Gartner projected back in 2024 that traditional search volume would drop 25 percent by 2026 as users shift to AI assistants, and the shift hit financial queries harder than most categories. Money questions are exactly the kind people prefer to ask a private chatbot instead of a public search box. Nobody wants their browser history to read “do I have enough to retire,” but they will type that sentence into ChatGPT at midnight without hesitation.

That behavioral quirk is the entire case for AEO financial services work in 2026. The questions that used to start at Google, where your firm could buy its way onto the page, now start inside an answer engine that cites three sources and ignores everyone else. If your firm is not one of the three, you were never in the conversation. Here is the playbook, built to survive the one constraint every other AEO guide ignores: your compliance department.

Run the diagnostic before you touch anything

Before optimizing, measure. We ran a live test for this article. On June 10, 2026, we asked Claude: “What should I look for when choosing a fee-only financial advisor?” The answer named fee structure transparency, fiduciary status, CFP credentials, and Form ADV review, and it pointed toward verification resources like NAPFA and the CFP Board rather than naming individual firms. Ask the same engine about advisors in a specific city, though, and it starts naming firms, pulling from directories, review profiles, and local coverage.

Advisors reviewing charts with a magnifying glass during a client portfolio audit

That two-part result tells you where the opportunity sits. Generic education queries cite institutions and credentialing bodies. Local and niche queries cite firms. Your move is to dominate the second category, because that is where the engines still need help and where a buying decision is actually forming. Run ten queries your ideal client would ask, note who gets cited, and you have your competitive map in an afternoon.

Build the query set from three sources. First, your own intake notes: the questions prospects ask in discovery calls are the questions they asked a chatbot the week before. Second, the objection layer: what stops people from hiring you (“are advisor fees worth it for a $400k portfolio”) is asked privately and almost never typed into a public search box. Third, the life-event triggers that start advisory relationships: inheritance, equity windfall, divorce, business sale, the first child. Phrase each query the way a nervous non-expert would, including the wrong vocabulary they would use, because engines answer the question as asked and the firm whose content matches the asking wins the citation.

Repeat the ten-query test monthly and log who gets cited, not just whether you do. Citation patterns shift visibly after engine updates, and watching which competitor gains ground tells you what the engines started rewarding before any industry blog writes it up.

Pick the niches where engines still lack a confident answer

“Financial advisor” is a lost query. “Financial advisor for equity compensation at pre-IPO startups” is winnable. Answer engines hedge when their training data is thin, and thin data is exactly what specialty niches produce. Advisors serving dentists, federal employees, expats, crypto holders with tax problems, or recent widows all sit in low-competition answer space where one well-structured site can become the citation of record.

The pattern holds across the industry’s segments. A regional credit union cannot outrank Chase for “best checking account,” but it can own “best business checking for Washington wineries” forever. Niche depth beats brand breadth in every answer engine we test.

The economics of this choice are better than they look. A niche query produces a fraction of the volume of a head term, but the person asking it converts at multiples of the generic searcher, because specificity in the question signals readiness in the asker. Someone asking “fee-only advisor experienced with federal TSP rollovers” has already decided to hire; they are choosing whom. Engines also defend niche citations longer. A head-term answer gets recomputed against thousands of competing sources every cycle, while a niche answer with two credible sources stays stable for months, which means the page you build this quarter keeps answering into next year.

Picking the niche is a positioning decision, not a content decision, so make it once and commit. The test: can you name the professional association, the trade publication, and the three online communities where your niche congregates? If yes, you can build the citation footprint. If no, the niche is a keyword, not a market.

Build the question architecture compliance can live with

The core asset of AEO financial services work is a library of pages that each answer one real client question with the directness of a good advisor in a first meeting. “How much does a fee-only advisor cost,” “what happens to my 401k when I change jobs,” “Roth conversion before retirement, who should not do it.” One question per page, the answer in the first 80 words, the nuance below.

Hand scrolling market data on a tablet beside a laptop, researching a financial question

Here is what makes finance different from every other AEO vertical: each page needs to clear FINRA 2210 or SEC marketing rule review, depending on your registration. The workaround is structural. Write educational answers, not performance claims. “How target-date funds rebalance” needs no disclaimer gymnastics. “Why our fund beats target-date funds” dies in review and should. The educational frame is not a limitation, it is the format answer engines prefer anyway, because they are also trying not to give individualized financial advice.

Structure each page the same way a good answer reads. The question as the H1, phrased as the client asks it. The direct answer in the first 80 words, numbers included where numbers exist. Then the conditions under which the answer changes, because money answers are conditional and engines reward content that handles the conditions instead of pretending one answer fits everyone. Then the moment when general information stops being enough, which is your natural, compliant bridge to the consultation offer. FAQ schema on the page, the advisor who reviewed it named with credentials, a visible last-reviewed date. That last element matters more in finance than anywhere: rates, limits, and thresholds change every year, and a page citing the 2024 contribution limit in 2026 is training every engine that reads it to distrust your domain.

Cadence beats heroics. One page per week, every week, built from the query map, will put forty conditional answers into the record in a year. That is more retrievable, citable financial answering than most regional firms have published in their existence, and it is the asset that makes every other move in this playbook work.

Make the credentials machine-readable

Answer engines weigh trust signals more heavily in finance than in almost any other category, since their builders know wrong money answers cause real harm. Feed that hunger in structured form. FinancialService and Person schema on advisor pages. CFP, CFA, and AIF credentials written in text, not trapped in badge images. Form ADV and CRS linked from the footer. Bios that state fiduciary status in a quotable sentence. NAPFA, XYPN, and CFP Board directory profiles kept current, because engines cross-reference directories to verify a firm exists and is what it claims.

The cross-referencing deserves emphasis because it is where most firms silently fail. An engine asked to recommend an advisor checks whether the firm’s site, its SEC or state registration, its directory profiles, and its LinkedIn all describe the same entity doing the same thing in the same place. Every mismatch, the old office address on one directory, the advisor who left in 2024 still listed on another, the AUM figure that disagrees with the ADV, lowers the engine’s confidence, and low confidence means the engine recommends the competitor whose record is boring and consistent. Run the consistency check once, fix everything, and put a recurring hour on the calendar each quarter to keep it fixed. It is the least glamorous work in this playbook and the highest floor-raising value per hour spent.

One asset class of trust signal is underused across the entire industry: the advisor’s own publication trail. An advisor quoted twice in trade press, with one contributed column and a podcast appearance, carries a verification halo that no amount of on-site content produces, because the engines treat independent domains vouching for a person as evidence of legitimacy. Five hours per quarter of earned-media effort buys what fifty hours of on-site writing cannot.

Get cited where the engines already look

Your own site is half the work. The other half is presence in the sources engines trust for finance: trade press like InvestmentNews and Financial Planning, local business journals, and the personal finance outlets that AI answers quote constantly. One quote in a story about 529 plan changes does more for your citation footprint than ten social posts, and reactive source platforms hand those opportunities to whoever answers fastest. This is the same earned-media flywheel that powers AEO financial services results for our clients in every other trust-heavy category, from law to healthcare.

The cadence that works for a small firm: one reactive quote per month minimum, one contributed article per quarter in a trade or local outlet, and one data-driven story per year built from something only you can see, even something as modest as “what 150 client conversations revealed about Social Security timing anxiety.” Compliance reviews earned media more gently than advertising in most shops, since you are being quoted as an expert rather than making solicitations, but route everything through your review process anyway and keep the approvals on file. The placements compound in two directions at once: each one is a trust signal to engines deciding whether to cite your domain, and each one is a page a cautious prospect finds when they search your name before signing the advisory agreement.

Audit the brand answer monthly

Ask each major engine what it knows about your firm, by name, every month. Wrong custodian, stale AUM figures, a departed partner still listed as lead advisor, or a confusion with a similarly named firm in another state: all common, all fixable, and all the kind of error that quietly kills referrals when a prospect checks you out. Corrections flow from updating the sources engines read, your site, your directory profiles, and your coverage, not from arguing with the chatbot.

Assign the work or it will not happen

Every firm we have watched succeed at this assigned AEO to one named owner with a monthly cadence: ten tracked queries, one new question page per week, one earned citation per month, one brand audit. Every firm that treated it as a someday project is still invisible, and the engines have only grown more selective about who they cite since the March 2026 core update tightened quality thresholds.

The firms that win the next five years of client acquisition are being chosen right now, one answer at a time, by engines that have no idea which firms are actually good. They only know which firms are legible.