An insider’s look at navigating AI’s technical, organisational and regulatory barriers that hinder the digital revolution of financial marketing.
Executive Summary
The artificial intelligence bubble feels fit to burst. Digitalisation has been lightning-fast this side of the millennium, and AI represents one of its most seismic underground to mainstream hits. It’s not just a talking point: it’s a data processing behemoth with a predictive analytical mind beyond comprehension. So much so that an estimated $35 billion has been invested in fintech AI applications.
The returns of which, however, have been diminishing. For all its positives for personalisation, automated outreach and data segmentation, funds are struggling to see how it empowers their marketers to set their offerings apart. Investment is a people’s industry requiring customer psychology, storytelling and relationship building to excel – areas where AI capability is constantly playing catch up.
Given these costly moves the financial industry has taken and the technology’s growing sophistication, AI is not going anywhere.
In this comprehensive report, we’ll cross-examine why firms are not currently harnessing its forecasted ‘revolutionary’ use, citing data quality, lack of expertise and contextual understanding, availability, overreliance on automations, and ever-looming financial compliance, which are getting stricter with AI’s own legal hurdles.
Beyond the pitfalls that have claimed early adopters, we’ll also offer constructive ways to implement AI that do have proven marketing outcomes and leave room for the most powerful factor for decision-making: authentic, human creativity.
Introduction
The Unique Context of Fund Marketing
Under the wide umbrella of financial services, fund marketing is a different beast from simply serving products to consumers. Investors are highly sophisticated, and not interested in impulse purchases. Instead, they’re conducting exhaustive research into firms with established track records of returns and informative investment strategies.
Traditionally, marketing centred purely around performance as the truest indicator of the fund’s ability to deliver returns. But this attitude is changing. Consistent rapport is still imperative to driving investment decisions, as it was in the days when salespeople and investor relations teams built personal relationships with investors and consultants at conferences and coffee meetings. Now however, bolstering a fund’s expertise on deep subject matter to specific markets relies on digital content, as well as reaching end users via industry publications, PR, and speaking engagements.
Considered financial decisions are taken after an investors’ culmination of multiple digital touchpoints across these online and offline formats, and across months, maybe years, dipping into fund philosophy and timely opinion pieces as much as fund data.
That’s tough to adhere to. Yet still, the fund sector’s sub-segments also pose distinct marketing challenges. Mutual funds utilise intermediaries to reach their ideal investors, compounding the existing heavy gaze of compliance. Hedge funds target institutional investors and high net worth individuals (HNWI) with complex financial needs, and private equity and venture capital firms transact intensively with limited partners that commit substantial capital for long periods.
All this has made fundamental hurdles that have existed before digitalisation even more complicated:
● Long sales cycles: Institutional investors may perform their own due diligence before committing for as long as 18 months, which requires marketers to provide progressively advanced engagements for hefty periods.
● High-value, low-volume transactions: When even single allocations can amount to hundreds of millions, every valued prospect requires a trusted white glove service relationship that mass-market strategies (including AI, for that matter) cannot match.
● Decision-making units: Institutional allocations can require approval from a range of committees, C-level executives, consultants and compliance heads, meaning marketers have to jump through many hoops to address multiple stakeholders. Marketing that talks to financial advisors who control allocation decisions requires a far different approach than speaking to the concerns of an end investor.
● Regulatory oversight: Fund marketing operates under local guidelines and cross-border equivalents, too, such as the Securities and Exchange Commission (SEC), or the Financial Conduct Authority (FCA), which limits messaging flexibility due to detailed rules around performance disclosure, risk warnings and fair presentation.
Is AI Nebulous, Or A Necessity?
As with every industry, traditional marketing approaches have a shelf life – no less due to the pandemic’s accelerated impact on digital adoption, with no option for face-to-face meetings. Investors are alert to streams of financial content in multiple areas of the internet across many devices. Customer bases are just as diverse as financial technologies offering new routes into investments through robo-advisors, and passive investment products that place active managers under pressure.
As such, these audiences expect the best from their digital-first experiences. Standing out is a sticking point, be it through creating more intuitive websites or using channels that have escaped funds for a long while despite success in B2C marketing, such as video content and podcasts. Even the traditional communication models have been overturned by investor protection compliance, including Europe’s Markets in Financial Instruments Directive (MIFID) II and the General Data Protection Regulation (GDPR).
It was inevitable that AI would become the next frontier for fund marketing. After being admittedly slow to embrace digital, the sector has had to adapt once again to accommodate AI’s solutions in achieving hyper-personalisation at scale, identifying investor insights in real-time, automating manual workloads and demonstrating return-on-investment (ROI) through segmented data visualisations.
This rise of AI has been so rapid that a buying frenzy has ensued. Vendors will, of course, be singing AI’s praises as all-encompassing solutions to marketing problems. Whether anticipating investors’ digital behaviours, answering queries around the clock, or generating content to drive down costs, the promises for asset and fund managers have been vast.
As such, digital investments have taken many forms, from AI-driven marketing technology (MarTech) platforms to hiring data scientists. Years on, those initial expectations have not borne fruit. Executives report that their AI initiatives have failed to improve pre-existing outcomes around lead quality, conversions, and asset raising. Because of this, AI projects have stalled in pilot phases or been abandoned altogether – particularly if they have consumed significant resources.
Continued AI failures strike a chord that technology investments in the future must be more greatly informed than initial enthusiasm, especially in such a specialised field as fund marketing.
Fund-Specific AI Technologies at Play[1]
The boom in AI companies has created a varied field of applications useful to fund marketers, which have been implemented through several models, as follows.
Marketing Application Categories
● Predictive analytics: Machine learning models can be trained against vast historical datasets – from website and email interactions to demographic information and investment experience – to identify patterns where prospective investors’ interests lie, redemption opportunities with existing investors, or the strength of marketing messages in nurturing engagement.
● Natural Language Processing (NLP): When macroeconomic shifts or geopolitical events can affect funds, AI systems are able to identify investor sentiment through analysing text in industry research reports and media articles. This can be applied to set up website Q&A chatbots or monitor trends on social media.
● Dynamic content: Personalisation engines enable fund websites, email and other marketing materials to customise content based on individual investors’ preferences, ensuring the most correct messaging gets tailored to its intended end user.
● Automation platforms: AI underlines the orchestration of nurture campaigns through integrated omnichannel marketing. With predefined rules, lead scores can be calculated based on engagement and trigger further actions from marketers and sales teams.
● Generative AI: Applications can produce email subject lines, newsletter copy, social media posts, and even fund commentary from user prompts to reduce costs and time (before editorial oversight).
Implementation Methods
● Internal builds: Proprietary AI systems offer full customisation and control, but require established data science teams, infrastructure demands, and continuous maintenance usually reserved for larger resource-heavy managers.
● Purchased platforms: Out-of-the-box marketing technology platforms are great for firms needing fast, simple deployments with included features and functionality (but perhaps less specialisations that fund marketers require).
● Agency outsourcing: Marketing agencies may provide fruitful partnerships through AI-driven campaign management, which reduces implementation risk but perhaps less strategic match-up with C-level growth objectives.
● Hybrid models: Vendor technologies can be re-kitted and combined with internal development or in-house applications.
Fund Management’s Enduring AI Challenges
Current Technical Limitations
For all of AI’s use cases and the appetite for it, it’s the following stumbling blocks that can cause funds of all sizes to be reluctant to replace their traditional investor-facing methods.
Data Quality, Availability, and Constraints
AI, in its simplest terms, involves feeding a system with an input, and for it to produce a desired output that can drive human decision-making in a fraction of the time. These AI systems require a great deal of implementation and vast data volumes in order for their performance to be as accurate as possible. Greater data scarcity restricts pattern recognition and AI’s ability to predict consumer behaviour.
Why funds lack high-quality datasets to train models against in comparison to, say, huge tech brands and e-commerce companies can be explained by the more ‘niche’ customer groups they serve, the financial products that they sell, and the rate of their digitalisation:
● Firms may only onboard hundreds of thousands of new investors each year, and not require the processing volume of multiple customers who may transact millions daily. One investor might only attend one event in their life, or open the occasional monthly newsletter.
● Traditionally, fund marketing systems are disconnected for each branch of the stack – think event platforms, an email automation service, CRMs, website analytics or data visualisation tools. Fragmented data sources are tough to draw together for AI’s holistic data processing capabilities, especially with limitations from legacy technologies.
● When databases are filled with incomplete or outdated information (following company mergers and acquisitions, or professional job movies), AI recommendations will not match up to the real-world environment.
This brings into question the general limitations of trained models and why generated outcomes need to be taken with a pinch of salt. For instance, overfitted models can only perform well with historical investor data, and are not so applicable to prospects or the quick-shifting market conditions. Black box deep learning models are opaque and do not explain how they reach their final outcome, which can invalidate their decisions – especially in light of compliance. Concept drift means that models can see correlations in data (the number of page visits leading to investments, for instance), but not the direct causation. It’s therefore difficult to gauge which manual marketing interventions can then better influence outcomes.
In order to be developed effectively, AI models require clear business outcomes. Long-term asset retention is a different time-sensitive goal from one-time event sign-ups, making it difficult to optimise machine learning to provide value for each aspect of a fund marketers’ strategy. This only adds to the general confusion of what defines marketing success; often ambiguous and without hard metrics.
Lack of Nuance
Financial brands are still utilising NLP technologies to (as accurately as possible) adhere to investors’ very personal queries. This is the underlying AI technology for website chatbots, with the ability to generate image and text-based content or to gauge market sentiment.
The trouble with NLP lies with how diverse human linguistics are, and the context in which they are used. NLP cannot be fully reliable, and is often utilised for very single-metric specific jobs where knowledge of complex strategic and psychological factors cannot be replaced.
Generated text may lack logic, coherent structures or correct grammar. And when trained on only English-language data, NLP will poorly understand other languages or produce culturally inappropriate content. Investor psychology is already a tough nut to crack for human-penned content, when there’s cognitive bias and emotion behind investment decisions, and knowledge of a prospect’s past disappointments or celebrations will not be understood by AI.
In fact, AI-generated content will always favour the straightforward face-value of things, when subtlety is so necessary for delivering bad news, differing messaging for experienced investors to not sound condescending, or modestly addressing performance in commentary. It’s not just the financial world’s idiosyncratic terminology (“factor exposure”, “alpha generation”, etc.) that causes problems for NLP systems, but regional slang, intent, and irony may cause them to misinterpret a sarcastic comment as positive.
In these cases, writing from scratch is less burdensome than reviewing NLP generation, particularly when AI may omit disclosures that are integral to regulatory measures.
Matters of Personalisation
This ‘contextual blindness’ is a real hindrance considering the hyper-personalised marketing that AI is meant to enhance. It is great to be fed content that adheres to our individual tastes, but AI engines may feel one visit to a fund page for a specific allocation, or just to be generally aware of the market, is enough to overwhelmingly bombard investors with the wrong content across every channel. There’s even the case that investors might feel ‘under surveillance’ from dynamic content being so selective according to their search histories.
Conversely, poor AI personalisation can oversimplify outreach to some audience segments. Dynamic content can come across as very superficial if a name is lazily inserted into a template or banner addressing regionally popular fund content compared to individual preferences. This does not make messaging relevant to personal challenges or characteristics – and a real problem when prospects with little to no touchpoint history are only able to receive generic, assumed communications.
There are other worst-case scenarios. When hyper-personalisation uses inaccurate data, it can produce embarrassing errors targeting the wrong people with investments they don’t hold, reducing the fund’s credibility. So-called ‘filter bubbles’ are like a content echo chamber that disregards diverse strategies or products that might meet an investor’s needs.
Used incorrectly, AI’s personalisation method is a quick-win to boost metrics, rather than to genuinely understanding the all-important end investor inside and out.
Is AI Truly Predictive?
The macroeconomic environment boomerangs, and fund marketers have to be constantly aware of the non-stationary world in which they operate. Preferences for new investors could be very different from the datasets of established long-term relationships, and so predictive models may not be fed the most up-to-date information they need for accurate decision-making. Investors that have not interacted extensively will be misrepresented by ‘sample selection bias’, and long sales cycles may display outcomes occurring years after the initial predictions, making them impossible to validate.
When many relevant variables in fund marketing (such as competitive dynamics) are impossible to quantify, they’re not ideal for models’ predictive analytics. Nor are their abilities to account for the genuinely rare investment commitments that do happen.
There’s a necessary interplay between the predictive models and the end user that has to exist, and often the most accurate models (like deep neural networks) are so tough to interpret that it brings their initial use case into question completely.
The Regulatory Spectre
As if fund houses’ compliance headaches were not strong enough, AI operations add more components to the mixture that can indeed conflict with existing mandates:
● Compliance team reviews are integral before distribution, which are bypassed by real-time AI-generated content. Attempts to personalise offerings are therefore at odds with necessary pre-approval tasks.
● Being able to customise presentations for fund performance (including benchmarking and standardised risk metrics) can only be completed in line with specified formats, leaving little wriggle room for AI differences that could violate terms.
● Inappropriate investment products must not be exposed to unsuitable investors. AI pattern-matching could increase the danger of this happening.
● Risks and benefits must be balanced according to financial regulation, where NLP capabilities tend to lean toward positivity to craft persuasive marketing copy.
Taking a look at the wider picture, international fund marketing is under the duress of many jurisdictions and their local regulators. As with NLP’s contextual troubles, AI systems can struggle to balance rules according to their governed region.
A similar concern extends to the collection of investor data, where marketers’ collection, storage and usage rules are dictated by area; prominent examples include GDPR in the EU, or the California Consumer Privacy Act (CCPA). Personalised AI communications that target multi-version messaging to various investors at different times make it extremely difficult to log and find audited data for supervisory checks.
A Compatibility Issue
Financial technology infrastructures come under the radar for being ‘behind the times’. Mostly, this refers to a reliance on legacy systems ill-equipped for changing regulatory measures and digitisation, as opposed to flexible cloud-based architectures.
For marketers, investor information may be split in various data siloes from CRMs to web analytics platforms, or through third-party services. Being able to unify digital touchpoints involves deep integrations – AI software integrations included – that are likely difficult or non-existent for outdated technologies to accomplish.
To avoid this, funds prioritise the improvement of existing risk management or compliance platforms, or sticking with proven vendors. AI projects are essentially null and void if base-level marketing technology built over decades is incompatible, where C-level budget allocators will put AI ventures onto the back burner.
Over-Reliance on Automation
It’s become easier to see why decision-making feels inherently a human trope with AI’s technical limitations. Even if data quality and quantity can account for successful AI model training, it removes the intuition of experienced marketers in various areas.
For example, chatbots’ stilted and formulaic conversation all comes through predetermined logic that is unable to deviate from a standard script. It’s true too that their predefined tasks will ignore any lucky, if unexpected, opportunities for an upsell that a human would pick up on and chase.
AI automations are time-savers behind the scenes at times, yet utilising them in areas where marketers will display more creativity, care and attention, shared learning, and accountability is a slippery slope that removes the personal flow of a trusted fund-to-investor relationship.
The Economics of AI Implementation
Getting every stakeholder at a fund invested in AI for marketing purposes is mired with troubles – no less from the technology’s long shelf-life, incurring huge fees that may be out of reach.
Resource Juggling
The sheer number of moving parts involved with AI implementation identifies where certain personnel or platform limitations become clear, including:
● A strong data infrastructure foundation, with cloud computing capabilities and data storage.
● Multi-channel automation platforms, add-ons and APIs.
● Specialist vendor services, including platform providers, support teams, and implementation consultants.
● Technical hires, expertise, and employee training.
Full-on attention in each area is integral for an AI strategy to be successfully launched and maintained for the runtime of future go-to market campaigns. The product’s lifecycle management can be hard to handle, and especially offputting when the opportunities offered by AI can be outweighed by investing in proven marketing tactics.
Total Cost of Ownership
It’s tricky to justify CFO buy-in when onboarding AI vendors, migrating data, system downtime, staff increases and training, and ongoing troubleshooting can be extensive and costly. At mid-sized firms, tech costs for AI platforms can already exceed $500,000 a year. This is exceedingly more for enterprise-scale deployments.
When projects extend up to 18 months, originally undervalued costs bloom to the point that they become unsustainable. And even when AI production environments become a reality, scalability issues involving greater customer onboarding and data for enterprise-levels become growing investments. This is where doubts begin about whether to carry on or abandon the AI jump, losing all that valuable time and money pumped into making it work.
Return on Investment
ROI is perhaps the ultimate sign of marketing attribution. When analysing the results of planned AI solutions, sometimes the desired business outcomes do not materialise as soon as the C-suite would like, questioning the legitimacy of how AI can fundamentally assist in asset-raising. Completed AI projects may not reap rewards for up to 5 years, and as with any new venture, hidden costs crop up all along the way to tarnish the new technology’s ROI picture.
If resources are too narrowed toward AI, it may also negate superior returns that other strategies may meet. This includes hiring personnel with strong industry relationships, conducting consultant and agency work, generating high-quality content with investment writers, video and social media experts, sales enablement tools, hosting smaller online events, or funding spots at famed industry conferences. There’s more human effort involved here, but it could pay dividends more immediately than stretched AI implementations.
Human Reservations
This again brings up the fact that human qualities often prevail over the most exciting digital advancements. Beyond AI, ‘tech fatigue’ is a very real thing now that devices are so intrinsic to our daily lives, and connections are as valuable (if not more) than they’ve ever been.
It’s also worth noting that AI has created significant cultural and organisational shifts which cannot be worked out overnight, and are perhaps widening as AI’s pros and cons are met with differing opinions inside and outside of the office:
For Funds
Adapting to the AI boom has seen a whole raft of operational changes. There are new platforms to get to grips with, altered workflows, roles and responsibilities, and new talent in the form of data science experts. Technical staff can do wonders for upping the fund’s handling of statistical modelling and data engineering, but can complicate life for a fund marketer. Their priorities are divided between building sophisticated, transformative systems and taking decisive action to raise AuM from investors, respectively.
No matter how comprehensive training programmes and shared expertise are, each department will speak a different language when it comes to AI. This can cause communication breakdowns, AI scepticism, and disruption to business-as-usual tasks. AI automation patterns can conflict with investor-first thinking or even diminish marketers’ decision-making control, especially if its output feels in direct opposition to their gut feeling and experience.
Similarly, the sheer number of AI-gathered touchpoints and interventions makes it tough to measure exactly who or what influenced final conversions or investment decisions, and the ‘metric gaming’ illusion of success (improved metrics over business outcomes) can skew results at quarterly meetings.
Seemingly AI can muddy marketing incentives when people and platform do not work together. The ‘job replacement’ discourse does abound, albeit a lot of fund marketers understand the co-dependent relationship with AI which, when fostered properly, does not take value away from the marketers themselves. A coherent dynamic relies on quality and healthy data, and defining marketing performance metrics separate from pure AI usage.
For Investors
AI is a great facilitator for increasing interaction rates and maintaining efficiency, but funds must never forget that investors are not algorithms. If too much outreach is masterminded and run by machine learning, it may not instil confidence in investors that genuine, trustworthy people are managing their assets – some of which can be very large sums indeed!
To heed their personal short- and long-term investment goals or assist them during market downturns, marketers have to demonstrate competence, emotional intelligence, and take accountability for the good times and the bad. This involves holding actual in-person discussions (as in the old days), and being responsive, timely and knowledgeable whenever investors need personal matters met. After all, everyone gets frustrated when a generic FAQ page or chatbot response beats around the bush and you cannot find a phone number to hand.
On top of this, the investor demographic is wider these days, and so are the varying comfort levels with conducting buyer journeys online. For instance, younger institutional investors may be from highly technical backgrounds, adept with a range of applications and scrupulous in identifying authenticity in generated content or other AI use cases.
Funds also have to take into account that some investors are genuinely anti-AI. HNWIs may be particularly worried about confidentiality around their data usage, and experienced investors can likely spot the quality difference in a firm’s content if too much is entrusted to AI. Trust erosion may exist, forcing investors to seek out funds offering more personal services.
Given how quickly digitalisation has gripped the financial services industry, there are high expectations for how technology enhances marketing efforts rather than replaces it. This is the crux that funds must understand in juggling AI with their audience’s preferences.
Case Studies: Where AI Can Pose Fund Marketing Problems[2]
Given the number of AI projects that could be undertaken, there are just as many hypothetical risks that can be encountered. Here are some potential pitfalls to account for when devising strategies to best utilise the technology.
Case 1: The Chatbot
The Scenario
A financial website offers an AI-powered chatbot to handle instant responses to global investor queries around the clock, without huge costs on call centres.
The Weaknesses
➔ Fund performance information given can be inadequately provided without proper disclosures, suitability analysis, and predictive statements against securities regulations.
➔ Misunderstood queries lead to incorrect strategy, fees, and holdings information, leading to legal complaints from investors who acted on the AI’s answers.
➔ Frequent “I don’t understand” answers and referrals to FAQ pages (and not humans) cause annoyance and a breakdown in the investor’s enthusiasm to engage.
Case 2: Over-Personalisation
The Scenario
A fund has customised dynamic email content using AI-driven segmentations – based on the recipient’s digital history, behaviours, and document downloads – to gain higher engagement metrics.
The Weaknesses
➔ The AI system is driven by open and click-through rates rather than investor interest, with sensationalised messaging outside of a brand’s tone-of-voice, as well as triggering multiple campaigns simultaneously, leading to upped unsubscribes.
➔ When prospects are misclassified, such as retail investors receiving institutional content, it marks an obvious lack of attention and honesty.
➔ Hyper-personalisation may feel encroaching or invasive when an email could reference highly specific fund pages or reports a user has looked into recently.
Case 3: Unedited GenAI Content
The Scenario
A small asset manager on a tight budget has utilised generative AI to produce market commentaries, based on the system’s analysis of past newsletter writing, news articles and industry data.
The Weaknesses
➔ The messaging may have been technically accurate and pertain to SEO, but generic and missing both the fund’s style guidelines and the distinct writing of portfolio managers that could offer fresh perspectives.
➔ Factual or mathematical mistakes could be made: either misidentified market trends or miscalculated returns, respectively, wrecking the manager’s credibility.
➔ The efficiency of generating content in seconds gets wasted from compliance officer reviews that require careful scrutiny of the AI’s judgment and output, particularly for risk warnings and fairness.
Case 4: Predictive Pattern Problems
The Scenario
A fund with a budget to play with looks to experiment with predictive analytics, analysing hundreds of variables from historical email enhancements to past investment patterns and letting the system self-learn from its findings for two years.
The Weaknesses
➔ Changing market conditions can deem a lot of the data the model was trained on insufficient. Volatility will increase, and audience expectations will change, making the system unable to adapt to relevancy quickly enough.
➔ Due to bias and model inaccuracies, ‘highly likely’ prospects may never invest, and marketers miss those labelled ‘low priority’ who could be essential AuM opportunities. AI’s false confidence can cause strategic errors.
➔ Investors’ personal relationships and investment guidance are all external factors limited to the models’ information they need for nuanced prediction.
Case 5: Promised Automation Pros
The Scenario
A firm has decided to implement an all-expansive automation tool comprising AI-powered campaign orchestration and lead scoring.
The Weaknesses
➔ Rigid workflows diminish manual intervention when needed, and could fail to integrate with the fund’s existing CRM and investor data systems, requiring manual uploads, exports, and costly deployments.
➔ Despite a range of configurations, this can lower the ability to create effective campaigns across outreach teams through tricky user controls, authentications and option paralysis for features, driving users to revert to familiar tools.
➔ Ongoing fixes for technical issues drive up previously calculated ROI, while persistent, poor support services from vendors can render platforms as ineffective money pits.
Going Hybrid: How Traditional and AI Marketing Can Co-Exist
For the many limitations of AI we’ve dissected, there are, of course, the human-prone marketing slips that have always existed. In today’s market, there’ll always be room for both to exist side-by-side for very different tasks, and identifying where people and platforms can spur AuM success comes down to comparing approaches across key performance metrics:
Lead Quality
The old school ways of gaining leads may feel long-winded, but the added effort could result in those of higher quality. If they’re gained from referrals through friends and colleagues who are existing investors, consultant introductions, or after a few minutes of genuine investment interest at forums, that already meets the minimum requirements for a follow-up. This level of personal connection is missing from digital lead generation, albeit that method maximises the potential to reach far more prospects for far less outgoings, across every available online channel where investors exist.
Conversion Rates
Research has estimated that conversion rates from traditional marketing exceed AI-driven processes by up to 60%. This may be due to the powers of persuasion through an in-person presentation or relationships built over a hefty period of time. It does not discount the effect AI can have on automated round-the-clock campaigns that continually contribute to the nurturing process from brand awareness to conversion, however.
Asset Retention
Intensive long-term periods of personal engagement will likely end with longer holder periods than those acquired by digital means, particularly through factors such as market instability, where trustworthy relationships will be valued the most.
Cost Efficiency
Outgoings are the CFO’s concern for marketing attributions, and both traditional and AI approaches can incur costs. While tech investments, connecting systems, and overall stack management is an ongoing expense, large-scale event sponsorships and traditional print advertising can also be one-off payments that make a dent. It’s important to identify the ineffective parts of both manual and automated marketing to double down on where conversions and ROI are boosted most.
Completely AI-free strategies do have their champions, dependent on the fund’s size and marketing scope. Boutiques may have successfully competed with bigger firms with less budget, focusing on their smaller pools of investors but juicing existing marketing techniques over and over again. Certain portfolio managers may be so esteemed in the industry that their words (written or videoed) will always hold an audience and do a lot of brand awareness and sales funnel legwork.
That does not mean, though, that AI cannot simply elevate these trusty methods in specific situations. The hybrid format is being leveraged by many funds to tactically reduce any pointless workloads and power up the ability to reach any investor, anywhere, with guided content. So long as AI-driven campaigns are iterated when they succeed, marketers can oversee a holistic strategy and focus on the creative aspects they favour, as follows:
● Identifying patterns from huge datasets is even more granular than finding a needle in a haystack, with AI surfacing investor topics of note (and where they’re popular) to fuel the next brilliant human-led nurture sequence.
● AI ‘suggestions’ are exactly that: outputs that are not gospel, but can point out missed opportunities, such as optimal contacting times or reactivated cooled leads for marketers to choose to act on.
● Routine tasks can be left to machines and free marketers’ time, including database entry, summarising documents, initial report generation, transcription services, and briefing meetings using recent investor activity and research cues.
● Basic tools for checking spelling, sentiment, and potential compliance gaps are available for articles and email correspondence, such as newsletters or commentaries.
Every part of the marketing machine has positives and limitations. Being solely one over the other could spell trouble, while a healthy mixture can be refined over time to achieve greater results.
Where AI’s Future Affects Fund Marketing
Regulatory Changes
The manner with which AI processes sensitive data, and can even manipulate a range of inputs into a different outcome, means that AI-led marketing systems are now coming under closer inspection from market regulators and newfound legislation.
One of the main scepticisms around AI was that it existed (and still does, to some extent) in a relative regulatory grey area. Rectifying this has had to be swift to protect the security of personal information, and hold financial institutions accountable for how they intend to use AI models – becoming more commonplace for such advantageous solutions as anti-financial crime controls.
Particularly for large institutions that may conduct business across the globe, systems have to be compliant with regional AI oversight. Some areas have heightened their national or multinational AI governance, while others remain lax. Prominent examples include:
● The SEC’s AI guidance relates to core principles as part of its SEC Marketing Rule, which emphasises firms’ responsibility for ensuring fair, representative and transparent communications (particularly for performance results) whether generated by AI or not. Technology cannot be made responsible for compliance errors.
● This is also the case under the Financial Industry Regulatory Authority (FINRA) for brokerage firms in the United States, where AI applications for communications (such as chatbots) must be supervised in line with applicable FINRA reviews, enforcing recordkeeping and documenting procedures for any correspondence that involves a firm’s investment banking or securities business.
● The European Union has aimed to unify its AI legal framework across all 27 member states through the AI Act. Financial applications, in a potentially high-risk industry, have come under close scrutiny in promoting the interpretability of algorithmic outcomes, boosting trust through transparent use, and implementing “explainable AI” (XAI) over opaque black box algorithms.
This therefore creates some novel challenges to mitigate any AI risk associated with regulatory red tape. Model risk management involves validating how robust AI systems are in determining their outcomes, where independent regulatory audits could prohibit any firms that are unable to disclose their data collection and AI usage practices to inform marketing decisions. Such recordkeeping becomes all the more unfeasible with multi-investor personalised content to track.
Being able to test that AI-led marketing platforms are functioning as intended (and therefore accurately, fairly and compliantly) is also a consistent monitoring and governance process that may be difficult to implement at resource-strapped firms. This only gets more complex when utilising third-party vendors who must be vetted for maintaining context-specific compliance protocols, as is ensuring model outcomes remain accurate across multiple input scenarios.
Legal liabilities are due to change, for both human errors and AI misuse. Compliance raises the risk level of AI-based marketing practice, when layer upon layer of financial regulation can already feel more of a pullback from AuM growth than a path toward it.
AI’s Realistic Evolution
Similarly worrying for fund marketers is that we feel on the bleeding edge of where AI is going. The revolutionary call may resound louder when AI capabilities continue to grow exponentially in a short time.
While it’s still speculation, improved NLP may blur the lines between what machine learning and humans can do in contextualising news sources, or finding ambiguity in language. Fund marketing inefficiencies may be all but removed by super-advanced computing power that gets closer to mimicking humans’ ability to create enriched personalisation experiences, all while regulators remain a little way behind in prosecuting businesses that have fallen foul of its existing limitations.
What funds can focus on are the smaller, campaign-ready AI applications that are either highly intuitive or even underlying ‘helpers’ in an existing CRM, email automation tool or other marketing platform. There’s greater AI implementation support than ever before offered by partnerships or consultancies, helping to accrue value for fund marketers and their investors in multiple automated ways:
Databases or business analytics platforms can quickly visualise immediately valuable information around digital behaviours and fund marketing performance to inform ways to tweak campaigns to align with weekly, monthly, or annual KPIs.
Content variations (such as segmented emailing steps based on investor interest) can be tested for success using AI-based A/B testing, across variables such as subject line messaging and appropriate send times per region.
Website interactions can be handled by chatbot triage rather than fully automated conversations, answering simple Q&A and handling complex inquiries to investor relations teams to manage expectations while providing genuine assistance.
Selective tasks allow fund marketers to fully optimise certain applications, and not feel too pressured from so many complex AI ethics and compliance concerns that can result in reputational flaws. AI adoption can remain a learning experience that heeds genuine ROI, and confidence in the people-platform joint model that looks to enhance fund marketing’s productivity.
Successful funds will identify AI’s direct influence on business outcomes and abandon any experiments that do not deliver in the short term. Human fund marketers will always be in control, with their asset-raising commitments augmented by technology, which the industry is already envisioning as a major part of their immediate future.
Our Recommended Next Steps
Implementing AI is a strategic choice that impacts all areas of a fund from the top C-level executive suite downwards. Given the breadth of AI-based solutions out there in the market and the various wins and risks they each provide, it can be confusing to map out a plan that makes the technology the operational helper it aims to be.
With that in mind, and the AI discourse building, here are some suggestions for those who are starting out, or early on, in their AI adoption phase.
For Fund Marketers
Start Small
While there may be a lot of noise, hype and hollow promises from AI vendors and tech fans, that will not amount to better engagement. By ruling out limitations that could be too risky, focus on AI being a single selective tool rather than an all-encompassing AuM saviour. Small-scale applications should be chosen to address any shortcomings, such as productivity or automated outreach, to then scale when they live up to a marketer’s expectations.
Keep Data Clean and Compliant
AI’s ability will be hampered by incomplete and outdated data. Before pursuing projects, make sure housed data (in the CRM and additional marketing platforms) is free from error and accessible to ensure smooth integrations. This is a massive factor in keeping investor communications compliant, too, where regulations should be heeded over efficiency gain or ‘a shortcut to innovation’.
Preserve the Fund’s Internal Value
Marketing technology is ultimately there to reinforce brand identity globally, with AI applications useful assistants for personalising email templates, sentiment checking, etc. When onboarding third-party vendors, shared values and market positioning must be maintained. This will also be crucial to build a marketing team’s understanding and enthusiasm to implement AI into existing workstreams and create a ‘human in the loop’ approach.
For Industry Leaders
Engage with Regulators
AI’s ethical minefield is a hotly debated thing, yet quick developments in the regulatory space are remedying how financial institutions can better understand AI usage. Proactively seeking advice from national watchdogs can shape AI oversight that does not hinder existing workflows.
Establish and Promote Best Practices
It’s recommended to develop organic AI best practice guidelines to keep teams on track as innovations continue. Key considerations around investor interests and sensitive data protection should be given within an AI framework and outside it, as well as methods to bolster the integrity of the ecosystem, such as where and how to make marketing communications transparent.
Create a Compliance Culture
Technological evolution starts with a conversation. Investing in educational sessions from AI experts builds literacy and even interest across all areas of the business, whether that’s in lunchtime talks or attending forums to collaboratively discuss productive AI use cases specific to fund managers. With more keen practitioners, a fund can make better tech-buying decisions.
Vendor Evaluation Framework[3]
With the chasm opened for AI marketing tools, the number of vendors out there has ballooned. This can lead to option paralysis and fuzzy thinking that do not align with fund marketing and sales’ asset-raising deliverables, which should remain the North Star when performing due diligence on any external AI services.
The first step is to consider how any new MarTech tool will contribute to performance. This comes from researching vendors with clear ROI-driven goals in mind, such as:
● Will the AI integration cost more than consolidating existing platforms?
● Can it reduce our manual marketing inefficiencies by x%?
● Will it help increase our investor subscription base by x%?
● Can we achieve x% number of new content per month?
These are holistic considerations maintaining eyes on the budget, which can then be allocated to specific AI functions for base fund marketing jobs:
Short-Term AI Wins
● Content: Ideas generation, structural tools, grammar and sentiment checks.
● People-platform hybrid: human-in-the-loop feedback to review and validate AI interpretations.
Then, benchmarking between vendors should take into account the following criteria to account for a considered lifecycle management plan from the start:.
Technical Capabilities
● Is the tool AI-native? These are core AI products that are sophisticated, easily updatable and scalable with no-code options.
● Is the tool AI-wrapped? These can complement existing software for personalisation features and abilities to streamline productivity or increase content outreach.
● Is it specialised for fund marketing or generalised? AI tools and support staff for financial marketing platforms will understand unique regulatory or communication style requirements rather than blanket all-inclusive solutions.
● Can it create audit logs to determine performance increases or decreases?
● Does it account for XAI or transparent AI model management?
Implementations
● Can it integrate with existing systems or ingest CRM data easily?
● Do implementation teams understand the knowledge and needs of financial services as well as complex technology sectors?
● Is the AI tool flexible for marketer-led human control and oversight?
● Does it have the capacity to accommodate increased data volumes?
Security and Regulation
● Are vendors compliant with protocols such as ISO27001?
● Are their data privacy and input data ownership rules clearly stated?
● Can they meet financial obligations from initial deployment, rather than reactive firefighting?
Cost Considerations
● What are the licensing fees, if any?
● Are there any hidden add-ons, e.g. ongoing support and staff training?
Considering A Strategic Compromise
There’s evidently been some stark misalignments between AI and fund marketing. Namely, the increasing scope of financial and AI regulations, bespoke investor communication requirements, realistic human and material resources, and the maturity of existing marketing platforms to accommodate automated organisational tools.
Considering the role of fund managers being so steeped in long, trusted investor relationships, the very idea of quick-win automation capabilities was already on the back foot, and only compounded by the complexity AI brings to funds. A lot has been asked of marketers, salespeople and IT services to learn complicated theoretical and practical AI uses quickly, and adapt to avoid potentially falling behind in the ‘tech race’.
However, there has not been a complete rejection of the technology. And there cannot be, considering the ways that human-led AI strategies can determine more effective communications with today’s digital investors. Each side should augment the other, with laborious manual tasks quashed, investor journeys tracked in realtime, and the marketing and sales process better informed by AI-analysed data connected in a MarTech stack.
Shared insight, training programmes and hands-on experimentation are all underway to get multiple sized funds clued-up with AI. That can only be good news for their transformation no matter how far their AI projects are taken. Understanding its current limitations and capabilities for raising AuM can prevent it from failing fund marketers, and should it become second-nature to the asset-raising process through time, steer funds confidently toward the next digital frontier.
References and Further Reading
Industry Reports and Research
● Cerulli Associates: The Future of Asset Management Distribution
● Deloitte: 2026 Investment Management Outlook
● McKinsey & Company: The Future of Wealth Management
