Differentia Consulting

12/07/2026

Excited to announce that Sarah Beams – The Harris Poll UK will deliver the keynote at our Qlik customer day on the 15th October 2026, Reading.

Title: If AI Can’t Understand Your Business, It Can’t Recommend You

Registration is OPEN:

The ABCs of AI. Qlik Customer Day with Differentia Consulting, Qlik & partners. Reading, 15 October 2026 - Register your place today - FREE

29/06/2026

Your RSS feed just became AI's best friend

Atom/RSS feeds were all the rage in the 00s then, since the advent of more reliable feeds such as those from Google, less so.

Guess what? They are back. Part of Microsoft's NLWeb spec is to call on RSS feeds to propagate LLM learning.

The problem. In WordPress RSS feeds, aka feed is fixed on post type off post. Which by design excludes custom post types and also pages.

If you only use 'pages', then your feed offers slim pickings.

Solution. A quick webhook to include pages and CPTs in your feed and all of a sudden AI can keep updated on your site.

Switched archives off in a plugin? Feeds give AI what it needs in JSON format to know what your site is about.

If there is a hole in my thinking, it would be good to know. How have you addressed this problem?

29/06/2026

Busted: Schema DOES Work for Agentic Discovery

At Differentia Consulting on our VISEON.IO site we decided to do something interesting. We built dynamically generated pages in WordPress that have as their content, Schema artefacts, only.

There is nothing new in dynamic builds, but I wanted to test a thing, and ONLY used Schema for content.

The test was to see if Agentic Discovery Search tools like Perplexity would cite the dynamic pages.

It does. And, by the way, it did so within 12 hours.

So, this tells me that we have our worlds inverted. We need a better technical foundation for Agentic Discovery and Search ( the two terms being mutually exclusive and successive in nature during Agentic 'search').

Schema to most is now as important as SEO and if used properly can avoid Digital Obscurity via Agentic discovery process.

We wrote a paper on mapping and the risk of not being on the map. Since the Domesday book being off the map has been catastrophic. Recently not being on Google Maps / Business has been the same. Today the new map is your map and AI tools need to consume it. AI does not have time to read your content, but it can read your content if presented via a knowledge graph, contextually,, as I've tested.

Your E-E-A-T signals it can grab first hand for citations, links and results.

Unless you want to trade using Agentic Commerce, or want to be discovered by new buyers you can rest easy, but any business needing to grow should be looking to alter its position on Schema. To the point, maybe of Schema only content. The royal jelly of SEO.

29/06/2026

Context and Digital Catalog Management, MDM for AI

As we add metadata (structured data about data) To data catalogs, who will own these catalogs?

Catalogs become the interface between your business, its data, and AI. But who will own them, and underlying MDM?

There are two hundred million active websites that are clambering to deploy Content via Schema and JSon-LD files as Context for AI to understand via Json-LD knowledge graphs (Digital Data Catalogs)

Does this then become an BI team problem, an AI team problem or stick with Marketing, and potentially neglected?

Centrally, metadata will become the corporate portfolio, catalogs the interface for the information super highway, and automation. For three important pillars: discovery (replacing web search), conversations (replacing website search) and Agentic Commerce (replacing legacy e-commerce).

But if we have lack of MDM/Catalog/DQ ownership how can we achieve better outcomes?

This is a huge opportunity for BI teams to own digital MDM and catalogs. It is that which becomes the semantic interface to your data, structured or unstructured for AI enabled BI tools to work with.

Fyi Catalogs can be exposed in tools that have graph capability, like Qlik for example, using graph objects to view. This we do every day with our web Schema audit tool.

29/06/2026

Service Now and Qlik announce partnership: What does the new Partnership Actually Mean?

**Qlik + ServiceNow**

Qlik announced a formal partnership with ServiceNow at Qlik Connect 2026 this week. Here's our read on it as a Qlik partner.

**What's been announced**

Two concrete things:

Qlik's analytics engine now feeds insight directly into ServiceNow workflows and AI agents — pulling in cross-system context from ERP, CRM, supply chain, billing, and beyond, not just what's inside ServiceNow.

New Qlik metadata collectors for the ServiceNow Data Catalog improve data lineage visibility and governance across shared assets.

**Why it matters**

The direction of travel in enterprise AI right now is clear: agents that operate with richer, governed context make better decisions than agents that don't. ServiceNow is already where a lot of workflow ex*****on happens. Qlik is already where a lot of governed enterprise data lives. Connecting the two properly — rather than with brittle point integrations — is genuinely useful.

The metadata collector piece is less flashy but arguably more important in practice. Knowing what data your agents are working with, where it came from, and whether it can be trusted is foundational. That's often the gap between AI that works in a demo and AI that works in production.

**What this means for clients**

If you're running ServiceNow and have Qlik in your environment — or are evaluating either — this changes the conversation. The question is no longer just "can we connect these systems" but "what context do our agents actually need to act well, and do we have the governance in place to provide it reliably?"

[Full press release](https://www.qlik.com/us/news/company/press-room/press-releases/qlik-partners-with-servicenow-to-bring-trusted-enterprise-context-into-ai-powered-workflows)

29/06/2026

Business Intelligence shifts to Semantic Intelligence

From Clay Tablets to Agentic Intelligence: We are hitting the "Semantic Wall"

​Something to ponder. For millennia, tablets of stone measured outputs. From Sumerian grain receipts to the modern SQL database, our tools have always been "dumb" repositories, working as passive mirrors reflecting what already happened.

​But we are entering a post-dashboard era.

​What will your tools look like in a world of semantic intelligence?

​They won't be windows you look through; they will be partners you work with.

​In the old model, the human was the "Semantic Layer." You looked at a chart, applied your 'tribal' knowledge, and decided to act.

In the Agentic era, the tool must possess its own context. If the data doesn't have a "sovereign" definition—a shared, machine-readable understanding of what a "customer" or "unit" actually is—the AI doesn't just fail; it hallucinates at scale.

​We’re moving from:

​Passive Recording (The Tablet)

​to Active Insight (The Dashboard)

​to Autonomous Ex*****on (The Agent).

​The tools of the next decade won't be measured by their visualizations, but by their comprehension.

Consider this. If your data architecture can't explain itself to an agent without a human translator, it’s still just a piece of stone.

29/06/2026

SCAS share how BI makes a difference

Great to see Mark Green at SCAS share how data is used to help keep ambulances on the road and ready to save lives.

29/06/2026

Google is reading SCHEMA TXT API Endpoints

We inject SCHEMA TXT API Endpoints into our HTTP headers and Google is reading them. Our DUNS number is ONLY visible via this method.

What this means-finally is the chasm of uncertainty is crossed. Legacy Search and AI are melded when it comes to site search.

Interesting times.

Image is the response to a Google mobile search.

29/06/2026

What will you do differently now Google will officially move to agentic search?

We see the #1 obstacle in the market the abundance of content and lack of machine readable quality -context

On every site there sits structured data. It needs to be machine- readable, then AI can ingest and make sense of it. Given the Google swap to an AI centric search, announced last week, that should be an imperative.

Do you even audit for accuracy, completeness, and framework compliance?

Have you exposed your structured data via API and MCP for AI access on demand. For Google's agents to discover products, discuss features and functions, price and availability, and then transact.

If not, how are you expecting AI to discover, discuss and transact - interact with your site. It needs quality -merchantable data. Is yours ready? Or are you suffering from digital obscurity when it comes to AI knowing what it needs to cite your brand, product or service?

29/06/2026

Check and Fix your LinkedIn website address - 'HTTPS'

The pas on the internet has to be the millions of website links on LI, and other social sites that point to HTTP and not HTTPS,

Also marketers around the world can ensure that the website listed on their social media accounts is in the correct canonical format for their homepage. Does it need www, does it need a trailing ' /'?

Don't know how to check? Do ctrl U on your homepage and search for 'canonical' that will show the official format for your homepage. Cite it any other way and it is incorrect and will cause the need for a redirect.

It matters, a lot. To web crawlers, AI and bots each is a different 'page'.

You need to own this, and fix. Then you will not get a million redirects a year for the first page your customers visit. Your homepage.

AI will then trust your domains homepage too -win .

Data quality and good data hygiene matters. Don't make AI have to guess.