The Insurance Broker’s Guide to B2B Lead Generation

The six things that make it hard

Every brokerage we have worked with, from the largest broker network in Australasia to a fifteen-page specialist in surety bonds, runs into the same six obstacles. They are worth naming, because the method in this guide is a response to each one.

1. The decision-maker is busy and already insured. The CFO, finance director or owner you need is being pitched constantly, and has a broker today. You are asking someone to change a relationship, not to start one.
2. Most leads are the wrong leads. Enquiries are easy to generate. Enquiries from businesses that fit your minimum premium, your classes of business and your appetite are not. A lead engine that cannot tell the two apart is a cost, not an asset.
3. The cycle is long and committee-led. Commercial cover is decided at renewal, by more than one person, over weeks. A specialist strata product we ran paid search for had a two-to-six-week cycle between quote and bound policy, and every short-term metric looked wrong until the pipeline matured.
4. Loyalty is sticky. A business that has had the same broker for a decade needs a concrete reason to move: a class of cover the incumbent does not understand, a market the incumbent cannot access, or evidence that someone else takes their risk more seriously.
5. Volume and value pull against each other. Bidding on broad terms fills the inbox and empties the diary. The trick is to buy fewer, better conversations, and this guide spends a whole chapter on it.
6. The rules are real. An AFS licensee in Australia cannot call itself independent if it earns commission. A UK brokerage’s advertising is a financial promotion under FCA rules. A New Zealand financial advice provider has disclosure duties. Every page, ad and email has to be written inside those lines.

Why it is worth the effort

Broker economics reward acquisition more than almost any other service business, because clients stay. NCI, a trade credit insurance brokerage we have run Google Search Ads for since 2018, retains a client for seven years on average. That single fact turns a 434% return on ad spend in Year 1 into a 2,037% return across the life of the clients acquired. The cost of acquisition is paid once; the commission is paid every year.

• 9.7×Broker leads vs the prior agency on equivalent spend, Year 5 (Steadfast)Case study →
• 2,037%Lifecycle ROI by Year 7 from Google Search Ads alone (NCI)Case study →
• 10Enquiries in month one after eleven years without a website lead (BCS Broking)Case study →

What follows is in the order we now build it: the website first, then organic and AI search, then paid search, then the outreach, conversion, nurturing and measurement that turn interest into bound policies. The 2024 edition of this guide started at step four. The brokerage in chapter one is the reason it no longer does.

Eleven years without a lead

BCS Broking place commercial insurance and surety bond facilities for Australian construction and mining companies turning over A$20 million and up. Contract works, professional indemnity, mining rehabilitation bonds: cover a finance director cannot buy from a comparison site. They are good at it, and they had been good at it for eleven years without their website producing a single lead worth the name.

The reasons were visible the moment the admin was opened. Fifteen pages in total, on a WordPress theme with a page builder layered over it. The homepage had last been published in 2016. Policy pages carried no focus keyword and no meta description. There was one general Insurance page and one general Surety page, so a search for “mining rehabilitation bond” or “civil construction insurance broker” had nothing specific to land on. For a business whose entire value is depth in a narrow field, the site said nothing a specialist would recognise.

The rebuild followed the structure of the market rather than the structure of the business: nineteen insurance verticals, each with its own page; eleven surety bond types, each with its own page; comparison pages for the questions that come before the purchase, such as surety bonds versus bank guarantees. The titles follow one pattern because the searches do: [vertical] insurance broker Australia. Underneath it, a static build served from an edge cache. The homepage answers in as little as 56 milliseconds and weighs 124 kilobytes.

What a brokerage website has to do

• One page per class of business you place, and one per industry you serve. A CFO searches for the cover they need in the words they use. A page called “Insurance” competes with every insurer on earth; a page called “Contract works insurance for civil contractors” competes with almost nobody.
• Answer the question that comes before the purchase. Surety bond or bank guarantee? What does a broker do that going direct does not? What will the underwriter ask for? These pages earn the trust the policy pages convert.
• Say who you are for, and who you are not for. Minimum premium, turnover range, appetite. It costs you the wrong enquiries and it is why the right ones arrive already qualified.
• Name the people. A broker is a professional relationship. Photographs, names, authorisations and a line on what each person actually places. Anonymous brokerages are hard to trust and, as chapter two explains, hard for an AI engine to recommend.
• Make the next step obvious and cheap. A phone number that is answered, a short enquiry form, a booking link. Not a nine-field quote form on the first visit.
• Carry the proof. Named clients where they will allow it, the classes you have placed, the markets you have access to, professional memberships. Vague claims of “tailored solutions” are the default; they persuade nobody.

Speed is a ranking factor and a trust factor

Google measures how real visitors experience your site and uses it in ranking. The three Core Web Vitals are Largest Contentful Paint (the main content should appear within 2.5 seconds), Interaction to Next Paint (the page should respond to a tap within 200 milliseconds) and Cumulative Layout Shift (nothing should jump around after it loads; a score under 0.1). Most brokerage sites we audit fail the first one on a phone, because a theme and a page builder ship a megabyte of scripts before a word of copy.

The commercial cost is simpler than the technical one. A finance director on a train opens two brokers from a search. One paints in a second, the other in six. The second broker does not get a second chance, and never finds out.

Our free website audit measures six things on any site, from the bytes it actually sends rather than from a score: whether the final URL is secure, whether the homepage title says what the business does, whether anything on the page tells Google it is a local business and where, whether the page tells a phone how to size itself, whether there is a description for search results to show, and how heavy the document and its render-blocking assets are. A brokerage that fails three of those is invisible for reasons that have nothing to do with the quality of its advice.

Writing inside the rules

Compliance is not a reason to say less; it is a reason to say precise things. In Australia, an AFS licensee that receives commission cannot describe itself as independent, impartial or unbiased (Corporations Act s923A), and a comparison page that implies otherwise is a problem in an ASIC review. In the United Kingdom, advertising by an authorised firm is a financial promotion and has to be clear, fair and not misleading, with the FCA’s rules on prominence and balance. In New Zealand, a financial advice provider has disclosure obligations under the Financial Markets Conduct Act, and the FMA takes an interest in how advice is advertised.

The practical version: describe what you place, for whom, and through which markets. Use client outcomes with permission and with the caveats the licence requires. Have your compliance officer read the site once, and keep a note of what they approved. Then never write a page that could not pass the same reading.

SEO for a brokerage is topical depth, not keyword stuffing

Google ranks the site that demonstrates it understands a subject, not the site that repeats a phrase. For a brokerage that means covering a class of business completely: the policy page, the industries it applies to, the questions asked before buying, the claims scenarios, the underwriting information required, the renewal timing. BCS’s nineteen vertical pages and eleven surety pages are not padding; each one is the only page on the internet written for that exact search.

• Map the searches to pages. Every class of business you want to grow gets a page whose title uses the buyer’s words and the market: “[class] insurance broker [country or city]”.
• Build the supporting layer. Two or three articles per class that answer the pre-purchase questions, linked to and from the policy page. This is what earned BCS its citations: the AI answer quoted their guides on bank guarantee alternatives and government contract requirements.
• Google Business Profile, properly. Category “Insurance broker”, every office, the service areas you actually cover, and reviews from clients who will give them. For city-plus-class searches the map results are half the page.
• Earn the links you deserve. Industry association directories, insurer partner pages, underwriting agency panels, local business bodies, the trade press for the industries you serve. Nobody buys links; brokerages have more legitimate places to be listed than almost any other business.
• Fix the foundation once. Fast pages, correct canonical URLs, structured data for the organisation and each office, a sitemap that reflects the real site. This is unglamorous and it lifts every page at once.

Being the broker the AI names

More of your prospects now start with a question to an assistant than with ten blue links. Google’s AI Overviews and AI Mode, ChatGPT, Gemini, Copilot and Perplexity all answer “which insurance broker should a mining contractor use in Western Australia” by assembling a shortlist from what they can find and verify. Getting onto that shortlist is answer engine optimisation (being the cited source) and generative engine optimisation (being the named entity), and for a brokerage the requirements are specific.

1. Be a resolvable entity. One consistent name, address and description everywhere the engine looks: your site’s structured data, Google Business Profile, LinkedIn, the association directories, the insurer panel pages. An entity the engines cannot pin down is an entity they will not recommend.
2. Have a page for the exact question. The AI answer for BCS did not cite a homepage. It cited the surety bond hub and two guides. Engines quote the page that answers the question the user asked, and a general page answers none.
3. Put named, credentialed people on the pages. Engines weight experience and accountability. A broker with a name, an authorisation and a track record on the page is a source; “our team” is not.
4. Say the specific, checkable thing. “We placed the contract works programme for a A$40 million civil project in 2025” beats “extensive experience in construction insurance”. Specific claims get quoted; vague ones get skipped.
5. Get corroborated. Reviews, mentions in the trade press, listings in the directories the engines already trust. These are the sources an AI answer cites beside your own site, and they are how it decides you are real.

One channel, done properly

NCI’s acquisition engine is Google Search Ads and nothing else. No display, no social, no remarketing. Every high-intent search for trade credit insurance in Australia is captured, converted at the best possible rate, and the compounding economics do the rest. Every dollar invested returns A$5.34 in Year 1 commission, a 434% return, and because clients stay for an average of seven years the same programme compounds to 2,037% across their life. A$9.6 million in cumulative revenue by Year 4, from one channel.

That result is not a product of a large budget. It is a product of restraint: exact and phrase match on the searches that mean “I need this cover”, a negative keyword list built before the first ad ran, a landing page written for the buyer, and a bid strategy calibrated to the value of a bound client rather than the platform’s idea of a conversion.

Phrase and exact match, and why broad match empties the diary

Broad match asks Google to show your ad for anything it decides is related. For a brokerage that means “insurance jobs”, “cheap car insurance”, “insurance courses” and “is insurance broking a good career”. Phrase and exact match show your ad for the words a buyer actually uses: “surety bond broker perth”, “professional indemnity insurance for engineers”, “trade credit insurance broker”. Fewer clicks, more expensive clicks, and far more of them from someone who can sign.

Here is the shape of that trade-off, modelled for one commercial brokerage account. The same monthly budget, two ways.

The broad account looks busier on every dashboard Google shows you: more clicks, more “conversions”, a lower cost per lead. It produces a third of the conversations at nearly four times the price, and the brokers stop trusting the leads within a month. The specialist strata insurer in the next section ran phrase and exact only from day one, for precisely this reason.

Structure that respects a committee-led sale

A specialist residential strata underwriter came to us with a challenge most agencies would sidestep: prove that paid search could grow a product with a two-to-six-week committee-led sales cycle, where strata committees rather than owners make the decision and every state uses different words for the same thing. The CFO set the bar at a 1.5× Year 1 return before any larger commitment.

• Three Search campaigns, each with its own keyword strategy, geography and bids. Match types strict: phrase and exact only, to keep renters and unrelated searches out.
• Negatives before launch. A comprehensive negative list built before a single ad went live, and reviewed weekly against the search-terms report.
• Geography by risk, not by map. Strict postcode targeting to avoid the high-rise towers that inflate cost per click without converting.
• One decisive landing-page test. Half the traffic to the homepage, half to the direct quote engine. The quote engine cost over A$980 per quote. The homepage, which gave committee-led buyers the context they needed first, came in under A$80. The quote-engine destination was killed inside weeks.
• Cut what does not bind. Display and remarketing produced four quotes and zero policies. Both were paused by week eight and the spend redirected to search.

At the eight-week mark the account showed 111 quotes and 30 bound policies, a 27% strike rate against a projected 40%. On paper it looked like underperformance. It was pipeline sitting with committees. Pacing was slowed so the spend and the pipeline matured together, and the programme closed its first phase on 213 quotes, 68 bound policies and a 1.59× Year 1 ROI. Phase two was approved.

• A$80Cost per quote after the landing-page test, down from A$980+Case study →
• 68Policies bound from 213 quotes in Phase 1Case study →
• 1.59×Year 1 ROI on a committee-led productCase study →

The working rules

1. Search network only. No search partners, no display expansion, no “performance max” until the search account has taught you what converts.
2. Campaigns by class of business or by market, never one campaign for everything. Each has its own budget, so a strong class cannot be starved by a weak one.
3. Ad groups of five to fifteen tightly related terms, phrase and exact, with the same words in the headline and on the landing page.
4. Location targeting set to people in the area, not people interested in it. Business hours scheduling, with a bid reduction outside them if the phone is not answered.
5. One landing page per ad group, written for that buyer, with the phone number and the short form above the fold. Never the homepage for a class-specific search, and never the quote engine for a committee-led one.
6. Call tracking and form tracking into the CRM, so the conversion the account optimises towards is a qualified enquiry, not a page view.
7. Manual or capped bidding for the first thirty days; automated bidding only once an ad group has enough real conversions to learn from.
8. A weekly search-terms review for the first two months, adding negatives every time. The list is never finished.

Broker networks: the Steadfast pattern

Steadfast Group is the largest general insurance broker network in Australasia. In 2018 they moved their digital marketing to us from Bohemia, part of M&C Saatchi, with one mandate: more “Find a Broker” leads from the same annual budget, benchmarked against the outgoing agency’s numbers. Within three months the campaigns delivered five times the broker leads Bohemia had achieved, using 23% of the annual budget. By the end of Year 1 lead volume was six times the prior baseline. By Year 5, 9.7 times, on equivalent spend, across more than two thousand micro-campaigns in search, social, audio and display managed with a media-mix model that moved budget to whatever was converting that week.

For an individual brokerage the lesson is not the scale. It is the single-minded call to action, the willingness to test every channel small before scaling any of them, and the habit of measuring against the only number that mattered to the client: leads to brokers, per dollar.

Know the four people in the room

• The owner or managing director of a small or mid-sized business. Wants the risk handled and the premium fair; will move for a broker who explains cover in their own industry’s terms.
• The finance director or CFO. Thinks in total cost of risk, cash flow at renewal, and what a claim would do to covenants. Responds to numbers and to evidence that you have handled businesses like theirs.
• The risk or operations manager. Lives with the exposures day to day: contract requirements, site conditions, contractual indemnities, licence conditions. Responds to specificity about their class of work.
• The incumbent’s weakness. Not a person, but always in the room. Every prospect has a broker; your message has to name what that broker is likely to be missing for this kind of business.

Segment, then say something specific to each

One message for everyone is the same as no message. Pick the segments you actually want, by industry and by class of business, and write for each. A civil contractor, a private hospital and a food manufacturer are three different conversations even when the policy class is the same.

• To a finance director: “Contract works and PI placed for civil contractors turning over A$20–200 million, with a bond facility that does not tie up your working capital.”
• To a risk manager: “Mining rehabilitation bonds, environmental liability and plant cover for operators who have to satisfy state regulators before they can dig.”
• To an owner: “Business insurance for trades and construction businesses, reviewed every renewal by a broker who has read your contracts.”

Lead magnets that a CFO would actually open

A lead magnet trades something useful for a name and an address. For commercial buyers it has to be genuinely useful and specific to their situation; a generic “guide to business insurance” is a brochure with a form in front of it. What works is the document they would otherwise have to write themselves.

• A contract-review checklist: the insurance clauses to look for in a head contract, and what each one commits you to.
• A renewal-preparation pack for a specific industry: what the underwriters will ask, what evidence lowers the premium, the timeline that avoids a lapse.
• A comparison that answers the pre-purchase question honestly: surety bond versus bank guarantee; captive versus market; direct versus broker.
• A short, dated market note for a class you place: what capacity looks like this year, where rates are moving, what is being excluded.

Each of these becomes a page on the site in its own right, which means it ranks and gets cited as well as gating an email. The guide you are reading is built exactly that way.

Four channels, one month, one Melbourne broker

Tudor Insurance is a Melbourne broker with decades of reputation and, in 2022, a digital presence that had not kept up. The brief was growth, fast, with the numbers to prove it. We built a four-channel system: paid search to intercept buyers at the moment of decision, personalised cold email to open conversations, LinkedIn InMail to reach decision-makers directly, and display remarketing to stay in front of anyone who had shown interest. Each channel had a distinct role, and the creative for all four carried the same positioning: an established Australian broker, digitally confident.

• 107Leads in the first monthCase study →
• 47New clients in the first monthCase study →
• 1,392%Return on investment, month oneCase study →

LinkedIn: the only channel where the job title is a targeting option

• Sponsored content to segments, not to “business owners”. Finance directors in construction, operations managers in mining services, owners of businesses in the industries you serve. Promote the lead magnets from chapter four, not a brand video.
• InMail that reads like a broker wrote it. One paragraph on their situation, one on what you place for businesses like theirs, one clear ask. No “I hope this finds you well”, no attachment, one follow-up a week later.
• The brokers as the voice. Posts from named brokers about real placements (with permission), regulatory changes in a class, what the underwriters are asking this renewal season. This is also the corroboration the AI engines look for.

Email: personal at the top, automated underneath

Cold email to commercial decision-makers works when it is specific and legal. Both are non-negotiable. In Australia the Spam Act requires consent or an existing relationship and a working unsubscribe; in the UK and New Zealand the rules differ in detail and not in spirit. Business-to-business outreach to a role at a company, about that company’s insurance, written by a named person, is a different thing from a blast, and it is treated differently by the recipient and the regulator.

• Lists built by industry and class, from sources you are entitled to use. Small, accurate lists outperform large purchased ones every time.
• Sequences of three, each shorter than the last. The first names their industry and one specific exposure; the second offers the relevant lead magnet; the third asks for fifteen minutes or closes the loop.
• Replies routed to a broker within the hour. The value of the channel is the conversation, and a conversation that waits a day is a conversation with the incumbent.

Remarketing and social: useful in their place

Display remarketing kept Tudor in front of prospects who had already engaged, and earned its place in that mix. For the strata insurer, display and remarketing produced four quotes and no policies and were cut inside eight weeks. Both are true. The rule is to test every supporting channel small, measure it against bound business rather than clicks, and keep only what earns its spend. For Steadfast, a media-mix model did this continuously across search, social, audio and display; for a single brokerage a monthly review of cost per bound policy by channel does the same job.

What a converting page looks like

1. It matches the search or the offer exactly. A click on “contract works insurance broker” lands on the contract works page, not the homepage. A click on the renewal-pack lead magnet lands on a page about that pack.
2. The headline says the outcome in their words. “Contract works and liability cover placed for civil contractors, with a bond facility that does not tie up your capital.”
3. Three to five lines of benefit, not a wall of copy. What you place, for whom, through which markets, and one line on how the process works.
4. Proof above the fold. Named clients with permission, the classes and markets you have access to, professional memberships, the brokers’ names.
5. One clear call to action, two ways to take it. A phone number that is answered and a short form: name, company, email, phone, what they need covered. Nothing else on the first visit.
6. Fast. Everything in chapter one applies with more force here, because this visitor is paid for.

Straight into the CRM

A form that emails an inbox is a lead that gets lost. Every enquiry should land in the CRM with its source (which campaign, which keyword, which email sequence), its segment and its timestamp, and be assigned to a broker with a service level for first contact. That record is also what makes chapter eight possible: without it you cannot know which channel produced the clients who bound and renewed.

• Capture source and campaign automatically; never rely on the prospect to tell you how they found you.
• Progressive profiling: ask for the minimum on the first form and gather the rest over the next interactions.
• Automated acknowledgement within a minute, and a human within the hour during business hours.
• A qualification step the brokers actually use, so “genuinely qualified” in the paid search model is a field, not an opinion.

A sequence that earns the right to ask

1. Immediately: deliver and introduce. The guide or pack they asked for, one paragraph on who you are and what you place, the name of the broker who will be their contact.
2. A few days later: teach something. One useful, specific piece for their segment: what changed in their class this year, a clause to check in their contracts, a claims example that shows what good cover does.
3. A week later: show the proof. A short case of a business like theirs, with the outcome and how it was achieved. Permission and compliance first, always.
4. Two weeks later: remove the objection. Switching brokers is easier than they think, and here is what it involves and when to do it relative to renewal.
5. Then: the ask. Fifteen minutes with the named broker, a booking link, and a note of their renewal date so the follow-up lands at the right time.

After the sequence, a quiet monthly note keeps the relationship alive: a market update for their class, a regulatory change, an invitation. The prospect who did not move this year is often the client next year, and the cost of staying in touch is close to nothing.

Automation that stays personal

• Segment by industry and class so the content is relevant; a strata committee and a mining contractor never receive the same email.
• Send from a named broker, with their signature and direct line, and route replies to that person.
• Trigger on behaviour: a prospect who reads the surety page twice gets the surety content next, not the generic sequence.
• Record the renewal date at the first opportunity and build the sequence timing around it. Insurance is bought on a date; the nurture should know it.

The metrics that matter, in order

1. Cost per qualified enquiry, by campaign and by keyword or sequence. Not cost per lead; the qualification field from chapter six is what separates them.
2. Cost per quote, and the enquiry-to-quote rate. This is where landing pages and broker response time show up.
3. Quote-to-bind rate and cost per bound policy. The number a CFO recognises. For a committee-led product, read it over the length of the cycle, not the calendar month.
4. Premium and commission per bound policy, by source. Some channels produce smaller clients; some produce the ones that matter. Budget follows this number.
5. Retention by source. The final word on lead quality arrives at the first renewal.

Optimising without breaking what works

• Test one thing at a time: a headline, a landing page, a bid cap. The strata landing-page test was decisive because nothing else changed.
• Cut what does not bind. Display for the strata insurer produced quotes and no policies and was paused in week eight; a channel that generates enquiries and no clients is a cost centre with good reporting.
• Refine audiences and negatives from real data every week for the first two months, then monthly. The search-terms report is the cheapest insight in marketing.
• Close the loop with the brokers. They know which leads were real; that knowledge belongs in the CRM and in next month’s targeting.

The calculation

Return on investment is the commission earned from clients acquired, less the cost of acquiring them, divided by that cost. A programme that costs 10,000 and produces 50,000 in first-year commission returns 400%. That is the number most dashboards stop at, and it is the wrong one, because the clients it acquired are still paying commission in year five.

NCI’s clients stay seven years on average. Their Google Search Ads programme returns A$5.34 for every A$1 in Year 1, a 434% return. Carry the same clients across their average life and the return on the original spend reaches 2,037%. Year 1 commission of A$961,714 became A$9.6 million in cumulative revenue by Year 4, with a 20% monthly budget-scaling model that lets the programme grow as the evidence accumulates. In the first quarter of 2026 alone the campaigns produced A$291,720 in projected commission, a 434% return in ninety days. Lead quality held, conversion held, cost per acquisition held, at scale.

• A$5.34Returned for every A$1 invested in Year 1 (NCI)Case study →
• 7 yearsAverage client retention, the multiplier on every acquisitionCase study →
• A$9.6MCumulative revenue by Year 4 from Google Search Ads aloneCase study →

Six ways to raise it

1. Narrow the targeting. Every dollar moved from a broad term to an exact one raises the qualified rate. The model in chapter three is the whole argument.
2. Raise lead quality at the source. Qualifying language on the page, minimum-premium honesty, forms that ask the one question that separates a fit from a tyre-kicker.
3. Test the destination and the offer before the bids. The largest single improvement in any account in this guide was a landing-page change, not a bidding change.
4. Follow up fast and for longer. The hour after an enquiry and the months before a renewal are where most brokerages leak the value they paid for.
5. Automate the repeatable. Acknowledgements, sequences, CRM entry, reporting. The broker’s time goes to the conversation, which is the only part a machine cannot do.
6. Retain and expand. A client kept for a seventh year and cross-sold a second class is the cheapest revenue a brokerage will ever earn, and it is why the acquisition cost was worth paying.

What this looks like over a year

• Quarter one: the site rebuilt around classes and industries, fast, compliant, with named brokers; tracking into the CRM; Google Business Profile and directories done; the first paid search campaign on phrase and exact for your strongest class.
• Quarter two: supporting content for the top three classes; the first lead magnet and its LinkedIn and email programme; negatives and landing pages refined from real search terms; the first citation checks.
• Quarter three: scale what is binding; add the second and third classes to paid search; nurture sequences keyed to renewal dates; cut every channel that is not producing bound policies.
• Quarter four: read the year on cost per bound policy and retention by source; set next year’s budget against lifetime return; add the markets and classes the evidence supports.

None of this requires a large budget. It requires the discipline to do it in this order, the patience to judge a committee-led sale on its own timeline, and the honesty to cut what is not binding. The brokerages in this guide did that, and the numbers are theirs.