Read on FFnews 

Clearspeed voice analytics represent a quantum leap in risk accuracy for insurers, says Jules Ehrlich, its Chief Product and Strategy Officer. Here, he explains how it’s achieved those targets – and at scale


Voice has a mixed reputation in insurance. Caller authentication technologies have been well-received but voice stress used to assess claims risk less so, for many reasons – including questionable accuracy.

Clearspeed voice analytics are nothing like the voice technology of the past. There is no use of voice stress or lie detection, no biometrics, no emotions, no inflections or hesitations, no natural language processing. This is new. And so much better.

The core technology uses automated questionnaires to ask exactly the same questions that claimants are asked now, such as ‘was someone other than you operating the vehicle at the time of the accident?’, ‘does your claim include property damage not caused by the fire/flood/storm?’, ‘do you know where your lost item is now?’.

In the past, it was nearly impossible or at least very cost prohibitive, to assess the accuracy risk of the answers that claimants provided. Today, insurers can leverage big data and AI capabilities with historical and even real-time information for more insight, but that can bring complexity, increase false positives and may not provide the validation needed to process claims quickly and accurately.

As speed is increasing, so is fraud, especially in these challenging times. Clearspeed brings new and unique data to risk assessment. And with greater than 97 per cent accuracy, it’s helping insurers increase the number of claims they can process straight through with very high confidence by leveraging Clearspeed’s automated questionnaires, either from within their call centres, or as an instant follow-up to ENOL/FNOL or other methods of claims submission.



“It’s impressive how something so simple can have such an immediate, positive impact”

Tara Shelton, Head of Desktop Solutions and European Development, The Cotswold Group



How it works

With Clearspeed voice analytics, insurers deliver short, automated voice questionnaires to their claimants, connecting with them by phone, app, chatbot or web. Questionnaires can include both yes/no and open-ended questions. Insurers can choose from the Clearspeed library of questions that have been successful with other insurers or create their own specific questions.

There are a few essential features to understand with Clearspeed.

  • Automated delivery means that the questions are asked the same way, in the same voice, without judgment, every time. And it offers consistent around-the-clock availability.
  •  Any language means that the questionnaires are language- agnostic, making it very easy to accommodate every claimant.
  •  No personal data collection means Clearspeed does not need to know who the claimant is or store any personally identifiable information (PII) to provide a risk assessment.

There is no bias in the automated delivery, nor in the scoring of responses. You can think of it pretty much like a metal detector at an airport when you go through departures – it doesn’t care who you are or where you’re from, just whether you’re carrying anything made from metal. Clearspeed takes the recorded claimant responses and removes all noise from the voice signal other than the universal voice characteristics that we know to be associated with risk. It then checks for the presence or absence of these voice characteristics and scores them on a risk spectrum of low to high.

Low-risk claimants can move forward quickly, with a straight-through or expedited process that gets them paid faster. But this does not mean that higher risk claims are denied. Clearspeed simply provides a low to high risk alert, just like the metal detector. Getting flagged doesn’t mean that you don’t fly, it just indicates the need for someone to follow up. If you’re clear, you’re good to go. I

In the same way, Clearspeed is used to clear low-risk people very quickly and with a much higher level of assurance than insurers have been accustomed to. But it is never used to deny a claim. On follow-up, insurers can find that, even though the alert was warranted, the claim is indeed valid and gets paid. In a smaller number of cases, that follow-up determines there is actual fraud.


“Clearspeed brings a new and highly effective approach to screening for fraud at scale during claims processing”

Clare Lunn, Head of Counter Fraud, Markerstudy


Optimising the claims process

Insurers have many opportunities to benefit from deploying Clearspeed within their claims process. At the initiation of the claim, they are usually less concerned about pinpointing the risk in each response and instead are looking for a red/green indicator about whether to expedite or follow up. Here they will either use a low-risk Clearspeed result on its own to process the claim or they will add it as a contributing data point to their AI/data analytics to make a decision.

Further downstream, an adjuster or SIU investigator will relish the precise risk data that Clearspeed provides, to inform their next steps, confident that it is not a false positive or unproductive alert. Clearspeed is effective, either standalone or fully integrated within an insurer’s existing processes and platforms, with results delivered via API. Since there is no data to gather, no AI to tune, etc, implementation is comparatively short and generally takes less than a month.


Supporting the ecosystem

Clearspeed is a highly secure Cloud-based service that operates worldwide and contributes effectively to the insurance ecosystem. It can integrate easily across all ecosystem platforms, from automated communications via Hi Marley, the intelligent communication platform for the insurance industry, to fraud and other analytics systems such as Shift or Friss, and to claims management via Guidewire or similar products.

Insurers often use Clearspeed data to enhance the accuracy of what they are seeing from their predictive analytics, either inserting the automated questionnaires at the very tip of the process or further downstream, once they have all the supporting data.

Some insurers actually apply a red/green short form questionnaire early on and more precision-based questionnaires when they require
more precise risk data.

In either scenario, the data flows easily across the ecosystem. Powerful and informative The reason that the assessment is so accurate (greater than 97 per cent) is because our initial use case was all about accuracy. We needed to help our military screen for risk at scale in austere, remote combat environments.

Lives were on the line. Getting it wrong was just not an option.

So, we had to iterate over many years until no high risk was missed and false positives were miniscule. Once an insurer begins getting data from Clearspeed, it will find there are many ways to use the questionnaires to drive different and impactful outcomes, which:

  • Increase the percentage of claims that go straight through… quicker pay, lower claim handling time, higher assurance, higher customer retention
  • Reduce your false positives
  • Reduce soft fraud
  • Increase walkaways – insurers are seeing very strong numbers here
  • Reduce investigation downstream with precise risk data so investigators can easily focus in on where the claim risk lies

It also means insurers can potentially rethink parts of their process. For example, if Clearspeed provides a low-risk score on a low-risk travel claim, maybe you won’t bother incurring the time and expense to request a receipt, review it, etc.

In fact, you will absolutely delight your policyholders. And, even if you’re already doing this today, Clearspeed will provide a new level of assurance for those decisions. Expedited and automated claims settlements also naturally means lower operating costs.

Together with our customers, Clearspeed is redefining risk assessment with unparalleled accuracy, speed and simplicity.

Clearspeed will be at ITC Vegas 2022 (September 20-22) in booth 3159, where visitors can learn more about how it’s helping process claims faster, reduce fraud, lower false positives and deliver a better customer experience. To find out more, go to