Are you still treating brand and demand as a budget tug-of-war? AI-search has made that a lot more expensive

Every home has a heating bill and somewhere behind the wall there’s insulation you can't see. Heating is the bit everyone notices. It's immediate, easy to point to, and you can turn it up whenever you get cold. Insulation doesn’t cross your mind that often, at least not until the bill arrives. That’s because it's not doing anything visible, it's just quietly determining how much heat you lose and how hard the boiler has to work to replace it.

Most B2B marketing budgets are run similarly. Demand generation is the heating: quick, measurable and satisfying to report on. Brand building is the insulation: slow, invisible and the first line item cut when a quarter gets tight. The problem is that a house with no insulation doesn't just feel colder, it costs more to heat every single month for as long as you own it.

That's the uncomfortable truth behind the latest B2B marketing research and it's come into sharper focus in the last two years. Buyers spend much of their research time inside AI tools before a vendor ever hears from them, and those tools decide who gets mentioned largely based on brand signals. Treating brand and demand as competing budget lines in an AI-search world is effectively an act of self-sabotage.

Are you still running the heating flat out anyway?

Les Binet and Peter Field's original IPA research put the optimal split for consumer brands at roughly 60% long-term brand building to 40% short-term activation.¹ When the LinkedIn B2B Institute commissioned the pair to run a B2B-specific cut of the data, the number shifted slightly to around 46% brand building and 54% activation, which reflects B2B's longer sales cycles and larger buying committees, but the core finding remained: under-investing in brand quietly punishes your future growth.²

Where this gets interesting is in WARC's 2024 study of marketing budgets, where they found that actual spend had inverted the recommended ratio entirely: 68.8% of budget went to short-term performance tactics, leaving just 31.2% for brand.³ In the same survey, respondents said their ideal split was close to 50:50. Marketers know the ratio, but they’re not tuning the boiler that way.

Turn the heating up and you feel the benefit instantly, so there's less debate about funding it. But if you skip the insulation, the boiler has to work even harder to keep the room at the same temperature. In marketing terms, you're left dealing with rising customer acquisition costs. Worse still, some rooms never warm up at all, which means you're out in the cold and not on the buyer's shortlist.

Being found by AI now depends on the brand-building many brands are skipping

LinkedIn's B2B Institute, working with the Ehrenberg-Bass Institute, published research in December 2025 arguing that B2B brands need to shift from "rented prominence" (placements you only occupy while you're paying for them) to "owned prominence": distinctive, recognisable brand assets that get you found without ongoing spend.⁴ It's the direct B2B translation of insulation versus heating. Rented prominence is warmth you're paying for by the day, whereas owned prominence is the thing that keeps working after you've stopped paying attention to it.

The reason this now matters commercially, and not just philosophically, is because of AI-search. Seer Interactive analysed 3,119 informational search queries between June 2024 and September 2025 and found that when Google shows an AI Overview, organic click-through rates for the ranked pages beneath it fell by 61% and paid click-through rates tumbled by 68%.⁵ But brands that get cited inside the AI Overview saw a 35% higher organic click-through rate and a 91% greater paid click-through rate than brands that didn't, on the same queries.⁵ Citation isn't determined by bid or by keyword density, but rather whether an AI model already recognises your brand as an authoritative, well-referenced answer for the category, which is exactly what brand building is supposed to produce.

6sense's 2025 Buyer Experience Report, based on nearly three years of longitudinal data and almost 4,000 B2B buyers, adds the buyer-behaviour half of the story. 94% of B2B buyers now use LLMs somewhere in their purchase journey, and 85% of buyers have prior experience with the vendors they end up evaluating, the highest level 6sense has recorded.⁶ Their conclusion is rather blunt: on average only around 10% of your total addressable market is in-market at any given moment, and 77% of buyers contact the vendor who eventually wins first. If you're not already recognised when the other 90% enters the market, the report argues that "you've likely already lost."⁶

None of this replaces demand generation. 6sense's own data shows buyers still average 16 interactions with the winning vendor and evaluate 5.1 vendors before deciding, so the conversion, nurture and sales-enablement work of demand still has to happen once a buyer is in motion.⁶ But it confirms what the insulation analogy has been arguing all along, which is that the heating only works efficiently in a house that's already insulated. Demand converts interest that brand has already created, but it's not particularly efficient at creating interest from scratch.

Why B2B teams keep overlooking the insulation anyway

A few reasons this gap persists, even among marketers who can quote the Binet and Field numbers back to you:

• Brand results arrive on a timeline finance doesn't reward. Quarterly targets reward demand, not brand, because brand's payoff shows up in next year's CAC, not this quarter's pipeline report.
• Attribution tools were built to measure short term, not long term. Last-click and form-fill attribution can trace a demand campaign to a deal. They were never designed to trace the years of brand recognition that got a buyer to type your name into ChatGPT in the first place.
• AI citation looks like an SEO problem, so it gets routed to SEO. But the Princeton 2024 GEO study found that what improves AI citation is largely the same thing that improves brand recognition everywhere else: original data, credible third-party sources, and consistent presence across authoritative platforms, not keyword optimisation, which actually reduced visibility by 10% in their testing.⁷

How Invera builds both at once

Rather than running brand and demand as two budgets with two owners and two sets of numbers, Invera treats them as one engine with two cylinders. The brand cylinder earns citation and recommendation inside AI-assisted search, builds awareness across your ICP through programmatic display and paid social, and establishes authority through thought leadership from your own executives. The demand cylinder captures and converts the intent that brand has already created, through quality content syndication, multi-touch nurture, and sales enablement that hands leads over with context instead of cold contacts with a snowball's chance in hell of converting.

Both report into a single account-level dashboard, so you can see brand spend priming an account and demand converting it, on the same view, at the same time. Critically, AI-search visibility isn't treated as a bolt-on SEO task. It sits alongside brand and gets benchmarked against the category entry points that actually decide whether an AI model cites you, which is the mechanism the research above points to.

Close the gap between what you know and what you fund

None of this is new. Binet and Field have been making versions of this argument for over a decade. What's new is what's at stake. Underinvesting in brand used to mean higher acquisition costs. Now, with AI doing the shortlisting, it can mean you're not even in the running.

The fix isn't abandoning demand generation, any more than the fix for a cold house is ripping out the boiler. It's making sure the insulation gets funded on the same schedule as the heating, tracked with the same seriousness, and not treated as the thing you cut first when the pressure on numbers hits. It's worth checking which one your budget has actually been running on, not which one you'd say if asked.

If you'd rather have that conversation with someone that builds always-on, full-funnel brand and demand programmes for a living, Invera is a nifty place to start.

Sources: ¹Les Binet and Peter Field, The Long and the Short of It, IPA, 2013. ²Les Binet and Peter Field, and the LinkedIn B2B Institute, The 5 Principles of Growth in B2B Marketing. ³WARC, Performance budgets rise at the expense of brand, 2024. ⁴LinkedIn B2B Institute and Ehrenberg-Bass Institute, Easy to Find: Being Where B2B Buying Happens, 2025. ⁵Seer Interactive, AIO Impact on Google CTR: September 2025 Update. ⁶6sense, 2025 Buyer Experience Report. ⁷Princeton University, Generative Engine Optimization (GEO) study, 2024.