Personalised Marketing in B2B — What AI Makes Possible and What Still Requires Human Judgement

Personalised Marketing Insights: The B2B Truth Nobody Wants to Admit

1. Why Personalisation Is the Wrong Starting Point
2. Using Data to Inform Strategy
3. Audience Segmentation and Targeting
4. Creating Content That Works
5. The Personalised Marketing Myth — and What Actually Works
6. Key Takeaways
7. FAQs

1. Why Personalisation Is the Wrong Starting Point

Every vendor selling marketing technology wants you to believe the same thing: that if you could only know more about each individual buyer — their behaviour, their preferences, their browsing history — you could tailor your message so precisely that deals would close themselves. That belief has cost B2B businesses billions. And most of them still have nothing to show for it.

We are told personalised marketing is the answer. But in B2B, it is almost always the wrong question. Gartner predicted that by 2025, 80% of marketers who invested in personalisation would abandon their efforts because of a lack of ROI, the complexity of customer data management, or both. That prediction has largely come true. The technology works. The strategy behind it does not.

I started my career cold calling at 18. I ran technology businesses for 30 years and watched marketing budgets get consumed by ideas that sounded sophisticated but produced nothing. Personalisation technology is one of the most expensive of those ideas. It presupposes you can identify who is in-market, what they want, and when to reach them — when the reality is that 95% of your market is not actively buying at any given moment. You cannot personalise your way around that fact.

The answer is not more data about individuals. The answer is consistent, visible, credible presence in front of your entire total addressable market — so that when a buyer does enter a buying cycle, you are already known to them. That is what B2B Performance Marketing is actually built to do. And it is what this article will walk you through.

2. Using Data to Inform Strategy

Data has a role. But most B2B businesses use it the wrong way — chasing individual-level signals rather than understanding market-level patterns. Before you attempt to personalise anything, you need to understand what your audience is actually doing in aggregate.

• Use Google Analytics 4 and Google Tag Manager to track how visitors behave on your site — which pages hold attention, which ones lose it, and where your content is earning genuine engagement versus where it is not.
• Identify patterns in what topics attract the right type of company, not just the highest volume of clicks.
• Monitor which content formats — video, articles, podcasts — drive return visits versus one-off reads.
• Use that intelligence to sharpen your content strategy and your outreach cadence, not to build individual buyer profiles you will never fully trust.

The data tells you what is resonating with your market. That is enough to act on. You do not need to know who specifically is reading. You need to know that the right kind of business is engaging with what you produce.

3. Audience Segmentation and Targeting

Segmentation matters, but not in the way most marketing teams practise it. In B2B, the relevant segmentation happens before you create a single piece of content. You are defining who you are talking to at the market level — not building individual buyer profiles.

Think about it this way. According to our research, 83% of B2B buyers complete their research digitally before speaking to anyone. By the time a buyer picks up the phone or fills in a form, they have already formed a view of the vendors they are interested in. The Anonymous Buyer problem is real — and no amount of personalisation technology solves it. What solves it is being visible, credible, and consistent throughout the period when buyers are researching but not yet identifiable.

The practical segmentation steps that actually work in B2B are these:

• Firmographics first: Industry, company size, geography, and likely revenue. These define your total addressable market.
• Buying authority: Who has the budget and authority to make a decision — CEO, CFO, VP of Operations? Write for those people, not a generic "decision-maker."
• Problem-first positioning: What specific operational or commercial problem does your product or service solve? Segment by problem, not by persona demographic.
• Platform targeting: Use LinkedIn's targeting capabilities to reach your defined segments by job title, company size, and industry. It remains the primary platform for B2B audience reach — LinkedIn drives 75 to 85% of all B2B leads generated through social media.

4. Creating Content That Works

Once you know who you are talking to, build content that earns their attention. Not content that talks about your product. Content that addresses the problems they are sitting with right now — the ones they are already searching for answers to.

• Copywriting: Write directly to the buyer type you have defined. Every article, every email, every LinkedIn post should sound like it was written for one specific person even if it reaches thousands. Read more in our Marketing-Tactics articles for practical guidance on copy that converts.
• Visuals: AI image generation has matured significantly. Midjourney v7 leads for artistic quality and brand visual work. Adobe Firefly is the safest choice for commercial use — it is trained only on licensed content and provides copyright indemnification. DALL-E (now integrated directly into ChatGPT) is the most accessible starting point. For photorealism, Flux is worth serious consideration. Use the tool that fits your workflow and your legal requirements — they are genuinely different in what they do well.
• Video: A weekly live stream or a regular video series is the single most efficient way to demonstrate knowledge at scale. Video production has been transformed by AI — platforms like Higgsfield give you cinematic AI video generation from text or image prompts, without a production crew.
• Podcasts: A podcast works as a long-form credibility vehicle. It gives buyers something to listen to across an entire buying cycle. It compounds over time in a way that a single sales call never can.
• SEO and organic search: Google's ranking now rewards demonstrable expertise and page experience — Core Web Vitals on a well-built standard page outperforms anything more complicated. If you are still maintaining AMP pages, you can stop. Google removed preferential treatment for AMP in 2021 and that advantage no longer exists. Fast, high-quality standard pages are what rank now.

AI tools — Claude, ChatGPT, Gemini — can accelerate content production significantly. They are good at drafting, structuring, and adapting content at pace. But they amplify the model you give them. If your strategy is unclear, AI produces more unclear content faster. Fix the strategy first. Then use AI to execute it.

5. The Personalised Marketing Myth — and What Actually Works

Here is what every MarTech vendor omits from their pitch: in B2B, you are almost never going to know who is reading, watching, or searching. The buyer is anonymous for most of the process. Gartner's own data shows that B2B buyers spend just 17% of their total buying time in direct contact with potential vendors. The other 83% happens without you knowing they exist.

So the question is not "how do I personalise my outreach to known individuals?" The question is "how do I make sure my business is in front of every potential buyer, consistently, so that I am already known when they decide to engage?"

The answer is straightforward, even if the execution takes discipline.

Start with your total addressable market. In the UK, there are approximately 46,000 businesses employing 50 or more people — medium-sized companies with 50 to 249 staff and large companies with 250 or more. If you sell B2B and your threshold is businesses with a turnover of roughly £5m and above, that is your universe. You can buy a list of that market from a reputable data broker tomorrow.

Then set up two parallel campaigns. First, an email sequence that invites your total addressable market to watch your live stream show or listen to your podcast — consistently, every week. Second, LinkedIn banner advertising targeted at the same audience by job title, company size, and industry. LinkedIn's visitor-to-lead conversion rate is 2.74%, comfortably ahead of other platforms, and its targeting precision for B2B is unmatched.

Of your total addressable market, roughly 1% to 15% will be somewhere in a buying cycle for your category of product or service at any given time. They will not all buy — some decisions are a flat no. But if 46,000 businesses are your universe and even 2% are actively evaluating something in your category each week, that is 920 potential buyers you have the opportunity to stay visible to. That number compounds over months.

The point that seems counterintuitive but is supported by everything we have observed over 30 years is this: in B2B, you do not need deep personalised marketing insights at the individual level. The volumes are manageable. The market is definable. What you need is reach, consistency, and a reason for buyers to pay attention. A weekly show does that. A podcast does that. A credible content engine does that.

Personalisation at the individual level is a consumer marketing concept applied to B2B because MarTech vendors needed a bigger market to sell into. It inflated GTM team sizes, added software costs, and produced form-fill rates that have consistently sat between 2% and 5% — a figure that has not moved in a decade despite billions spent on optimisation. That is not a technology problem. It is a strategic one.

6. Key Takeaways

1. Personalisation technology has not delivered the ROI promised — Gartner predicted 80% of marketers would abandon it by 2025 due to lack of ROI, and that prediction proved accurate.
2. 83% of B2B buyers research digitally before speaking to anyone. Be present and credible throughout that period, not just at the point of contact.
3. Define your total addressable market precisely — around 46,000 UK businesses employ 50 or more people. That is a manageable, reachable universe.
4. Use email campaigns and LinkedIn advertising to invite your entire market to a weekly live stream or podcast — consistently and at scale.
5. Create content that speaks to the problems your buyers are sitting with, not content that describes your product. Use AI tools to produce it faster, but make sure the underlying strategy is sound first.

7. FAQs

Q: What live streaming technology do I actually need to get started?

A: You do not need much. Blackmagic Design produces affordable broadcast-quality camera and switching hardware that works well for a weekly show. Pair it with a platform like StreamYard or Riverside.fm for multi-guest capability and recording. The technical barrier is lower than most businesses assume.

Q: How do I create effective LinkedIn adverts for a B2B audience?

A: Keep the creative simple and direct. Test two or three versions — different headlines and images — and let performance data tell you which works. Single image ads and video ads both perform on LinkedIn. The copy should address a problem, not promote a product. LinkedIn's own data shows that video posts achieve a 20% engagement rate on the platform, which makes a short clip from your live show a natural ad creative.

Q: Can AI tools genuinely help with B2B content production?

A: Yes, practically. ChatGPT, Claude, and Gemini are all capable of producing well-structured drafts, repurposing long-form content into shorter formats, and maintaining a consistent output volume. For visuals, Midjourney produces strong results for brand imagery, Adobe Firefly is the safest option if copyright indemnification matters to your business, and DALL-E integrated into ChatGPT is the most accessible entry point. For video, Higgsfield gives you AI-driven cinematic video generation from prompts or still images. The tools are genuinely useful. The strategy they execute still has to be yours.