
.png)
There are tons of infographics and insights out there on how to use AI to improve your marketing and business processes. But what about the other side, when your customers are using agents and AI for their buying process?
AI is shifting how we think, work and do business. It’s been on my mind a lot, but in particular, I’ve been looking into how it’ll affect marketing a B2B business. How will B2B businesses attract customers when those customers are sending out bots and agents to do most of their procurement work?
My particular interest is B2B because that’s where my career has been focused over the past decade plus.
Individuals are starting to use AI agents to book holidays, buy clothes and electricals, entrusting them with credit card details and to make important financial decisions. There’s about to be an influx of retail agents from the likes of Google and Meta over the coming months.
In some cases, the human will never interact with the brand at all, apart from maybe a final approval to go ahead with the transaction. This means that the whole marketing funnel piece is being turned on its head. How are we meant to attract and nurture customers now? What rules should we be playing by?
There’s a lot that B2B can learn from B2C when it comes to agentic procurement and purchases, but there are still a good few differences between how the two might operate.
Generally, B2B transaction value is higher, typically there are more stakeholders and decision makers and the buying cycle is longer, but AI agents can still be involved and businesses need to be prepared for them.
For years, I’ve been harping on that B2B marketing is actually like B2C because you’re still talking to humans, even if you’re marketing to a business, it’s not some faceless corporate entity, you’re always talking to real people.
But what happens when there are no humans making the purchasing decisions? Is in-bound dead? Can we still use emotional advertising and messaging or do we have to become bland and functional?
I’m just a curious person here, I’m on a quest to find the answers, but because everything is moving so quickly, these ‘answers’ are sure to keep evolving.
But in this snapshot, I’m hoping to find out how B2B businesses can respond to agentic customers, maintain a brand and keep those leads and sales coming in. If you have thoughts, please send in your answers on a postcard!

The dead internet theory is something I’ve long found fascinating. If you’re not familiar with it, it’s the concept that if all content and interactions are automated and AI generated, there will be a point at which humans on the internet become a minority.
And if the AI algorithm is being taught on its own outputs, it’ll create more slop and the internet will just be a senseless mess. Time will tell on this one I guess!
When bots are ruling the majority of the internet, then what’s the point? Do we create a new humans-only internet with authentic connections… or is that what we call real life? I digress, this is philosophical talk for another day.
Let’s get back to the dead internet theory and how it might play out for B2B businesses and change the whole relationship between business and customer.
The Claude website
You’re the founder of a B2B SaaS and run up a website in Claude in a few hours. It’s automatically optimised for schemas and SEO and you’re finding that most of the traffic is bots from LLMs or agents doing their research on your services.
If no humans are visiting the site, then what’s the point in having a fancy website? It’s like the ghost kitchens built for Deliveroo, they don’t have actual premises (in this case, an attractive website) where people can visit, or indeed stumble across, instead they have a menu which people access though the Deliveroo app and a kitchen where they prepare the food.
The LinkedIn post
Or alternatively, you have a LinkedIn page for your B2B business. As a busy founder, you don’t have time for marketing and so outsource it to an agent to generate and publish your weekly posts. It looks like a potential customer has liked and commented on your post, but they’ve actually outsourced this to an AI agent. In this whole interaction, no humans have actually seen this content.
When AI bots and agents are generating the marketing content as well as consuming the marketing content, it creates an empty loop of machines talking to machines.
More questions than answers here, these are extreme scenarios but it’s worth keeping them in mind as they could well become a reality, impacting the fundamental way that we do marketing and how ‘customers’ interact with brands.
My focus for the moment isn’t on how to use AI in your own marketing processes, there’s a lot of stuff about that already.
I’m interested in how marketing will have to change when a potential customer starts outsourcing their B2B procurement decisions to an AI agent, when one day you wake up to discover that most of your inbound visits and enquiries come from bots and agents.
This change will involve a whole rethink on how companies market themselves and attract business and not enough people are talking about it yet.

Before we dive into how agentic commerce might affect B2B, it’s easier to understand how it’s likely to change B2C when individuals are using agents to help them with their everyday shopping decisions.
The most concrete examples of agentic commerce are coming from B2C scenarios: people are starting to outsource some of their purchasing decisions. There are varying degrees of autonomy but agents are more and more involved in the comparison, shortlisting and sometimes purchasing.
B2C agentic scenarios might include the following:
Shopping for clothes
“I’m going to a wedding and need some navy blue leather court shoes to match this dress, they must be closed-toe, with a heel that’s no higher than 2 inches. Please bring up a list of options, from UK sellers, my budget is £80.”
Home improvements
“Here’s a photo of my broken Hoselock spray gun, please find and order the cheapest replacement with the fastest delivery.”
Booking a hotel room
“Please book me a 2 night stay in a hotel room for two people. The hotel should be within 1 hour’s drive of Birmingham. It must be in a market town which has a highly-rated tea room for a cream tea, countryside walks nearby, spa facilities and breakfast included. The budget is £500, please book the cheapest weekend in October”.
While not everyone is entrusting their purchasing process to agents yet, we know that the way people are shopping is changing, 67% of shoppers now use LLMs for search and discovery when buying online, meaning the traditional Google Search and website browsing is becoming less prevalent.
There are still many questions on the B2C side about how commerce agents might affect the marketing, customer service and loyalty journey which you can read more about in this Substack article.
As a business owner, your first question should be: how do I build my brand, stay discoverable and maintain sales when the audience is no longer human and instead being taken over by agents?

Agentic commerce isn’t as simple as B2C for businesses buying from other businesses. On the whole, the B2B buying journey is longer, higher value and more complex.
B2C commerce agents will mostly be optimised for convenience and price, while B2B agents are a little more complicated in that they’re focused on risk, contracts, reputation, usage and compliance.
With 86% of business leaders claiming their companies now rely on AI agents for daily operations (Sifted), there is no doubt their function will spill into the procurement space.
Again, Sifted have said that fully agent-run workflows remain uncommon at the moment. Typically, agents are handling one or two steps in a larger process, with humans overseeing and managing the rest (Sifted - Rise of the AI Agents) but even out sourcing one or two steps in the procurement process will have a profound effect on the way that marketing is done.
Outsourcing even a couple of phases in the procurement cycle will fundamentally alter marketing strategy. Below is a breakdown of the 5 key stages in the B2B buyer journey, along with real-world agent examples:
Customers are using agents and bots to manage their social presence, from liking and commenting on articles, to auto-generating posts.
It’s becoming harder and harder to rely on social engagement to measure intent as there are fewer humans in the process interacting with your brand.
Real people are no longer being influenced or getting brand awareness through the traditional social routes.

The AI agent is becoming the inbox gatekeeper, summarising what matters and forgetting the rest.
This could mean that the traditional email nurture sequence isn’t quite as effective as it once was as you’re no longer vying for the attention of a curious human.

Customer agents might already be scraping the web and consolidating data on your brand and comparing your services against competitors before they have even filled in an enquiry form.
The ‘auto RFP’ is a real possibility: that you’re pitching for work with data and content that’s publicly available, without even realising it.
It’s likely that you’ll lose deals without ever knowing you were considered for them in the first place, just because some information was missing from your website.

There might be a time when your potential customer’s agent visits your website sending questions and queries automatically, then shortlisting autonomously based on specific criteria.
Fielding these queries manually will no doubt be quite arduous and you might miss your chance if you take too long to respond.
I wouldn’t always recommend fighting AI with AI but at this stage, if you’re starting to get an influx of agent queries, then maybe it’s time to consider it.

This one’s specifically for SaaS providers: renewal agents are now able to track and monitor internal usage data for all types of software, specifying exactly how much the software is being used, by whom and what functionality is needed.
When it comes to renewal, the agent is able to drive a hard bargain if certain features or licenses aren’t actually being utilised. It’s also able to write an RFP ready to send out to the competition if the software isn’t fit for purpose.
Having a few super-users on side and a brilliant account manager probably isn’t enough if the renewal agent data shows that your software is overpriced and hardly used.


Simplistically, the traditional marketing funnel has 3 parts…
At each stage of the funnel, you’re establishing a relationship with someone in your target market, from knowing about the existence of your brand, to building up trust then an eventual conversion.
But if you add AI procurement agents into the mix, the whole marketing funnel and what it stands for quickly becomes irrelevant, because AI agents operate differently to humans.
If an AI agent is asked to shortlist some providers and send out an auto-generated RFP, in a sense that means the middle of the funnel has been flattened.
In the classic B2B funnel, buyers spend weeks moving through blog posts, gated white papers and email nurture sequences. When the agent is involved, the entire search, comparison and recommendation phase can happen in seconds.
So humans are only really going to be actively involved at the top of the funnel and at the bottom.
At the top of the funnel some B2B buyers will have their own shortlist in mind after attending an event, or ask trusted peers in groups what their recommended tools are.
And at the bottom, once an agent has shortlisted and recommended a possible vendor, the human is likely to step in to evaluate brand reputation, legal safety as well as the ‘gut feel / je ne sais quoi’.
Spending all your time trying to ‘crack the code’ by thinking solely about attracting the AI agent will only lead to dull and homogenous companies.
To really market effectively when your B2B customers are sending out AI agents, you need to have a two-track system: one track built for machine-readable logic and another reserved for human trust.
Jamie Smith, expert in B2C personal agents stresses the importance of brands to create a connection with their customers. Agents risk making the transactional experience 'functionally excellent, but emotionally invisible', meaning you might provide an excellent service or appropriately-priced product but the customer won't remember who you are. In order to adapt, he says that brands will need to create emotional connections elsewhere, through demonstrating values, creating experiences and helping customers accomplish something meaningful.

It’s becoming clear that getting the attention of humans is becoming harder and harder. When you can create content at the click of a button, there’s no end in the supply but when it comes to giving it attention, we still have the same amount of hours in the day, with a million other things to look at.
But when you replace that attention with AI agents being sent out to engage with posts, carry out research on companies, create shortlists and negotiate deals on behalf of their human clients, then that changes the whole way we market and get sales in B2B.
I don’t believe that the B2B agentic commerce wave will happen overnight, and when it does happen, companies won’t be entrusting their whole purchasing decisions to an agent. There will ALWAYS be humans in the process.
And as a business owner dealing with in-bound customer agents, be aware of their imperfections. As Michelle Johnson says, the more AI agents that are used, the lower the success rate is. If 10 agents each have a success rate of 95%, then the end-to-end success rate is just 60%. So if you get an absurd question or request, it’s probably the agent’s fault and you’ll really need to reach out to a real human for clarification.
There will be a small number of companies entrusting everything to the AI agent, others who outsource a couple of purchasing stages, and another group who will be forever skeptical.
The old way of winning in B2B is to build up a content list and following, the new way is to get onto the AI Agent’s ‘list’ then work on winning the human over.
I am convinced that one day soon, there will be a rebellion on AI homogeny and brands which set themselves apart from others will be those that can prove their ‘humanness’ will succeed.
We’ve seen ‘analogue’ resurgences like this before…

Business owners and marketers will have to adapt to non-human AI agents but we must not turn this into a systematic and mindless checklist of an exercise. By focusing all our attention on how to win over the AI agent, don’t lose sight of the soul of your business, what makes it different and unique.
Here’s how we can adapt:
The way to win and keep the sales coming when your customers start becoming agents isn’t all about being technical and ‘marketing to the bots’. Realise that the agent can get you onto the shortlist but the human should often have the final say.
This is very much a work in progress, if you have any ideas about how B2B agent procurement will affect the future of marketing, please reach out. You'll also find some bitesize content on this on my LinkedIn page.
If you have a B2B company and need to figure out how to adapt your marketing for AI agents, get in touch to arrange a call with me.
" Are you ready to bring in a fractional marketing director for your B2B company? From ongoing support to one-off projects, strategy to delivery, harness marketing expertise at a fraction of the cost of an employee."
- Georgina Willison, Fractional Marketing Director, Marwhal.