July 28, 2015
How Quote Data Can Optimize Pricing
Insurers must follow the lead of airlines and retailers and use quote data to fine-tune prices and features based on each customer's situation.
Retailers do it. Auto dealers do it. From wholesale parts suppliers to craigslist sellers and kids with lemonade stands, everyone knows that if you are going to take the trouble to sell something you should sell it for its full value. Many insurers, however, are stuck within semi-fixed pricing models that don’t allow them to capture the most profit they can from each policy.
Today, insurers can change that because they have the ideal vehicle to help them optimize pricing and improve their margin — quote data. Quote data, when analyzed and tested on a continual basis and kept within the boundaries of the rate filing, can yield dramatic insights into purchase patterns and price tolerance. Plus, optimizing price with quote data is an analytics concept that will excite nearly everyone in the organization.
Why should insurers consider using quote data to modify pricing or products?
Insurers have actuarial models and underwriters who understand the market, plus they have rate plans that have already been filed for specific products. Quote data is ripe with excellent, relevant insights. The reason we see Google, Overstock and Amazon dipping into insurance quoting is because they grasp the potential in marrying purchase pattern data with price testing. For insurers, quote data tested against purchase patterns is a gold mine waiting to be tapped. What do insurers have to gain?
- New data yields new insights and can result in new decisions. (The ability to analyze multiple risk factors, even at the quote stage, is improving.)
- Insurers can decide to charge more based on what they learn.
- Insurers can decide to go after lower-margin, high-quality business.
- They can go after low-margin, high-efficiency business.
- They can identify business that they don’t really want.
- They can answer the competitive threats of new entrants that are poised to capture an increasing share of the market.
Is optimization the right way to make decisions?
For the most part, the days of “from the gut” decisions are over. Human brains are predictable enough that they can be mined for decision data and yield well-patterned insights across similar individuals with similar decision patterns. Amazon, Pandora and Google can effortlessly predict a consumer’s next areas of interest and likely purchases without the individual ever telling them anything. The messages we receive from nearly everywhere are “optimized” because they are proven to most likely produce a positive reaction from us. Optimization is data science that works. Pricing is the second step of optimization; it concerns itself with how much a certain type of prospect will pay at that point in time through that particular channel.
As an example, consider a couple purchasing a boat two days before Memorial Day weekend. They are in the showroom using a quote aggregator on her mobile phone. They may be willing to pay more for insurance because of the need to move through underwriting quickly. Quote data over time may also prove that two boater certification questions need to be added to the quote process for first-time boat purchasers to keep the product profitable, either through adjusting price or filtering out applicants.
Insurers have a leg up on traditional online retailers because prospects do tell us something about themselves before they purchase, to get an accurate price. This kind of pricing optimization isn’t limited to online purchases. It can be done through agency channels and even through traditional direct mail. But the best data accessibility and ability to test is through website and mobile channel metadata.
How insurers optimize price — finding opportunities among the limits.
There are several areas for insurers to consider when optimizing through quote metrics. First, insurers should be tracking every bit of data and metadata surrounding the application. Every submission document has the bits of a consumer story to tell. For example, how many days is it until renewal? Is a client making a last-ditch effort to get better auto pricing with you before turning elsewhere? Is a prospect shopping around in the last week before her home policy auto-renews? How many apps are coming through a particular channel in a particular day? All of these questions and many more could lead to pricing revisions based upon consumer behavior in the application process.
Next, insurers should become highly adept at A/B testing. Consider variables as levers and raise and lower them to reach their limits, then continue monitoring and adapting. For example, begin with quote take-up rates on all submissions. Insurers should consider testing the limits available to the market. Do take-up rates improve when limits are raised?
Website metadata can be informative in this regard, as well. What pages do consumers visit and when? Is there a standard path for the person who seems to rush through shopping, quoting and purchasing? Can the insurer raise the price for those who seem to decide quickly in their first visit and lower it for someone who has come back to the site repeatedly, conceivably price shopping?
There are hurdles, however. Price testing must be done within the boundaries of the filing and the specific products. Some pricing changes may be able to be implemented immediately, but many will need to go back through the filing process. Pricing always has to happen within the regulatory box, so what is possible in testing may not always be feasible in pricing.
But pricing optimization is only one part of the A/B testing equation when it comes to quoting. Quoting data can also be used to more finely tune risk factors and their relationship to take-up rates and claims. This kind of profit optimization is just as critical as pricing optimization, and it requires no regulatory refiling. It is data that can be fed back into actuarial models and may ultimately be useful when used in conjunction with mobile telematics data and a host of other data sources. Even if an insurer planned no immediate repricing of products, the ability to understand price tolerance based upon other quote factors (e.g. age, income, take-up rates, property value) would be helpful in the development of new products.
The nuts and bolts of pricing optimization will vary with each insurer’s unique quote process and current market. But the promise it holds is not only a better overall margin per policy, but also the potential to grow volume through unexplored insights and the opportunities to deeply understand individuals, groups and their motivations to purchase insurance.
Consumer data analytics is here to stay. The value in quote data is continuing to grow.