do we make it useful?

Most businesses don't have a data problem. They have a listening problem. Customer signals exist across your website, CRM, support conversations, marketing platforms and purchase history. The challenge is connecting those signals to understand where customers are progressing, where they're getting stuck and what they need next. The most valuable data isn't what customers tell you. It's what their behaviour has been telling you all along.

How do we optimise our website for ChatGPT and AI search?

AI rewards businesses that communicate clearly, build trust and organise information around real customer needs. The businesses that will stand out aren't the ones producing the most content—they're the ones creating the clearest understanding of who they help, what they sell and why customers choose them. AI doesn't reward better content. It rewards better understanding.

We're growing, but everything feels harder than it used to. Is that normal?

Very. Growth creates complexity. Complexity exposes weak systems. The businesses that continue to grow aren't necessarily the ones spending the most on marketing—they're the ones building better operating systems that scale with their customers, teams and ambitions. Growth doesn't create chaos. It reveals it.

We know something isn't working, but we can't figure out what's holding us back.

That's where we start. Rather than jumping straight into solutions, we identify the constraint that's preventing customers and the business from progressing. Solving the right problem creates momentum everywhere else. The hardest problems aren't difficult because they're complex. They're difficult because they're usually the wrong problems.

How do we build a business that continues to grow?

Sustainable growth isn't built campaign by campaign. It's built by creating systems that consistently attract, convert, retain and grow your most valuable customers while learning from every interaction along the way. The businesses that grow the fastest aren't the ones with the biggest budgets. They're the ones that learn the fastest.

How to Build a Business AI Can Understand

🎧 Listen to this Insight

AI is becoming a new interface to commerce. The businesses that benefit most will not simply be the ones using AI. They will be the ones whose businesses are clear enough for AI to understand.

1. Business Problem

For years, businesses have optimised for people discovering them.

Search engines.

Marketplaces.

Social platforms.

Ads.

Email.

Retail shelves.

The next interface is increasingly conversational.

A customer may not begin with:

“Find me Brand X.”

They may begin with:

“I need something for sensitive skin that fits into a simple morning routine and doesn't contain ingredients I want to avoid.”

The interface then has to understand businesses well enough to answer.

That creates a new problem.

A business can have excellent products, strong reviews and years of expertise—and still be difficult for machines to understand.

The information may be scattered across product pages, PDFs, blogs, FAQs, collections, reviews, social posts and third-party platforms.

The business knows what it sells.

Its customers know what it does.

But the machine has to reconstruct the story.

That reconstruction is not guaranteed to be accurate.

AI can only reason effectively from the information and relationships it can discover.

This means commerce is moving toward a new requirement:

**Businesses need to become understandable systems.**

Not just discoverable websites.

Understandable businesses.

2. Why Conventional Advice Falls Short

The first response to AI discovery is often to create more content.

Publish more articles.

Add more keywords.

Create more FAQs.

Generate more product descriptions.

Write more social posts.

Some of this helps.

But content volume does not solve a knowledge architecture problem.

Imagine a brand sells 40 products.

Each product has ingredients.

Each ingredient has benefits, exclusions and sourcing information.

Each product serves different customer needs.

Some products are complementary.

Some should not be used together.

Some are suitable for particular routines.

Some are designed for specific customer progress.

The information exists.

But if the relationships are implicit, inconsistent or scattered, a machine has to infer them.

That creates ambiguity.

The answer is not necessarily more words.

It is better structure.

A well-understood business has clear entities, relationships, definitions and evidence.

It makes it easier for machines—and humans—to answer questions such as:

What is this product?

Who is it for?

What problem does it address?

What makes it different?

What does it contain?

How does it relate to other products?

What should someone use before or after it?

What evidence supports the claim?

What should a customer know before purchasing?

This is the difference between **content** and **business knowledge**.

Content communicates.

Knowledge structures meaning.

3. The Operator Perspective

Operators know that a business is much larger than its website.

There is a system underneath it.

Products.

Customers.

Orders.

Ingredients.

Suppliers.

Claims.

Policies.

Markets.

Channels.

Use cases.

Customer questions.

Experiences.

Operational constraints.

Commercial relationships.

The website is simply one interface into that system.

For a long time, most commerce architecture was built around pages.

Home page.

Product page.

Collection page.

Blog post.

FAQ.

But increasingly, we should also think in entities and relationships.

**Product → contains → Ingredient**

**Product → helps with → Customer Need**

**Product → supports → Customer Progress**

**Customer → purchased → Product**

**Customer → experienced → Outcome**

**Product → complements → Product**

**Brand → operates in → Market**

**Claim → supported by → Evidence**

This is much closer to how a business actually works.

It is also much closer to how intelligent systems can reason about a business.

The opportunity is not to build a website that talks about the business.

It is to build a commerce system that represents the business clearly.

4. What Makes a Business Understandable to AI?

There are seven layers to think about.

### 1. Clear Entities

Define the important things in the business.

Products.

Brands.

Ingredients.

Categories.

Customer needs.

Use cases.

Markets.

Policies.

People.

Experiences.

These should have consistent names and definitions.

### 2. Relationships

Define how those entities connect.

A product is not just a product.

It belongs to a category.

It contains ingredients.

It solves needs.

It may complement other products.

It may be suitable for particular customer journeys.

Relationships create context.

### 3. Consistent Definitions

A business should not describe the same concept five different ways.

If “sensitive skin” means one thing on the product page and another thing in the blog, the knowledge becomes ambiguous.

Canonical definitions matter.

### 4. Evidence

Claims need context.

Why should someone believe the product works?

What supports the ingredient claim?

What customer evidence exists?

What certifications or testing apply?

AI systems need more than assertions.

They need trustworthy information they can ground answers in.

### 5. Customer Context

A product does not exist in isolation.

It exists in relation to a customer need.

The better a business represents those relationships, the easier it becomes to answer intent-driven questions.

### 6. Freshness

Business knowledge changes.

Products launch.

Ingredients change.

Prices change.

Policies change.

Markets change.

Outdated information creates bad decisions.

A knowledge system needs ownership and maintenance.

### 7. Machine-Readable Structure

Important information should not exist only as beautiful prose.

Use structured data, consistent metadata, clean taxonomies and connected commerce objects wherever appropriate.

The goal is not to “hack AI.”

The goal is to make the business legible.

5. A Real-World Scenario: From Product Catalog to Knowledge System

Consider an organic beauty brand with 60 products.

The traditional approach organises them into collections:

Face.

Eyes.

Lips.

Skincare.

Sets.

The customer asks:

“I need a simple makeup routine for dry, sensitive skin that looks natural and works for a long day.”

A traditional catalogue may struggle.

The business has the information, but the relationships are not necessarily represented.

Now imagine the underlying system understands:

- Skin concern
- Desired finish
- Product type
- Ingredients
- Ingredient exclusions
- Routine role
- Wear characteristics
- Complementary products
- Customer Progress
- FAQs
- Reviews
- Evidence

Now the request can be interpreted as a journey rather than a keyword.

The system can reason:

**Customer need → constraints → suitable product category → suitable products → complementary routine → education → proof**

This is much closer to how a good human retail advisor thinks.

The important insight is that AI does not magically create this understanding.

The business has to create the underlying clarity.

AI becomes the interface.

The knowledge system becomes the foundation.

6. Practical Framework: Build Your Business Knowledge Graph

You do not need to begin by building a sophisticated technical knowledge graph.

Start with the business model.

### Step 1: Identify the Core Entities

List the things your business needs to understand.

For a consumer brand, this could include:

Brand.

Product.

Ingredient.

Customer Need.

Customer Progress.

Concern.

Benefit.

Routine.

Use Case.

Market.

Channel.

Claim.

Evidence.

FAQ.

### Step 2: Define Each Entity

Give each one a canonical definition.

For example:

**Customer Progress:** The meaningful change a customer is trying to achieve through interacting with the brand.

This prevents the same concept from being interpreted differently across systems.

### Step 3: Map Relationships

Ask:

What connects to what?

Product → Ingredient.

Product → Concern.

Product → Benefit.

Product → Use Case.

Product → Customer Progress.

Product → Complement.

Claim → Evidence.

This becomes the connective layer.

### Step 4: Assign Sources of Truth

Decide where each piece of information is maintained.

Product data may live in Shopify.

Customer data may live in the customer intelligence layer.

Editorial knowledge may live in the content system.

The important thing is that each concept has an owner and a trusted source.

### Step 5: Expose the Knowledge

Make important relationships available across the interfaces where they matter.

Website.

Search.

AI discovery.

Customer service.

Email.

Merchandising.

Internal decision systems.

### Step 6: Keep It Current

A business knowledge system that is six months out of date becomes a liability.

Create ownership.

Define review cycles.

Treat knowledge as an operating asset.

### Step 7: Connect It to Decisions

Finally, connect knowledge back to the business.

What should we recommend?

What should we build?

What should we explain?

What should we stop selling?

Which customers need help?

Which market should we enter?

That is when a knowledge architecture becomes a Commerce Operating System rather than a documentation project.

7. What Leaders Should Do Next

Do not start with:

“How do we optimise for AI?”

Start with:

**“Could a new employee understand our business from our systems?”**

Then ask:

Could they understand our products?

Our customers?

Our positioning?

Our claims?

Our product relationships?

Our markets?

Our policies?

Our customer journeys?

Our operational constraints?

If the answer is no, AI will struggle too.

Start small.

Choose one commercial domain.

For example, your product catalogue.

Create a structured map of:

Products → Ingredients → Benefits → Concerns → Use Cases → Customer Progress → Complementary Products.

Then expose that knowledge through your website and internal systems.

Once the model works, extend it.

Customer intelligence.

Content.

Operations.

Markets.

Partners.

This is how businesses become progressively more understandable.

And that creates an important strategic advantage:

**The business becomes easier for people and machines to navigate, interpret and act on.**

8. The Bigger Opportunity: From Website to Commerce Intelligence Layer

The website used to be the destination.

Increasingly, it is becoming one interface into a much larger commerce system.

Customers may discover a brand through search.

Ask an AI assistant about a product.

See a recommendation in social.

Visit the website.

Speak to customer service.

Buy through a marketplace.

Return to the brand later.

The interface changes.

The underlying business knowledge should not.

That means the long-term opportunity is to build a shared commerce intelligence layer underneath those experiences.

One understanding of:

The brand.

The products.

The customers.

The relationships.

The evidence.

The journeys.

The commercial rules.

The operational constraints.

Different interfaces can then use the same underlying intelligence.

This is a foundational idea behind Rekindle's **Commerce Operating System**.

The goal is not simply to make a website AI-friendly.

It is to build a business whose knowledge can travel across interfaces.

Rekindle Framework

Want to see how understandable your business is?

The ***Growth Capability Canvas** can help you map the entities, customer signals, growth capabilities and systems your business needs to make better decisions and create better experiences.

The first step is not AI.

It is clarity.

Key Takeaway

AI is becoming a new interface to commerce.

But AI cannot understand what a business has never clearly represented.

The competitive advantage will not come from publishing more words or adding another AI tool.

It will come from creating a business with clear entities, relationships, evidence, context and ownership.

**The businesses that become easiest for AI to understand may also become the easiest businesses for customers, employees and partners to understand.**

Build the underlying system first.

Then let the interfaces evolve.

Is your business understandable beyond its website?

A Rekindle Operator Session can help map the knowledge, customer context and commerce relationships underneath your business—and identify where structure could create leverage.

We can help you move from disconnected information toward a clearer Commerce Operating System.

**Build Better Commerce.**

Book an Operator Session with Rekindle.