Some Clients are Likely to Leave

Indeed, according to recent Vertafore research, a modest 3 percent increase in customer retention, achieved consistently over a five-year period, can produce a revenue increase of 15 percent. So improving client retention is an essential growth strategy.

Achieving those retention goals, however, can be a challenge. Identifying at-risk accounts — and spotting them in time to take action — isn't easy.

The Numbers

An analysis of current policy retention, compiled by Vertafore from 3,700 agencies and over 150 million policies shows that the median policy retention rate among independent agencies is just 82 percent.

Routine policy renewals aren't enough to ensure long-term success. Moreover, many agencies don't know their true retention rate. In most agencies, 15 percent of total policies are at high risk of non-renewal, and another 10 percent are at somewhat elevated risk. If not for the hard work of landing new accounts and/or expanding existing accounts into new product lines, most agencies would experience extremely small gains year-over-year — or none at all.

So how can brokers and agents know when a customer is thinking of canceling or choosing not to renew, and why?

Predictive data analytics may have the answer for independent agencies.

Data Science for You

Predictive analytics isn't new in the insurance distribution channel. For example, large carriers with data science teams have been using predictive analytics techniques for some time. But until recently, unlocking the power of these analytics to truly inform key business decisions and outcomes has been limited to those with the time and resources to invest in a dedicated program.

Now, the convergence of three factors — cloud computing, the accessibility and breadth of data, and artificial intelligence (AI) and machine learning (ML) technology — is putting predictive analytics into the hands of agencies of all sizes. And advancements in data science are allowing predictive analytics to go deeper than ever before and target the factors that impact an agency's growth and success.

When combined with sufficient volumes of high-quality, relevant data, predictive analytics is one of the most powerful insurtech tools for independent agencies in years. And one of the most powerful use-cases for predictive analytics for independent agents is policy retention prediction.

Early Warning

Predictive analytics use math, statistics, and large computational power to uncover relationships between data and a business outcome. In the case of insurance policy retention and risk predictions, the prediction model analyzes an extensive list of variables from six categories: policy information, agent profile, brokerage profile, end-insured profile, end-insured behavior,and external market factors.

AI and ML technologiesquickly ingest all this information andsharpen their understanding with each successive pass. In other words, these tools learn as they go.

The effectiveness of predictive analytics depends on the breadth, scope, and quality of the underlying data. At Vertafore, for example, we have a system that can predict the retention probability for a policy. We can also show which variables are driving the risk, such as product changes, price increases, or recent claims. Having that kind of information can help agents and brokers address issues 30 to 120 days in advance of the policy renewal date.

The Right Information at the Right Time

When it comes to client retention, modern predictive analytics can — and will — have a strong impact on agency success. Retaining at-risk policies must be a priority — and predictive analytics provides the information that empowers agencies to act. By knowing which accounts are at risk, agents can focus their time and attention on the places that will have the most impact.

As an enabling technology, predictive analytics has the potential to radically transform the fortunes of insurance agencies by predicting at-risk policies, accounts, and relationships. Agents and brokers may never have a Magic 8 Ball for their business — but for anyone who wonders about the value and use of predictive analytics in the future, "All Signs Point to Yes."

---

Scott Ziemke is the director of data science atVertafore.

---

The original version of this story was published onThinkAdvisor.