What Really Works in B2B Tech Marketing (and What Doesn’t)

I recently joined Jakob Löwenbrand on the Tech Marketing Trends podcast to talk about what actually works in B2B tech marketing today – specifically when you’re selling into complex enterprises with long cycles and large buying committees.

That conversation reinforced a pattern I see over and over again: we over-complicate channels and tactics, and under-invest in the simple question that matters most:

Who should we be talking to next – and why?

This post is my attempt to take the themes from that discussion and turn them into something practical: how I think about B2B tech marketing in 2025, what’s changed, and what hasn’t.

The illusion of the “channel silver bullet”

If you hang around B2B marketing long enough, you’ll notice the same cycle:

• A channel or tactic starts working for a few companies
• It gets talked about at conferences and on LinkedIn
• Everyone piles in
• The early movers ride the wave; the latecomers see diminishing returns

We’ve done this with:

• Programmatic display
• LinkedIn thought leadership
• Content syndication
• Outbound sequences
• Now, of course, AI-generated content

The problem isn’t the channels themselves. The problem is when we treat them as silver bullets instead of as parts of a system.

The fundamentals haven’t changed:

• Who are we trying to reach?
• When are they likely to care?
• What story are we telling that’s actually different?
• How do sales and marketing work together around that?

If you’re unclear on those, no channel will save you. You’ll just burn budget faster.

Start from “who should we be talking to next quarter?”

Most tech marketers still think in terms of leads, forms and campaigns.

Enterprise sales teams don’t.

If you’re selling into large, multi-location organisations, the world looks very different:

• You close accounts, not leads
• There are multiple stakeholders with different priorities
• Opportunity timing is often dictated by external events, not your calendar

That’s why I always start with a simple, slightly uncomfortable question:

“If we looked at your market today, which specific companies should you be talking to over the next 90 days?”

If you can’t answer that reasonably well:

• Your ideal customer profile is probably too broad
• Your segmentation is likely cosmetic
• Your marketing and sales motions are misaligned

When we built Radiate B2B, we started from the account-level view: help companies see which organisations are in market, who inside those organisations is waking up to a problem, and how warm they already are to your story. (Radiate B2B)

Everything else – advertising, content, outreach – sits around that.

Advertising as infrastructure, not campaigns

Most B2B companies still run advertising as if they were in the 2010s:

• Big bursts around events or product launches
• Broad targeting based on job titles and industries
• Reporting that focuses on impressions, clicks and maybe some form-fills

The result is predictable: bursts of activity with long troughs in between, lots of noise, and very little clarity about what actually moves pipeline.

I’m more interested in advertising as infrastructure around your pipeline:

• Always-on, but focused on your named accounts
• Turning up and down around signals of interest, not arbitrary dates
• Designed to support sales territories and priorities, not live in a separate world

In practical terms that means:

• Your named accounts should see you regularly, long before they talk to sales
• When an account starts researching your category, your presence should intensify
• Sales should feel “pulled” into conversations because people have seen you, not pushed

If marketing can’t show how advertising changes what sales does next week, you probably don’t need more spend; you need a better system.

Using data and AI to change behaviour (not just dashboards)

It’s easy to talk about data and AI. It’s much harder to use them in a way that changes how teams behave.

The role of data and AI in B2B tech marketing – for me – is not to automate everything or to replace human judgement. It’s to:

1. Spot patterns earlier than humans can
2. Surface those insights in simple ways
3. Trigger changes in who we talk to and what we say

Examples:

• Identifying accounts that suddenly start consuming content related to a specific problem
• Highlighting clusters of stakeholders inside a company engaging with your site and ads
• Grouping companies by “readiness” rather than by static firmographics

What AI is good at in this context:

• Pattern recognition across large, messy datasets
• Summarising signals into something a human can act on
• Automating routine decisions (e.g. “when X happens, increase ad pressure for Y accounts”)

What it’s not good at (yet):

• Designing your go-to-market motion
• Deciding which customers you actually want
• Crafting a narrative that aligns leadership, sales and marketing

AI amplifies the quality of your underlying decisions. If your ICP, messaging and process are fuzzy, AI will help you do the wrong thing faster.

Designing for sales and marketing at the same time

One thing Jakob and I talked about is how often organisations still design for marketing or for sales, but not both together.

You see this in:

• Tools chosen purely by marketing, with sales as an afterthought
• Sales enablement content created without any feedback from marketing
• Attribution models that make marketing look good and sales look bad (or vice versa)

When you’re selling into complex enterprises, that split is deadly. The buying journey doesn’t care about your org chart.

So I tend to look at everything through a simple lens:

“If we do X, how does that show up in both marketing’s world and sales’ world?”

For example:

• Launching a new campaign → which accounts are we trying to warm up, and how will sales know when they are warm?
• Implementing new intent data → how will this change territory planning and outreach sequences, not just dashboards?
• Changing the narrative → how will this alter the conversations sales has, not just the headlines on your website?

If an initiative doesn’t have a clear answer on the sales side, it’s probably not ready.

Three questions every B2B tech CMO should be asking

If you’re running marketing for a B2B tech company, especially with enterprise deals, I’d start with these three questions:

1. Can we list the 50–200 accounts that matter most in the next 6–12 months – and are sales and marketing aligned on that list?
2. Do we know which of those accounts are warming up right now, and why?
3. If I shadowed a salesperson for a week, would I see our data, AI and campaigns changing their priorities and conversations?

If the answer to any of those is “not really”, you don’t necessarily need more tools or headcount. You probably need:

• A sharper ideal customer profile and segmentation
• A simpler view of account-level intent and engagement
• Clearer operating rhythms between sales and marketing

A simple 90-day experiment

If you want something concrete from this, try a 90-day experiment:

1. Pick a focused segment

   • 50–150 accounts that match your ICP and are meaningful for sales.

2. Agree the list with sales

   • No marketing-side “secret lists”. One shared list, one definition of success.

3. Build a basic signal layer

   • Combine whatever you have: website behaviour, email, ad engagement, intent data.
   • Score accounts at a very simple level: cold / warming / hot.

4. Run always-on, account-focused advertising

   • Modest budget, tightly targeted to those accounts.
   • The goal is familiarity, not just clicks.

5. Create a regular sales–marketing sync around the list

   • Weekly or bi-weekly, short and focused:

     • Which accounts moved from cold → warming, warming → hot?
     • What did they engage with?
     • What will sales do about it this week?

6. Measure pipeline movement, not vanity metrics

   • New opportunities from the list
   • Acceleration of existing deals
   • Quality of conversations

At the end of 90 days, you’ll know far more about what actually works in your context than any generic report can tell you.

Why this matters beyond Radiate B2B

Radiate B2B is one expression of these ideas in software – a way to bring targeting, advertising, data and sales alignment into one place.

But the principles apply whether you use us, build your own stack, or work it out on spreadsheets:

• Start from who you want as customers, not from generic channels
• Treat advertising and content as infrastructure around your pipeline, not campaigns in isolation
• Use data and AI to change behaviour, not just create dashboards
• Design everything for sales and marketing together, not one at the expense of the other

The tools will keep changing. New channels will appear. AI will get better.

The job stays the same:

Figure out who matters. Understand when they’re ready. Tell a story that makes sense in their world. And build systems so your teams can do that consistently.

That’s where I’m spending my time – through Radiate B2B, through consulting and fractional work with B2B tech companies, and through mentoring leaders who want their marketing to be something sales can actually feel, not just something that looks good in a slide deck.