Business development, marketing and sales teams at asset management firms are asking us the same question. How can AI make our outreach to allocators more effective, and where should we start?
The first thing to note is that AI does not change what wins new business. Clarity of message, credibility earned through performance, and the relevance of your strategies to the allocator still win. What AI does is make it far easier for an allocator to find the manager most relevant to their mandate.
This week I watched a Volvo YouTube ad make a simple pitch. It is the car parents choose for a first-time driver, because nothing protects them better. That claim lands because Volvo has spent over 65 years earning it. Safety is consistent through all their messaging.
Safety is not a slogan Volvo added later, it is the founding story. In 1959, Volvo invented the modern three-point seatbelt and gave the patent away free, deciding that saving lives mattered more than owning the technology. It has backed that up with public, measurable commitments. So when I type a question into an AI engine, the answer comes back already decided.
Asked of Claude, August 2026
“What is the safest car to drive?”
“If you want one model that most consistently gets pointed to as the safest, it’s usually a Volvo.”
Volvo has consistently articulated its edge as safety, and so safety-conscious, affluent parents and professionals who value reassurance can type that prompt into AI and be directed to the brand.
AI is becoming the matchmaker between intent and reputation.
The same mechanic decides which asset managers get surfaced. A firm articulating its edge across its website, social channels, consultant databases, factsheets, media commentary and thought leadership gives models something stable to retrieve. As allocators increasingly begin their screening with AI rather than a consultant's shortlist, the firms with the clearest positioning are the ones that get retrieved.
Any BD person will recognise the established machinery of institutional discovery, because they work it every day. Consultants such as Mercer remain genuine gatekeepers. They influence searches, shortlists and recommendations. Databases such as eVestment and Preqin are the infrastructure of institutional discovery and due diligence, and being accurately represented there remains part of the basic hygiene of getting found and considered. Citywire and the broader fund-selector ecosystem play the same role in the wholesale channel. None of this is going away.
What is changing is that this ecosystem is no longer the only route by which an allocator discovers a manager. Increasingly, allocators research managers directly, through AI engines, as well as traditional search and social platforms. At the same time, managers can use first-party data and digital targeting to match the right thought leadership with the right investors, bringing allocators to them. And once an institution becomes seriously interested, all the established machinery re-enters the picture, from consultant relationships and database presence to track record analysis, due diligence, ratings and investment committees.
So the strategic shift is not replacement. AI creates a cleaner route to the investor, before and alongside the traditional process. The firms that win will run both routes deliberately, and make sure they tell the same story. The pace varies by segment, with wholesale and wealth channels moving fastest, institutional mandates more slowly and private markets in between, but the direction is the same.
Two routes to the same investment committee
Both routes are live, and both run on the same story. The direct route is the one most firms are not yet resourcing.
Forrester’s 2026 research finds that generative AI searches are now the starting point for professional buying decisions,[1] and roughly six in ten searches end without a click to any website.[2] Institutional manager selection is conservative, fiduciary-bound and consultant-mediated, and it will move later than the markets around it, but it will not be exempt. Allocators, consultants and advisers are beginning to ask ChatGPT, Claude and Perplexity to assess firms. When these models cannot articulate a firm’s edge convincingly, it risks being screened out of consideration, well before the consultant call or the database screen.
To avoid this, asset managers need to embrace generative engine optimisation, or GEO, the discipline of ensuring AI engines describe your firm accurately and cite your content. It is much more than SEO with a new name. A February 2026 analysis of three million ChatGPT responses found that 44 per cent of the citations the model makes come from the first third of a page.[3] And the engines are drifting away from the old search hierarchy. In mid-2025 around three-quarters of AI-generated answer citations came from pages ranking in Google’s top ten; by early 2026 that share had fallen below 40 per cent.[4] Start by asking the engines what they say about your firm. The answer will be instructive.
There is a feature of this market most firms have yet to notice. The engines have remarkably little to work with beyond what you publish. Much of the top-tier press sits behind paywalls the models cannot read, and there is little third-party commentary about individual managers on the open web. The models therefore build their summaries substantially from your own website, your own insights and your own social presence, and present the result in a voice that sounds independent.
Allocators can now ask an engine to compare five managers in a niche, a question Google could never answer, and the material you publish is what decides how that comparison reads. Handled well, this is a gift. Visibility in the AI layer is earned with content rather than advertising budget, which levels the ground between mid-sized specialists and the largest brands. But it will not stay this way indefinitely. Publishers are steadily licensing their journalism to the models, and as more firms practise GEO, the engines will learn to weight independent validation more heavily, just as Google learned to see through manipulated SEO. Firms that move now can shape how they are described. Firms that move later will need independent validation to do the same job. Either way, your own channels have never mattered more.
An AI engine compresses an asset manager’s entire positioning into a few sentences. If the messaging architecture is woolly, the AI summary will be generic, and a generic summary is a weak starting point for diligence. This is why the message spine comes before anything else. The opportunity you address, the edge you bring, how you deliver it, and why now. That architecture is no longer just the foundation of your pitch deck. It is the raw input that every human reader and every machine summariser consumes. Sharpening it is the highest-leverage spend in the entire marketing budget, because everything downstream inherits its clarity or its confusion.
AI floods the market with competent but voiceless commentary, and allocators can tell. What they pay for is judgement from named humans. More than nine in ten human-advised investors say they would not switch to a purely digital service. [5] The implication is not just to produce more content. It is to focus on quality over quantity and to make what you do put out reinforce your positioning, with deeper, more opinionated pieces from your investment team, and video providing the personal voice that introduces text-based analysis, data and insight. The discipline is one core message, expressed across every touchpoint the engines scan.
The targeting layer is not new technology, but it has never mattered more. You no longer need to communicate with an entire market. You can communicate with a list. Audiences of a couple of thousand named decision-makers, built on professional data or first-party contact information that LinkedIn can target. Content triggered by what an individual did, whether the page they visited, the paper they downloaded or the theme they keep returning to. CRM-driven follow-up based on the next best action, rather than on memory and good intentions.
None of it works unless marketing and distribution data live in one connected system, because the firms that win will not know more about their prospects; they will act on intent signals faster, and a signal that takes three weeks to reach the salesperson is no longer useful. To be clear about what is being personalised here, the content itself is made once, by humans. What the system personalises is who receives it, when, and what follows.
AI-referred traffic is still tiny, typically under 1 per cent of volume, which is exactly why most firms ignore it. But it is high-intent. In consumer commerce, Adobe found AI-referred visitors converted roughly a third better than traditional traffic,[6] because a visitor who arrives from an AI answer has already been qualified by the question they asked. The same logic applies to an allocator who arrives from an engine’s summary of your strategy. Judged by session counts alone, it could look negligible, which is how firms end up starving the one channel that is working.
Our answer to the question, then, is less about tools than discipline. A sharper story, told by real people, expressed across every touchpoint the machines (and humans) read, delivered with precision to the select allocators who matter. That is human narrative carried by machine distribution, and it only works when both halves tell the same story.
Firms that get this right now will be the ones allocators find first. Please get in touch to find out how Peregrine can help.
Sources
Forrester, The State of Business Buying, 2026 (January 2026).
SparkToro / Datos zero-click search analysis (Rand Fishkin), which found 58.5 per cent of US and 59.7 per cent of EU searches end without a click (2025).
Kevin Indig, analysis of three million ChatGPT responses and 30 million citations, reported in Search Engine Land, February 2026.
AI Overview citation tracking, 2025-26. The share of citations from pages ranking in Google’s top ten fell from 76 per cent (mid-2025) to 38 per cent (early 2026).
Vanguard, Quantifying the Investor’s View on the Value of Human and Robo Advice, which found over 90 per cent of human-advised investors would not switch to digital-only advice.
Adobe Digital Insights, holiday season 2025. AI-referred visitors converted 31 per cent better than non-AI traffic.