Choosing a fintech SEO agency: nine questions that separate specialists from generalists

The problem for a finance brand hiring an SEO agency is not finding candidates. It is that the good ones and the confident ones look almost identical in a pitch.

Every agency will claim finance experience, name a framework, show a dashboard and promise a strategic partnership. Distinguishing them requires asking questions that a generalist cannot answer well, and paying attention to which answers arrive without hesitation.

This is a buyer’s guide, and it is written on the assumption that some readers will conclude an agency is the wrong answer for them. That outcome is worth the read.

What a specialist should be able to show before you ask

A finance specialist has usually done a few things as a side effect of working in the category, and those are the fastest signals available.

They can describe your market’s search landscape before you brief them. Not perfectly, but specifically enough to name the clusters that matter and the type of property holding them. A generalist explains their process at this point. A specialist tells you something you did not know about your own market.

They know what a compliance reviewer will reject. Financial promotion rules, disclaimer requirements, restricted claims and jurisdiction-specific restrictions shape what can actually be published. An agency that has never worked with a regulated marketing team will produce content that dies in review, and will blame the review process for it.

They can talk about entities, licences and markets without prompting. Multi-entity brokers, exchanges operating under several regimes and fintechs with a parent brand and regional subsidiaries all create search architecture problems that are specific to regulated businesses. If the conversation never reaches that subject, the specialism is probably decorative.

Nine questions worth asking

1. Which commercial clusters would you target first, and why those? The answer should reference the buyer’s decision path rather than search volume alone. If every recommendation is a high-volume term, the strategy is a traffic plan.

2. What is the qualification event for organic traffic in this business? A funded account, a verified deposit, an approved merchant account and a subscribed publisher are different outcomes. An agency that cannot name the right one will report on the wrong one.

3. Who is holding the SERPs today, and what would it take to move them? The answer should name affiliates, comparison publishers, regulators, marketplaces and communities as appropriate, and should include a candid view of which positions are not worth contesting.

4. What will you refuse to do? A useful answer includes link schemes, guarantees, and content the compliance team cannot approve. An agency with no exclusions is selling production rather than strategy.

5. What does your reporting separate? The right answer distinguishes branded from non-branded demand. Without that split, sponsorship and paid activity inflate organic reporting and hide whether the programme is creating demand or collecting it.

6. How will this survive a compliance review? The answer should describe a workflow, not a promise. Briefs that flag claims needing substantiation, reviewers who understand the formats, and turnaround expectations set in advance.

7. Who does the work, and what changes if they leave? This question is uncomfortable for every agency and the answer is informative. Ask about continuity, handover documentation and whether the strategist in the pitch is the person in the work.

8. What would make you recommend we stop? A specialist will have an answer about sites that cannot convert, categories that are not winnable, or businesses whose constraint is product rather than channel. A generalist will treat this as a trap.

9. What have you learned from a client engagement that went badly? The ability to describe a failure crisply, without blaming the client, is a better indicator of working practice than any case study.

Red flags, in rough order of seriousness

Guaranteed rankings or traffic. No provider controls the ranking algorithms, and a provider who guarantees a position has either misunderstood the assignment or intends to deliver it through terms nobody read.

Anonymous results. Percentages with no client, no market, no timeframe and no baseline. Either the client is identifiable or the number is fabricated.

Case studies that cannot be checked. Ask which brands a claimed client relationship was with and what the basis was. Agencies that describe experience vaguely are usually describing experience held by someone else, or by nobody.

A framework with a trademark. A named process can be useful shorthand. It is not evidence of anything, and a proprietary-sounding methodology frequently wraps standard work in language that makes it harder to scrutinise.

No interest in your analytics. An agency that proposes a strategy without asking how the business currently converts is guessing at the commercial model.

A large content volume in month one. Thirty articles in the first month usually means the strategy is production volume rather than market position, and the resulting pages are the ones you will be pruning in a year.

Reluctance to write down scope. Timeframes, deliverables, access requirements and terms should be documented before work starts, particularly where the agency is also selling the measurement.

Agency, in-house or fractional

The honest comparison depends on what your team already contains.

In-house is stronger when you have someone who understands the category, the compliance process and the product, and the gap is capacity rather than capability. An internal team also holds context that takes an agency months to acquire. If your constraint is that nobody has time, hiring for time is usually better value than buying the same time at an agency rate.

An agency is stronger when the missing ingredient is category knowledge, market research capability, or a technical specialism you do not need permanently. It is also the right answer when you need to move quickly in a market you have not competed in before.

Fractional leadership sits between them. It suits a team that can implement but has no senior search owner setting the strategy and holding the standard. The output is a roadmap, a measurement frame and someone in the room with enough authority to say no.

The wrong answer is hiring an agency to fill a capability gap that is actually an organisational one. If nobody internally can approve a page, the constraint is approval capacity, and no external team will fix it.

What the first ninety days should contain

Any credible plan covers the same ground in roughly the same order, regardless of the framework attached to it.

The first phase establishes the market position: which commercial searches exist, who holds them, what suppresses your site and which opportunity is worth funding first. Nothing should be recommended before that exists, and a programme that starts with content in week one has skipped it.

The second phase fixes the constraints that would waste the content: rendering, indexation, entity and market architecture, and the internal process for getting pages published. A site that cannot get a page indexed cannot benefit from a better page.

The third phase begins publishing against the priority book, with reporting that tracks the agreed qualification event rather than sessions.

If a plan arrives with more than one of those phases missing, ask what the agency expects to happen instead.

How to read a pitch deck

Deck order usually reveals priority. A deck that opens with the agency’s culture and closes with your market has its priorities inverted.

Look for: an early, specific claim about your market rather than your industry. Named people and what they will do. A stated exclusions list. A measurement plan that uses your qualification event. And an honest section about what will be hard, which is the single most reliable indicator that the agency has done the work before.

Be sceptical of any deck that contains a projected traffic number with no stated assumptions, and of any deck where the only case studies are in categories that are easier to rank in than yours.

Measuring whether it is working

Judge the programme on leading and lagging indicators separately, and expect the leading ones to move first.

Leading: share of non-branded commercial queries where you appear, number of clusters where you hold a position, pages published and indexed against the plan, and the time from brief to published. Lagging: non-branded organic registrations, qualified accounts or deposits, and revenue where attribution permits.

The measure that matters most is the one most agencies avoid: whether non-branded organic demand reaches your qualification event. A programme can grow traffic, sessions and rankings indefinitely without moving that number, and if it does, the reporting will not mention it.