Introduction

Hi everyone, my name is Vincent.

I am the co-founder and CEO at Rembrandt.

And as you can see, we're gonna talk today about AI market intelligence.

But before I get started, maybe to segment the audience a little bit, who here works on AI professionally on a daily basis?

Right, who here uses AI on a weekly basis?

Okay, great, all right.

Defining AI Market Intelligence

So yeah, we're gonna talk about AI market intelligence, and in short, what that means is AI agents that track your entire total addressable market, aligning leaders, go-to-market teams, and product teams to live market facts.

The problem that we're solving is that companies still navigate on static siloed reports, and this leads to sales wasting weeks and months on dead accounts, marketing, burning budget on narratives that do not land, and product shipping features that nobody asked for.

From Static Reports to Live Signals

1The solution is a live market intelligence layer that turns public data into clear, timed actions and insights.

The Three-Step Approach

And it's a three-step process, so we basically continuously track every prospect that you care about.

We detect intent signals through many different public data sources like job vacancies, news articles, web search, public tenders, customer reviews, and annual and quarterly reports.

And basically that leads to the ecosystem image that you can see here.

Ecosystem Overview and Data Channels

So on the left, this is the total addressable market.

We ingest every piece of public information that we can find.

And then we basically model it to the logic of each customer that we work with.

We then store every single piece of public data forever.

And we then make that information logically available and queryable through four different channels, of which three are displayed here.

One is a web application.

And by the way, I'll dive into each one of these in a second.

But one is the web application.

The other one is the CRM, obviously a single source of truth.

And then finally, the LLM of choice.

So whether you would like to use JetGBT, Gemini, Entropic,

Doesn't really matter.

And the fourth method is a weekly newsletter, basically updating you on everything that changed in the market that the AI agents have picked up.

Who Benefits Across the Organization

And this is something that can be used across the entire organization because whether you work in marketing, whether you work in sales, whether you work in product, whether you're a leader or maybe even a shareholder, you would still like to know what's going on in the market of the company that you're working at.

And this is what we enable companies to do.

I'm sorry?

We're gonna get to that in a second.

Case Study: Performance and Active Nutrition

So today, to make this a bit more tangible, I'm gonna talk about a case study of one of our current customers who is selling dairy ingredients to performance and active nutrition brands.

So basically, they sell an ingredient to a protein bar company somewhere in the world.

And to scope out what we're doing, we're tracking 400 brands in the performance and nutrition industry in North America on a weekly basis, with one single goal, which is to identify which are the top performers, who is going to win that specific market, because that's the company that our client would like to work with.

Scoping the Total Addressable Market

So we start by defining the actual total addressable market, so which companies are part of the market that our client would like to win.

And we do that together, but it's primarily the job of the actual company.

Normally they already know this, they have this information ready to go, but of course we need to set the parameters in order to start researching.

Collecting Public Signals

Then we start searching every piece of data.

Basically, to make it a bit more tangible and to answer your question, here are a couple of examples.

News and Partnerships

This is the actual news scanning that we do.

On the left, you can see David Protein, a protein bar company raising 75 million.

eyeing retail expansion, which is a super interesting signal because they're clearly having traction.

And on the right you can see another company called Pure Protein collaborating together with Walmart to get a huge distribution channel through all of their stores.

Job Vacancies as Rich Intent Data

But then a job vacancy, why a job vacancy?

Well, actually a job vacancy is full of super interesting information because a job vacancy is meant to, well, do like the first shift of people that look at it because if they don't think that's something for them, they're not going to even apply, right?

So you're gonna sell it, you're gonna sell your company, you're gonna explain what you're gonna want them to do, which KPIs you care about, what tools they need to be able to use.

the whole thing.

But in this case, just to make it tangible, 4LIFE Research is looking for a senior director of product development, and this entire document is full of interesting information if you're selling ingredients to performance and extradition companies.

But in short, this person is going to be responsible to review all the collaborations with external ingredients providers.

which is a big signal that they're going to have an overhaul in their entire collaborations with all their suppliers.

Additionally, they talk about their R&D initiatives, which is very interesting, because if they don't have their own R&D in-house, why even want to collaborate with them?

So there's a ton of information in here, and this is a specific example, but if you work in technology, for example, you would like to know which technology stack they use, which KPIs they care about, which job roles they're actually hiring for, et cetera, et cetera.

Does that answer your question?

We dive a bit deeper in a second as well in this specific example.

From Qualitative Noise to Quantitative Insight

1So then what we need to do is we need to transform all this qualitative data into quantitative metrics, because qualitative data is overload.

There's so much, and the LLM is really great at reading all of it, but you need to actually interpret it in a single, nicely, organisable metric.

Mapping Signals to the Buying Journey

So what we do is we scan through the sources, we then map all of these signals to the different buying journey phases that our client has identified.

And what is a buying journey phase?

It's basically how mature is the customer in the problem recognition and the search for a solution.

And we kind of reverse engineer what these signals look like in the early stages of looking for a solution and in the later stages of looking for a solution.

Awareness, Consideration, Optimization

In this case, awareness, consideration, and optimization.

Awareness, I have a problem and I need a solution.

Consideration, I have a problem and I'm actively looking for a solution.

And optimization is

I probably already have a solution in place, whether I build it myself or I bought something, but it's suboptimal to what we as a company would like to offer.

And then what we do is all those, we call that actually the intent quality.

Building the Urgency Score

And then in the urgency score, we combine the quality, the quantity, and the recency into one single metric.

So how mature are the signals?

How often do we actually pick up the signals?

And how recent are the signals?

And these three metrics combined

become one single metric, which I'll show you in a second.

Yeah.

We're going to go in in a second.

I'll explain here.

So let's zoom in a little bit.

Operational Framework and Pillars

And still, this is not the end level detail, because we go way deeper.

But just for the sake of this conversation, we need to keep it a little bit understandable.

So for this specific case, we have identified three pillars.

And by the way, we create a custom framework for each company that we work with.

Pillar 1: Organizational Growth

So the first pillar on the left is organizational growth.

Does the company have internal momentum?

So we're trying to figure out if they're actually growing as an organization and if they're actually doing the right things.

And the way we do that is by looking at the early stages, awareness, the initial hiring, and team formation.

So for example, I'm looking for a marketing manager who is responsible for social media, performance marketing, branding, and offline marketing.

That clearly signals that they don't have a very mature marketing department, so they're very early in their marketing efforts.

Now we're looking at the consideration phase.

So they're starting to hire someone who is in charge of social media, someone who's in charge of performance marketing, someone who's in charge of all these other channels.

So that's a signal that they're getting more mature.

They're hiring specialists, right?

So now we're looking at the optimization phase.

So they're looking for a director of marketing, someone who is really heavy profile, someone with a lot of experience, which shows that they're really building out the capacity on a marketing point of view, which means they can pay for all of those headcounts.

That's one signal, but second is,

they're taking the growth very, very seriously in this company.

And this is just one of the role developments that we track for this internal momentum that we're trying to capture.

Pillar 2: External Momentum

Then on the second level, we are tracking external momentum, brand and market presence.

Because what we want to know is, hey, this is a brand.

Are they getting regional exposure?

Are they getting national exposure?

Are they getting international exposure?

What's the level of exposure that they're getting?

How often are they being featured in the relevant news outlets?

Pillar 3: Expansion and Capitalization

And then finally, we're looking at expansion and capitalization.

So are they raising seed funding?

Are they raising series A funding?

Are they acquiring other companies to expand their presence?

Are they basically acquiring production facilities?

Are they opening up new locations?

What are they doing specifically?

Because they talk about all of this stuff all the time, because any chance they get to be in the news, they will grab, right?

So we pick up that data, we reverse engineer it into the

logic of our client to make sure that the client knows who to focus on and who not to focus on.

Is that more clear?

Compounding Market Memory and Why Not Just ChatGPT

The compounding market memory.

Weekly Tracking and Stage Shifts

Because we track every week the same companies, the same data sources, the same logic, we start a compounding market memory every week.

And this allows us to see who's heating up, who's cooling down, and who's shifting stage.

And in a perfect world, we can actually predict when the perfect buying moment actually will appear because we see them coming from so far away that by the time they get into from the latent need to the actual need, the sales team or the marketing team or whoever is in charge will be ready to strike.

Limitations of Generic LLM Research

And I hear a couple of you thinking, so I can just use ChatGPT or perplexity in this case to do deep research.

Why can't I do that?

Well, you can, but there's a couple of differences.

The first one is some signals only stay online for a bit, right?

So a job vacancy specifically is not there forever.

It's gonna be there a couple of weeks and then it's gone.

Second, ChatGPT, you ask it the same question five times, it will give you a different answer every single time.

And that is because it has such a wide array of solutions that it can apply to your question that it will actually search for a new way to solve your problem each individual time.

And what we do is we apply, again, consistent logic for consistent data sources, and we basically communicate everything together and then organize it in a way that it's very queryable by the actual LLM.

And that's something I'll show in a second.

So that's all nice, theoretical, like so many AI solutions.

So what is the actual output and how does it really add value?

Outputs That Drive Action

Territory Analysis and Prioritization

So, first of all, territory analysis.

Because you are a go-to-market motion, you wanna know, okay, today's my Monday, what am I gonna work on this week?

And everybody has 200 accounts on their name, where shall I start?

Right, well, actually, we know where to start because we apply that single metric, which we call the urgency score,

And we actually have buy and journey phase tracking, which is a compounded metric over time.

So we really get to know these prospects, so we really understand who you should be focusing on first to have the highest chance of success.

But also the second and the third.

And we might not be super right about every position, but over time, sample size of the data set grows, accuracy grows as well.

So the longer you track, the more accurate you can predict which company is most likely to buy your solution.

Pipeline Pulse: The Weekly Newsletter

Then the pipeline pulse, something we made up ourselves.

It's basically a weekly newsletter which shows you who is moving and why.

And on the left you can see a quick scan which is always the top of the newsletter which says for the accounts that are in your name, only those,

what happens, so which use case actually we picked up intent for, for what country, what the buying journey phase is after we picked up the intent, and then of course what the urgency score is after we pick up the intent, including the deltas because you want to know is this moving in the right direction or in the wrong direction.

And then on the right you see an example from David Protein where we actually showcase for each of these companies in that newsletter

why this matters now, so literally what we picked up and how it actually ties into the opportunity you're hoping to find, and what we've seen so far, which is a chronological timeline of all the relevant data points that add to that specific why this matters now, because you don't just want to see what happens this week, you want to see what happens across the board, because that ties the argument together.

Web Application and Agent Configuration

And then there's the web app.

So this is basically a video of our application where you can see the dashboard, which is the pine journey phase monitoring.

You can see the actual signals.

So in detail, as you can see here, the actual evidence quote that the AI agent found in that specific job vacancy, that was a reason to pick it out as an actual signal.

And then here you can see the agent itself.

So we picked up almost a thousand signals so far.

260 job vacancies, 728 news articles in nine weeks.

Here in the actual configuration of the agent, you can see the actual accounts we're tracking, you can see the sources we're tracking, and you can also see for each buying journey phase how many signals we have specifically.

So imagine every job vacancy we search for approximately 18 signals, every news article we search for approximately 25 signals for each of these brands on a weekly basis, and that's only possible because of AI.

Because I would have to do that myself, I would not have a fun job.

LLM Integrations and Example Prompts

And then of course the final one, which is the one I'll showcase in a bit more detail in a second, is the connection of our dataset into the AI chat of choice.

So whether you like OpenAI or Gemini or whatever you prefer to use, our dataset is structured and organized in a way that the LLM knows what to look for, and then you can actually apply whatever the logic is you have

that you want as an output to our data set and use the functions, the great functions of these LLM that are already there.

So if you work on market intelligence, you can ask questions like, how is the market going to develop in the coming years?

If you work in marketing, what message will best resonate with North American brands?

Or in sales, how should I approach David Protein, the company we used in the example before?

What that actually looked like in real life is demonstrated here.

So can you act as a trend spotter and make a detailed analysis on new trends in the functional food, beverage, and supplement market where you expect growth in the coming five years and explain why.

And this basically has all these signals, almost a thousand, organized and structured for each of these prospects.

And it interprets that with the actual knowledge that the LLM has already.

And it comes with super detailed conclusions with the evidence that we actually picked up in the data set of all of these prospects.

all of a sudden you have something that nobody in the market has because you've been tracking every company every week with your logic.

And if you work in marketing, you... Wait, I think...

Please tell me who we should target and what narrative would have the highest chance of success because you want to use the right messaging for the right company to make sure that they either get educated or get challenged depending on what their current beliefs are.

So we know exactly what they're thinking because we're tracking every single piece of data that they're putting out there.

And by us knowing exactly what they're thinking, we should be able to come up with marketing narratives that actually will strike home.

And that's exactly what you can do if you basically want to do that from a marketing point of view.

Applications for Sales and BD

And then finally,

There is what's in it for sales, business development, sales development representatives.

You can basically say, help me convert David Protein as a customer.

In this case, I applied a couple of frameworks, Medpick and the YUYU now, which are like sales frameworks that are widely adopted by business development teams.

And what it does, it uses our data set, it adds web search on top, and it says, okay, let's have a look what we know.

And then it starts applying these frameworks.

And eventually it will come up with a really, really highly personalized messaging which includes the actual pain points that we've identified and which includes a solution from the customer that we have in order to make the most.

Here you can see the actual messaging subject.

And this is basically all applications on top of the data set with the LLM that you already use, so there's nothing really special about it in that sense.

Conclusion

Final message, if we know each player in the market intimately, we can answer the questions that truly matter.

Thank you very much.