If you are reading this, you are looking for a CRM for your insurance practice. Short answer: there is no single CRM that fits every broker. The right one depends on what kind of insurance you sell.
Insurance is two businesses sitting under one roof. The first is acquisition, where you sell the policy. The second is retention, where you keep the client (e.g. when you sell for a second time when policy expires OR when a risk happens and you need to help with reimbursement). Both happen on a scale that runs from pure commodity like car insurance to fully bespoke like insuring artwork. At one end of the scale you do not need a CRM at all. At the other end you cannot use one. The middle is where the CRM works best. The picks at the middle, in the order I would think about them, are Pipedrive, Salesforce, and Attio. The platforms to avoid are Zoho and Microsoft Dynamics.
In this article I will walk you through the framework that decides which one fits you. I will tell you when a CRM matters, when you should absolutely forget about implementing one, and what to ask for when you will be on the market looking for CRM. I will also flag the trap of hyper-local insurance CRMs, the puppy excitement effect that catches every broker the first time they sit through a demo, and a short note about AI agents. By the end you will know what to pick and what to skip.
Let me illustrate this for you
Now, let me illustrate this for you. Two brokers. Two completely different days.
Broker A sells car insurance in Rotterdam. Her job is speed. Each morning she gets fifty inquiries through her website and issues fifty policies same day. She does not speak to a human buyer. The site does the work. Her margin per policy is thin. Her competition is fierce. Twenty other carriers sell the same product. Her main asset and tool is her website.
Broker B sells art insurance. He has been on one project for four months. A private collector wants to insure a Renaissance fresco that has never been to auction. Broker B is on the phone with a museum conservator, a lab in Florence, and an underwriter who needs three more pieces of paperwork. He will close five of these projects this year. He will earn well.
Same industry. Different tools. The reason is the scale they sit on. Stick with me. I will walk you through that scale and tell you what to pick at each end.
The customer journey is the first map you need
Before we talk about software, let me lay out the framework I use whenever a brokerage asks me what to buy. It is called the Customer Journey Map. The simplest piece of strategy I know.
The journey runs from the moment a client has no idea your firm exists, all the way to the moment that same client is at a barbecue telling a friend how brilliantly you handled the time their car wrapped around a tree. Five stages.
Stage one is awareness. The buyer has a problem. They start to research. They Google. They ask a chatbot. They send a quiet “hey, do you know a good broker?” over WhatsApp.
Stage two is consideration. The buyer is comparing. They call brokers. They fill in a form on your website. They land in your inbox as a lead. This is where pipeline management starts to matter, because prospects begin to pile up.
Stage three is purchase. The transaction. The simplest transaction in the world is buying gum at a self-checkout. The most complex is building a factory. Depending on what insurance you sell, it could sit closer to chewing gum or closer to building a factory. It’s a complexity scale. More on that in a moment. You collect data, send it to the insurance company, wait on a confirmation, sometimes chase a missing document.
Stage four is delivery. In insurance, delivery is the policy term. One date in your system, the end date. During that window the client may file a claim. They may wrap their car around a tree while drunk, driving someone elses’s wife’s car that was insured with you and even walk away unhurt from the accident, and call you Tuesday morning expecting you to make it right.
Stage five is advocacy. The client tells the story. They tell it at the barbecue. They tell it on LinkedIn. They send their cousin to your office.
Now, you need CRM in stages two, three, and four. In stages one and five, the work is marketing and trust. Hold this journey in your head while we walk through the rest.
Every brokerage runs two businesses under one roof
Once the journey is clear, the next thing to notice is that an insurance brokerage runs two businesses sitting under the same roof.
The first is acquisition. A new client walks in. You sell them their first policy. You collect a commission. Most of stages one through three live here.
The second is retention. The policy will expire in twelve or twenty-four months. The client may have a claim in between. When the policy ends, the client renews with you, switches carrier through you, or quietly leaves. Stage four lives here. Stage five depends on how you handled the other three stages before that.
These two businesses use different muscles. Acquisition is a marketing-and-funnel discipline. Retention is a service-and-renewal discipline. Software that is good at one is often weak at the other. Most brokers I meet pour attention into acquisition and treat retention almost like an afterthought. The CRM you choose MUST do both. Period.
The scale that decides everything else
As I’ve said, insurance covers a wide product range. Car policies that cost €40 a month sit at one end. Art insurance that takes six months to underwrite sits at the other. Each product fits on a scale from 1 to 10. Where it fits decides what software you need.
A 1 is pure commodity. The buyer cannot tell you which carrier sold them the policy. Think about how a car has changed in a hundred years. In the 1920s a car was a luxury. Today it is an appliance. You can buy a working used car in Amsterdam for €2,000 if you look hard enough. The price collapsed. The product commoditized. The insurance that goes with it commoditized too. Twenty carriers sell the same coverage. The buyer compares them on a single screen. The policy is signed before they have read the company name.
A 10 is fully bespoke. Insuring a Renaissance fresco is legit. Authentication takes weeks. Valuation takes months. You may have to ship the work to a museum lab in Florence to confirm it is real. There is no comparable sale on record. Each project is structurally different. You might close five a year.
In between, the products fall along the scale. Life insurance sits around a 4 in the Netherlands. In the United States, where it sells more often, it is closer to a 2 or 3. In Serbia it is a 6, sometimes a 7. Pension and social products sit around a 2 in most of Europe. They barely exist in some former Soviet states. Professional liability cover for a freelance consultant, which I pay €120 a month for myself, is a 2. Health insurance in the Netherlands is a 1. Twenty carriers sell variants of the same regulated product.
The biggest job a brokerage can do for itself is to find its own number on this scale, per product line. Once the number is in front of you, you can make the correct judgement. Which bring us to the next point.
At a 1, your CRM is your website
If most of what you sell sits at a 1 or a 2, you can forget the CRM. There is no sales pipeline to manage. There is no sale to manage. You cannot afford to sell one policy an hour. You need to sell fifty policies an hour. The margin is thin. Your weapon is speed and zero friction.
What you need at this end is a self-service portal. That sounds simple. It is a serious engineering project. The site has to talk to a national insurance database via API. It has to pull live offers from every carrier you represent. As a broker, you are licensed across maybe ten carriers in your country. The site has to hold buyer data securely. It has to produce a valid contract at the end. There is no off-the-shelf cheap version.
If you find one, beware. The easier it is to build, the more brokers in your country have already built one. The space is already commoditized.
Latvia and Estonia in the EU have made this easier than anywhere else. They built central insurance databases. From a single source you can pull every offer from every carrier. As a result, every broker in those countries already has a portal. Starting a car insurance business there as a standalone is rarely worth the fight. The exception is the broker who runs a one-stop shop, where a client walks in once and gets every policy from a single counter. In that case the portal becomes a feature inside a wider portfolio.
Where the integration is hard, the difficulty is your moat. Every carrier has its own quirks. One asks for one set of fields. The next asks for a different set. Connecting six carriers can be a six-month project. The harder it is, the fewer competitors will pull it off. You set the local standard.
So at the commodity end, the real question is which engineering team will build the portal. That conversation almost always sits with a development partner rather than a CRM vendor.
At a 10, the CRM gets in the way
Travel to the other end of the scale. If you insure fine art, sculpture, or private collections, your tooling is a phone, a calendar, and a shared folder. A spreadsheet for project tracking is fine.
A CRM at this scale becomes a problem. CRM software is built around standard stages: prospecting, qualification, proposal, close. None of those stages map onto a fresco in a church, a Michelangelo statue, or a private collector’s drawing room. The work is bespoke from the first call. Every project is a one-off. Every project has a different shape and a different team of experts around it.
The cost of a deal at this scale is high. The volume is low. The value of standardizing the workflow is close to zero. Five to ten cases a year, each paying enough to fund the rest. CRM software at this end slows you down. Every project becomes a fight with the system. The system loses. It cannot bend.
If your average deal runs longer than two months and is worth more than a hundred thousand euros, the CRM is the wrong problem to solve.
The middle is where the CRM earns its keep
Now we arrive at the most interesting place on the scale, somewhere between a 3 and a 7.
You might be asking a question right now. Why do we need a CRM at all? If the commodity end runs on a website and the bespoke end runs on a phone book, where does the CRM fit in?
Here is the answer. In the middle, the volume of prospects is too high to hold in your head, and the value per deal is high enough to justify human follow-up. That is the gap a CRM was built to fill. Life insurance is the textbook example.
Life insurance is a hybrid product. It is part savings account and part death benefit. The buyer accumulates value over time. During the years before the value reaches the target, the policy guarantees the family will receive the agreed sum if the buyer passes away. I once looked at a million-euro policy on my own life. The carrier assessed whether my career trajectory plausibly justified that figure. They wanted to know whether I could earn a million euros over the working decades I had left. The premium decreases as the savings approach the target, like a mortgage in reverse.
A solo broker carries 20 to 100 prospects at any time. A firm carries 300 to 400. Each prospect sits at a different point on the journey. One is waiting on a medical exam. One is reading a proposal that landed yesterday. One has gone quiet and needs a follow up phone call. The volume is too high to hold in your head. The volume is too low to fully automate.
That is the textbook definition of pipeline management. That is what a modern CRM was built to do.
The CRM picks, ranked
Here is how I think about the picks at this end of the scale. Start at the top of the list. Move down only if the previous one does not fit.
Pipedrive is the simplest pick that works
For pipeline management with nothing fancy on top, Pipedrive is hard to beat. Built for the job. Fast to set up. Affordable. The team will open it on a Monday morning without complaint.
The trade-off is custom integrations. Pipedrive does not extend well. The day you decide to add a self-service renewals portal, plug in a chatbot, or sell professional liability cover through the same engine, you will hit a wall.
If you sell one product line and you live in stage two of the customer journey, Pipedrive is the answer. Stop reading about CRM and buy it.
Bigin CRM is the cheap alternative
Same job. Lower price. The technical foundation is rough. I would not call it a beautiful piece of software. It works.
If budget is the deciding factor, Bigin is where to look. Set expectations. The system is going to feel clunky. Your team will get used to it. But it’s VERY similar in function to Pipedrive.
Salesforce is the proven path when you need extensibility
If you need more than pipeline management, this is the safe bet. Expensive. Slow to implement. Reliable. Find a competent partner (the kind of firm I run is one example) and the project will solve your problem. Guaranteed.
The catch is that it will be slow. It will cost more than it should. It will still be there in ten years. Inertia is real. No big incumbent loses its customers fast even when it should. Even if Salesforce stumbles, it will keep its enterprise customers for a decade through sheer switching cost.
Pick Salesforce when the project is big enough, the budget is committed, and the firm cannot afford to bet on a young platform.
Attio is the newer alternative
I discovered this platform myself this year. I should admit upfront that I am in the middle of a puppy excitement phase with it. I will explain that phrase in a minute.
Attio has fewer features than Salesforce by a wide margin. The architecture is modern. The integrations with AI tooling are something Salesforce cannot match without buying or bolting on extra products. You can plug a chatbot into the bench. You can plug Claude in next to it. Customers can talk to your CRM through chat or voice. You can put digital agents on your website that talk to your CRM in the background. Customers feel like they are talking to a member of your team while they are talking to a model. The technology is here today.
The risk is the youth of the platform. Will Attio be here in ten years? I do not know. Will Salesforce? Yes.
Pick Attio when the firm wants to move fast, the team is technical enough to live with rough edges, and the AI layer matters to the business model.
Salesforce Agentforce is not the answer
A short word on Salesforce’s own AI agent product. It’s essentially their version of ChatGPT. On paper. But. The marketing is bullish. The reality is slow, expensive to configure, and built on top of an old database. Salesforce sits on Oracle. Oracle is a fine database for a 2002 enterprise application. It is a poor one for a chat interface. Agentforce will get better. Today it is not where the leading edge is.
Avoid Zoho and Microsoft Dynamics
Both products look complete on the brochure. Their feature lists match the modern CRM playbook on paper. In production you spend months working through small holes. Flows do not behave the way the documentation promises. Many companies use both. I do not begrudge that. I do not recommend either.
The trap of the hyper-local insurance CRM
Insurance also has a category of software I have to address. Every brokerage owner asks about it within the first week.
Specialized insurance CRMs exist. Most of them are hyper-local. A system that works in Delaware has been built around Delaware regulations and Delaware carriers. A system that works in Switzerland has been shaped around Swiss carriers. The technical term is hard-coded. The process is baked into the program at the source code level. You cannot reconfigure it through settings.
If you find a hyper-local system that fits your country, your product, and your process, use it. Run a one-month pilot first. Push real cases through it. Watch where it bends and where it lacks functionality. The conversation only becomes adequate once you start hitting limitations: this is bad, that does not work, this part is fine. That is the moment you can decide what trade-offs you are willing to accept.
There is a bigger point here. Adopting a local specialized system rarely gives you a competitive edge. If many brokerages in your country use the same system, you have raised yourself to the local market standard and no further. My job, when a brokerage hires my firm, is to produce an edge. Do something competitors cannot do. Or do the same thing twice as cheaply, three times faster. Off-the-shelf local software almost never produces that edge. Platforms, configured around your process, sometimes do. That is why I lean towards platforms even when a hyper-local product looks attractive on the brochure.
The puppy excitement effect
A short detour. Everybody falls into this trap. I am no exception.
When you sit down with a vendor for the first time, the first impression is overwhelming. The demo hits all of your pain points. The salesperson knows what you are about to ask. The company employs sharp copywriters and a sharper pitch than the technology behind it. You leave the call thinking you have found the answer.
I call this the puppy excitement effect. I am in the middle of one with Attio right now. I know it.
The defense is to ask a different kind of question. Vendors will tell you everything that the system does. They will not tell you what it does not do. They will not tell you where it falls over, where it slows down, where the documentation lies. You have to ask that explicitly. You have to test the answers with real cases. The systems that survive a one-month pilot, with your real data flowing through them, are the ones worth buying.
A short note on AI agents and the data risk
Modern platforms ship with chatbot integrations and AI agents that can sit between the client and your CRM. Attio supports this well. So does Claude. The dream is that a client can chat with the system, ask questions, change a policy detail, and feel like they are talking to a member of your team. The technology is here today.
The risk is that we do not yet know how secure it is to put real client data behind these agents. There is a phenomenon called social engineering. It is the technique scammers use when they call a grandmother or a junior office worker and manipulate them into giving up a password or wiring a payment. We do not know how vulnerable AI agents are to the same trick. Even with restrictions in place, an agent that has access to private client information might be persuaded to release it through a clever prompt sequence. We will learn over the next five years. Some form of protection will emerge. Today, the right approach is to pilot AI agents on non-sensitive interactions and keep the sensitive ones inside human-supervised workflows.
We are at the earliest stages of this technology. The systems need a few more years to mature. The industry needs time to find the failure modes and document the limits. The brokerages that move first will learn the most. They will also pay the most. The brokerages that wait will pay less and arrive later.
Find your number this week
If you have read this far, your homework is concrete. Pick a single product line you sell. Run three questions through it.
- How long does an average sale take, from first inquiry to signed policy?
- How many sales does one agent close in a typical month?
- How much does the buyer remember about you a day after the policy is signed?
If the answers are minutes, hundreds, and almost nothing, you are at a 1 or a 2. The CRM line in your budget should be replaced by a developer line. Build a self-service portal.
If the answers are months, fewer than five a year, and everything, you are at a 9 or a 10. Hire better project managers and forget the CRM.
If the answers are weeks, twenty to a hundred a month, and your name plus a vague sense of trust, you are in the middle. This is where you need CRM. The next decision is whether you want simplicity (Pipedrive, Bigin) or extensibility (Salesforce, Attio).
Run this exercise per product line. A mixed brokerage will sit at different points on the scale for different products. That is fine. What you do not want is a single CRM strategy that pretends the whole portfolio sits at one place on the scale.
Closing
Most brokerages I meet want to start the conversation by asking which CRM is best. That is the wrong starting point. The first conversation is about the customer journey. The second is about the two businesses inside your firm. The third is about where each product line sits on the scale from commodity to bespoke.
Once you have those three pictures on a single page, the software decision becomes obvious. The implementation becomes survivable. Skip them and you will join the brokerages that pay for years.
If you want to know what your customer journey looks like in practice, where each product sits on the scale, and which tools fit the work, book a free audit.
We spend an hour inside your setup. We look at the products you sell and the deals in flight. We send back a plain-language report on what to fix first. Book it here: https://muncly.com/crm-audit/