How Web Analytics Improves Insurance Businesses

  • Ethereum
  • MetaMask
Scan to Donate Ethereum to 0xe95BF07c484dcC33Af16a57A7a1cB725906bc57D

Donate Ethereum to this address

If you find the blog useful and would like to show some support, a little donation of Ethereum would help to keep this site alive. Appreciate it!

Donate With MetaMask

Donate ETH Via PAY With Meta Mask

In ancient times, insurance contracts were used by merchants to avoid suffering huge losses should their goods be stolen or lost at sea. Fast forward to 2018, much more sophisticated insurance products have been created and sold worldwide. There is no lack of innovation in the product department, as insurers come up with all kinds of insurance products to provide better coverage in every aspect of our lives. Great Eastern's Great Mozziesafe

Great Eastern’s Great Mozziesafe

Having said that, most traditional insurers are having a tough time getting into the digital part of things. Business model has remained the same for a long time. Many still relies on agents to secure sales as their main source of revenue. It’s not for lack of trying. Many traditional insurers actually do sell insurance products online, but the result is often disappointing in comparison to what their sales force is bringing in. Since most activities and purchases are taking place online, does that mean Insurtech companies will eventually make traditional insurers obsolete? Yes and no. In the long run, if traditional insurers fail to adapt to the world of digital, the inefficiencies in various areas will put them far behind those who manage to master the digital ways. But if they are able to find a way in, the advantage of having a sizeable sales force will act as a barrier for Insurtech companies. The question now is, how can traditional insurers protect themselves from an eventual doom? The key lies in how well they can support their sales force. So, how can they equip the sales force with an unfair advantage over those who generate leads using cold-calls and roadshows? Today, you are going to find out how web analytics improves insurance businesses.

Characteristics of an insurance website

Similar to e-Commerce companies, the main goal of an insurance company is to generate as many sales as possible. But instead of selling things online, most insurance websites are focused on generating leads. For an e-Commerce website, cross-selling is crucial in enticing customers to make more purchases.Amazon cross-selling   For an insurance company, it is difficult enough to get users to buy one product, let alone a second one. Reason being, most insurance products are financial commitments that last for an extended amount of time. Users tend to do a lot more research before deciding on making a purchase. Therefore, our effort should be focused on securing that one sale instead of diluting the attention of the potential customer. We can all agree that our goal here is to convince users to leave their contact details (lead generation) for our agents to call them back. And in order to do so, we first need to understand the needs and wants of our users. Insurance product isn’t something that I wake up one day and decide to buy. There must be an intent generated from somewhere, for example, losing a close friend to an accident. In that case,  it is impossible to obtain such sensitive information from strangers in the offline world, and we certainly do not encourage agents going around asking everyone “did you just lose someone close to you?”. Remember when we said, most activities and purchases are taking place online? That includes the due diligence carried out with purchase intents. On a normal day, no one would casually go to an insurance website just to view the latest offering. So for someone to make a visit to our site, she must have a decent purchase intent, or at least a certain amount of interest.

Ways to increase sales

There are 2 ways we can contribute to our goal:
  1. Generate more leads
  2. Assist agents in providing personalized and timely services

1. Generate more leads

This is rather straightforward. Probability tells us, the more leads we have, the higher the chance of making a sales. With a large enough number of leads, it is with certainty that sales number will go up. In order to generate more leads, there are a few things we can do:
  • Lead form optimization
  • Campaign optimization
  • Improve user experience

2. Assist agents to secure more sales

Ultimately, the sales force is our main source of revenue. Generating more leads is akin to increasing awareness, where we hope to land more customers with the help of probability. Another way we can ensure higher sales number is to improve the closing rate of insurance products by empowering our agents. In order for us to do that, there are a couple of things that would help:
  • Understanding our customers better
  • Personalization in a timely manner

Wait… Are insurance agents indeed still relevant?

As much as we are in the digital era, people trust each other more than robots, and we still very much desire the human touch. For most big ticket items, people are generally reluctant to make the purchase online. To them, it is a big financial decision that cannot be made hastily. Back in the pre-Internet times, people would go down to physical establishment of businesses to check out the products before making a purchase. Fast forward to today, there are tons of information and materials online, you can expect one to do all the research she needs without leaving the comfort of her chair. Does that mean there is no way our agents can participate in the buying process? On the contrary. Think about it. Do you make a purchase based on just one review you found online? No. We will continue searching for more and more reviews, as if we are intentionally looking for a particular one that tells us “This sucks!”. So why are we exhibiting such behavior? That is because, we still can’t fully trust what we see online. Our minds work in the way that, if we see enough good reviews, we will have more faith in the product and eventually make the purchase. That is also why online stores are so huge on the practice known as “review brushing”. By publishing enough good reviews, they hope to trick us into thinking that their products are worth our money. This is where our agents come into play. As the authority figures, our agents exert the most influence on potential customers with their expertise and advice. If that’s not enough, pair it with the human touch.

Useful web analytics insights

Now that we have clearly defined our goals and means to achieve them, let’s delve deeper into the inner-workings.
  1. Campaign optimization
  2. Lead form optimization
  3. Improve user experience
  4. Understanding our customers better
  5. Personalization in a timely manner

1. Campaign optimization

The why

Marketers often work within a certain budget, which means, we cannot simply acquire more users by spending more. This is why campaign optimization is important. We need to maximize the budget by increasing the effectiveness of our campaigns. Instead of getting 100 users with $1,000, we want to get 150, or even 200 users with that same $1,000. In order for us to do so, we need to find means of improving our campaigns. Traditionally, marketers build campaigns based on gut feels. If we launch enough of those campaigns, one is bound to become a hit. In the past, there is no way to tell what works and what does not. All we know are how much we spent, and how many leads we have gotten. With web analytics, we seek to quantify the performance of all campaigns, allowing marketers to better understand how effective were past campaigns, and to build future campaigns based on the lessons they have learned.

The how

Knowing just how much we spent and how many leads we have gotten is not enough. We need to know more. We need to know if there are lost opportunities that we could have otherwise captured. Did my campaign excel in generating awareness? Did my campaign excel in converting interested parties? Below are some great insights that we can rely on:
  • Click-through rate
    • Out of all the users who have seen our ad, how many went on to click to our campaign landing page?
  • Campaign landing page visits
    • How many users did we attract to our campaign site?
  • Conversion rate
    • Out of all the users whom we have attracted, how many went on to convert?
  • Bounce rate
    • High bounce rate often indicates the disparity in expectation of what they see on the ad, and what they find on the page itself
  • Campaign performance broken down by traffic sources
    • Of all the campaigns, which platform/channel performs the best in terms of effectiveness and cost efficiency?
To find out more, here’s an article I have written to further explain what each insight represents and what are the different opportunities for optimization.

2. Lead form optimization

The why

This is pretty self-explanatory. In order to generate more leads, we need more lead forms to be completed. Even with unlimited budget, there are only so many users we can bring on to our website. Eventually, we will need to find some other means to increase our lead numbers. One such way is to optimize our lead form. Take for example, for every person who landed on our lead form page, there is a 10% chance that she will leave her contact details behind. That means, if we have 1,000 users, the expected number of leads we will have is 100. What if, somehow, we are able to increase the chance to 90%? With 1,000 users, we will have 900 leads! Realistically, achieving that kind of number is impossible for various reasons, but there is nothing stopping us from improving the “10%” into something better.

The how

In order for us to start lead form optimization, we first need to identify the issues and areas for improvement. To help us do that, we need the following insights:
  • Conversion rate of all products
  • Step-by-step drop-off rate

Conversion rate of all products

Reason is simple. Sometimes, the problem does not lie with the lead forms. As much as we love all our kids, there are still some that are not as perfect as we would like them to be. Therefore, it is important to identify the products that are doing extremely badly, and remove them from the lead form optimization process. These are data we cannot rely on, and there should be a whole new investigation into why these products are performing so badly as compared to the others.

Step-by-step drop-off rate

Companies love to concentrate on overall conversion rate, but that tells us nothing concrete. In order to truly understand what works what does not work, we have to break down the whole lead form process into numerous steps. If the form has 5 steps, we want to know which of the 5 steps are responsible for the high drop-off and fix them accordingly. Common issues for high drop-off rate are:
  • Form too long
  • Bug/Glitch preventing users from filling up the form or proceeding to the next step
  • Form requiring sensitive data that the users are uncomfortable providing
  • Form not intuitive and lacks clear instructions
  • Poor UI/UX
The list goes on. To recap:
  1. Identify which steps are responsible for high drop-off rates
  2. Investigate and determine the root cause for the high drop-off rate
  3. If it is a technical issue, fix it. If it is user experience issue, perform A/B testing

3. Improve user experience

The why

There are thousands of reasons why someone would leave your website without making a purchase or, in our case, leaving her details. And what this means is that, there are thousands of opportunities for us to improve our businesses! I might be exaggerating a bit, but that’s not too far from the truth. First and foremost, everyone enjoys a website that has an updated look and feel. But we might be thinking to ourselves, as long as the website provides the necessary services and information, we’re fine. That is so not true. An outdated website gives off the vibe that the company just simply cannot be bothered and for smaller businesses, might even feel a bit dodgy. No matter how great the contents are, if you do not have an audience, there is no one to preach to. There are other aspects of a website that look inconsequential, but can instead play a huge part in determining whether an user stay or leave.  Some examples are banner type, button placement, wording, content length, font size, ease of use, etc.

The how

A google search will return millions of results on how to improve user experience. There are some fundamental rules that should not be broken (legible font size, minimize number of clicks to conversion, avoid over-lengthy contents, etc) but for the most part of it, no one really knows. If your website has not met the basic requirements, fix it. If you are past that, then it’s time to make use of A/B testing to further optimize your website and challenge existing assumptions.

A/B testing: EA SimCity

EA was gearing up to release the new version of SimCity and wanted to optimize the microsite, where the users can pre-order the game. EA SimCity Microsite On the microsite, they had a big banner showing a promotional offer in hope to lure more pre-orders. Somehow, the result was not showing, as the number of pre-orders did not experience a significant increase. This is when they came up with a hypothesis – the banner was taking up too much space, pushing the CTA (Call-To-Action) buttons down the page, resulting in less pre-orders. They made use of A/B testing and put the hypothesis to the test. EA SimCity Microsite A/B Testing On the right, we can see that the banner was completely removed, resulting in the CTA buttons moving higher up on the page. The result? Surprise, surprise! The variation without the banner obtained 43% more pre-orders! Without A/B testing, no one would have expected that the promotional banner was working against them. For gamers though, this might not come as a surprise. Before the launch of great games like SimCity, gamers would have already heard news from forums and peers. After doing their own research, they would have long come to a conclusion whether to pre-order the game or not. Offers are great incentives, but they would have heard from their friends or gaming sites before coming to the microsite. The only reason why they would come to the microsite, is to make a purchase and nothing else. If companies knew their target audience well enough, they could have saved on a lot of unnecessary costs. Alas, that is usually not the case, hence the need for A/B testing to test out every hypothesis possible!

4. Understanding our customers better

The why

We want to equip our agents with better understanding of their clients, in order to further boost their image as authority figures. Like we have mentioned previously, people crave for the human touch. Even though that’s the case, we are not ready to place our trust in just anyone. As much as I want to, I can’t trust someone who knows nothing about me, my wants and my needs. This is where we want to help our agents, to provide them with information on their potential customers, for them to build the much needed trust. We want our potential customers to see our agents as their trusted friends. Friends that they can turn to for advice and help.

The how

In the web analytics world, there are 2 forms of personalization we can provide:
  1. General personalization
  2. One-to-one personalization
In this case, we will start off with general personalization.

General personalization

General personalization is when we do not know the true identity of an user on our website. This happens to be the most common scenario, as most users do not have the habit of logging on to a website unless there is a need to. In such case, even though we can’t exactly tell who is doing what on our website, we can still tell the user web behaviors as a collective. If enough people are doing the same thing, we can make an educated guess that others might also be interested in that same thing. Here are some useful insights that we can provide to our agents to better their chances against potential customers:
  • Product popularity
    • “This is one of the most popular products we have and here is why..”
  • Trending products
    • “Recently there is a huge surge of interest in product X, you should really look into getting one as it can help you with..”
  • Trending articles
    • “We realized more and more people are worried about XXX, so I thought you might want to consider..”

5. Personalization in a timely manner

The why

If an agent is convincing enough, all it takes is one session for the customer to make a purchase. Hence, being the first to get to a potential customer can make all the difference in the world. How fast should we be? No longer than 7 days. Users generally have short attention span and might have move on to something else by the time we get to them. If a particular user had urgent needs, she would have gotten things settled way before the 7 days mark. Therefore, 7 days is the longest duration we can tolerate in terms of serving crucial insights to our agents. If we can reduce the information lag time down to a day, that would be fabulous! The advice is, start small (7 days) and slowly work towards serving near-real-time information (1 day or less).

The how

With the knowledge of who we are dealing with, time is of the essence in securing the sales.

One-to-one personalization

One-to-one personalization can only happen when we can identify the users coming on to our website. When users login on our website, we would have their personal information stored securely on our internal databases, thus, allowing us to link all their web activities back to them. Previously, I have covered the topic on how we can track individual users on our website, so if you are unfamiliar with it, do check it out. With that in mind, here are some useful insights:
  • Products viewed in past X days
  • Articles viewed in past X days
  • Tools used in past X days
  • Promotions viewed in past X days
  • FAQs read in past X days


At the end of the day, traditional insurers will still rely heavily on their sales force to drive the revenues. We should never be looking to replace our agents with emotionless robots, but instead, make use of technologies to help them do their jobs better.
About Zenny Tan Zhong Ming 32 Articles
Let's connect on LinkedIn ( ).

Be the first to comment

Leave a Reply

Your email address will not be published.