I've had some version of this conversation with almost every prospective client I've talked to this year.
Someone searches for a fund strategy once. For the next two weeks, an ad for that exact strategy follows them to CNBC, to their weather app, to a completely unrelated article about grilling techniques. They ask me how that's possible.
That part's easy to explain. It's retargeting, and it's been around for years. The part that's actually new, and the part most firms in this space haven't caught up to yet, is everything happening before that moment. The entire model financial services marketing was built on is quietly breaking, and it's breaking in a specific, traceable order.
Quick take: Search stopped sending people to a page of links and started answering the question directly. By the time someone's ready to talk to you, they've already done most of the deciding, usually without leaving a trace you can see. That means the old play, publish good content and wait to rank, isn't enough on its own anymore. The firms pulling ahead are reaching people during the research phase instead of waiting for it to end.
Here's that shift, broken into the order it actually happens.
For twenty years, the model was simple. Someone searches, Google returns ten blue links, they click one, and you win by being in the top three of that list. Nearly every dollar of SEO spend in this industry was built on that assumption.
That assumption is gone. Nearly 60% of Google searches now end without a click, because an AI Overview answers the question right there on the results page. Prospects are increasingly skipping the search bar entirely and asking ChatGPT, Gemini, or Claude to compare managers and build a shortlist directly. There's no page of links to rank on. There's one synthesized answer, and either your firm is the source it pulled from or it isn't.
That's Answer Engine Optimization, and it's not SEO's replacement so much as the thing that now sits next to it. The queries themselves don't even look like searches anymore. The average AI query runs 26 words long, built more like a real conversation than three keywords typed into a box. We've broken down exactly how those queries get built if you want the deeper technical version.
Here's the problem underneath that, though: almost nobody is actually measuring whether they show up in those answers. Most firms are optimizing for AEO blind, assuming they're either invisible or fine, with no actual data either way. An AEO audit fixes that. It runs your firm against a defined set of industry-relevant queries, across multiple AI platforms, ChatGPT, Perplexity, Google AI Overviews, Claude, and Gemini among them, and reports back how often you're mentioned, where you rank against named competitors, and whether the sentiment around those mentions is even positive. It's the AI-search equivalent of a keyword ranking report, and right now, almost nobody in financial services is running one.
Here's the uncomfortable part of all this. About 80% of the buying decision happens before your team ever gets on a call, and 78% of financial advisors research a fund or manager online before they'll engage with anyone from your firm. 6sense's 2025 Buyer Experience Report, based on more than 4,000 B2B buyers, found that buying groups have already ranked their preferred vendor before talking to sales, consuming an average of 13 pieces of content along the way, almost all of it anonymously.
By the time someone fills out a form, or your wholesaler gets a callback, the decision is largely made. Waiting for that moment to find you is waiting too long. Which is why the tools worth paying attention to right now don't wait. They go looking.
The most striking version of this is intent data. The underlying database covers more than 300 million households. You filter that down by demographic and firm-type criteria relevant to your business, then layer in real-time search behavior: who has actively searched specific, relevant keywords in the last two weeks. Not clicked an ad. Searched the term itself, on their own, unprompted. What comes back isn't a vague "this account is in-market" label. It's an actual person: name, email, phone number, and job title, tied directly to that search activity.
One honesty check belongs here: a December 2025 survey of 750 marketing leaders by DemandScience found that 87% say their intent signals are unreliable or inflated. A lot of what gets sold as "intent data" is noise. The version that's filtered by firm type, demographic criteria, and financial services-specific keywords, matched to a real, contactable person, is a different product entirely.
The same logic applies to your own website traffic. Most firms already see less of it than actually exists, because most visitors never fill out a form. Visitor identification tools match that anonymous traffic, at the company and individual level, against RIA and institutional databases, and hand you a list of who's worth a follow-up. If a compliance officer at a firm you've been chasing for six months just spent four minutes on your pricing page, that's worth knowing today, not discovering by accident next quarter.
And most firms already own a piece of this they aren't using. A platform like FINTRX, covering more than 40,000 firms and 750,000 contacts across broker-dealers, RIAs, and family offices, usually gets pulled once for a mailing list and then forgotten. The research is sitting there. The gap is almost never access to data. It's turning a static export into something that actually runs.
Once you know who's researching, the next problem is reaching them somewhere other than an inbox they're already ignoring.
Retargeting is the most familiar version. Someone reads your explainer on 1031 exchanges, leaves without filling out a form, and days later your ad shows up while they're reading something unrelated on Bloomberg. Keyword search retargeting goes a step further, catching people who searched a relevant term on a publisher site and never visited you at all. Keyword contextual targeting skips the person entirely and follows the content instead, placing your ad next to an article on the exact topic you cover, cookieless, so it works even in a privacy-first environment.
Geofencing is the one that gets the most attention, and it's really one mechanic with several applications. You draw a virtual boundary around a physical location, and any mobile device that enters it gets tagged for a campaign. Point that boundary at a verified home address and you're reaching a known prospect, like an RIA or an accredited investor, at close to 100% match accuracy. Point it at a conference venue and you're reaching every attendee without paying for a booth. Point it at a neighborhood or zip code known for a concentration of family offices and you've got a broader awareness play with no address list required. Point it at a specific office building, a prospect firm or a competitor's headquarters, and the mechanic doesn't change at all. It's the same infrastructure a car dealership uses to reach people who just left a competitor's lot. The one practical note: if you're naming a specific competitor, run the messaging past compliance before launch. The mechanic's fine. How you talk about it publicly is the only thing worth a second look.
And reaching people directly still matters, which is where cold outbound infrastructure comes in. The AI tools getting attention right now write and send emails at scale, some framed as tools that can basically act as a rep. The more durable version of this isn't about impersonation. It's a dedicated sending domain, separate from your primary one, with deliverability monitoring and gradual warming. A properly warmed domain can see inbox placement around 87% and reply rates near 4%. The same messages sent from a fresh, unwarmed domain land in the inbox only about 12% of the time. None of the sequencing or personalization matters if the infrastructure underneath it is broken.
Here's the part that cuts against the rest of this story. AI made content cheap to produce at scale, which means every inbox and every feed is now more crowded with generic, forgettable material than it's ever been. Getting someone's attention through the tactics above doesn't mean much if what they find next reads like everyone else's AI-assisted blog post.
Prospects are really asking the same five questions no matter how they found you: your team's background, your actual track record, your fee structure, your process, your product. Most firms answer one or two of those well and leave the rest to a salesperson to explain live, which only works once someone's already agreed to a call. The firms doing this well map content to all five, at every stage of the journey, so most of the questions are already answered by the time a real conversation starts.
The hard part is doing that consistently without it turning into more of the generic noise that's drowning everyone else out. That's the specific problem a platform like Market Voice AI is built to solve. Instead of a generic AI tool producing passable-but-forgettable content, it's trained on your firm's own existing thought leadership, expertise, and voice, with a real editorial calendar built around topics your actual audience cares about. The output still needs your team's real expertise behind it. It's a way to scale production of a genuine point of view, not a way to skip having one. None of this replaces a real website, real video, or real investor materials either. It just means the content filling the gaps between them doesn't have to sound like it came from nowhere.
Everything above gets someone's attention and starts to earn their trust. What happens in the next few minutes decides whether that turns into a conversation or quietly evaporates, because the window where someone's actually paying attention is shorter than it used to be.
A lot of firms build a lead scoring model in HubSpot once, watch it fail to reflect reality, and quietly stop trusting it. The fix isn't more automation, it's weighting the score around behavior that actually signals intent: three page views instead of one, a download instead of an open, a return visit instead of a single click. We've written more on building a lead scoring strategy that actually holds up.
The newest piece of this is an AI chatbot that sits on your site, trained on your company information and the questions prospects actually ask, running a real qualifying conversation before anything reaches a person. It routes the contact into your CRM automatically and assigns it to the right team member, so the first human touch is with someone who's already told you what they need, instead of a cold "just checking things out" form fill. For a team without the headcount to staff live chat, that closes a gap that used to just be a dead contact form. From there, sales enablement, sequences, playbooks, alerts, tells your team when a prospect is actually worth a call, built around the same signals as the scoring itself.
A few honest notes before any of this gets built out.
Retargeting, contextual targeting, AEO, and lead scoring aren't controversial. They're tools every other industry already uses without a second thought. Financial services has just been slower to adopt them.
Geofencing is genuinely flexible, and most applications, a household, a conference venue, a neighborhood radius, are no different from what retail and automotive marketing have run for years. The only application worth a quick compliance gut-check is naming a specific competitor publicly, and that's a messaging question, not a mechanical one.
Market Voice AI and the chatbot are new enough that the honest caveat is different. They're tools, not replacements for an actual point of view. An AI trained on your firm's expertise still needs your firm's expertise to train on. Garbage in, generic out, same as anything else.
The blue link assumed patience. It assumed someone would search, scan a page of results, and give your firm a fair shot at being read. None of what's actually happening right now assumes that anymore. The answer gets generated before anyone scrolls. The research happens before anyone fills out a form. The trust gets built or lost in content that's either genuinely yours or indistinguishable from everyone else's.
That's not a reason to panic. It's a reason to stop waiting. The firms adapting right now aren't doing more marketing than everyone else. They're reaching people earlier, in more specific ways, at the exact moments the rest of the industry still assumes don't exist.
If you want to see what a build-out of this actually looks like for your firm, our guide, Growing Assets with Digital Advertising, walks through how these channels fit together and what a realistic rollout looks like.
What's an AEO audit, and do we actually need one? It's a structured measurement of how often your firm gets mentioned or cited across AI platforms like ChatGPT, Perplexity, Claude, and Gemini, compared to named competitors, run across a defined set of industry queries. Most firms are optimizing for AEO blind right now. An audit replaces that guess with a prioritized list of what to fix.
Is any of this different from what other industries already do? No, and that's the point. Retailers and auto dealers have used retargeting and geofencing for years. Financial services is simply newer to applying these tools, mostly due to a compliance culture that treats unfamiliar tactics with more caution than they usually warrant.
Does something like Market Voice AI replace our content team? No. It's trained on your firm's existing expertise and voice specifically so the output doesn't sound generic, which means it still needs real input from your team to stay accurate and distinctive. It scales production. It doesn't supply the point of view.
Do we need a data provider like FINTRX or Discovery Data to do any of this? Not for all of it. Retargeting, contextual targeting, and AEO don't require a third-party data provider. Geofencing and intent data work best with one, since they rely on verified firm and contact data. If you already have a subscription, the opportunity is usually turning that static export into an active campaign.
Where should a firm start if none of this exists today? Start with website visitor identification and lead scoring, since both make your existing traffic and CRM more useful without any new campaigns. From there, layer in retargeting and intent data before moving into geofencing, AEO, and newer tools like Market Voice AI and an AI chatbot, which take more setup but compound the longer they run.