Technology that automates mortgage processing is the talk of the industry. No doubt you’ve read any number of vendor promotions about these solutions’ features, benefits, and ROI.
But how do you know which claims are realistic and which are merely marketing hype?
From Overpromises to Reality
Most recently, some vendors assert that mortgage industry companies can deploy automation solutions with limited IT investment and effort. But that’s highly unlikely. Given the complexity, high number of variables, and uniqueness of each company’s processes, you shouldn’t put much faith in claims of plug-and-play automation solutions.
These claims – and others like them – are creating cynicism in the mortgage industry. Companies are becoming wary of vendors, viewing these claims as overpromises on how leading-edge technologies such as AI and machine learning can help their business.
So how can you know what to realistically expect from automation solutions? What’s the best way to parse vendor claims about the value these technologies can deliver?
Consider the following three issues when evaluating the vendor promises.
Variability of Processes, Workflows, and Documents
No matter what its role in the mortgage lifecycle, every company uses different documents and rule sets. There are no industry norms to map a company’s taxonomy to commonly used forms and documents. A company in mortgage insurance finds some data fields to be critical, while a company in lending may rely on a different set of data fields.
A solution that can be used with “no preparation” takes no steps to account for this variability. By definition, it would be unable to map data fields and classes to your company’s data fields or classes. Think of it as the typical “garbage in-garbage out” scenario. The solution can be up and running quickly, but it won’t reflect your business processes, workflows, or workloads.
To realize true value, you need to map your use cases, documents, and field names to those of the automation solution. Highly advanced mortgage automation solutions can perform much of this mapping automatically by auto-generating the mapping files.
By choosing one of these solutions, you can reduce the timeline from solution deployment to ROI. Depending on the use case, auto-generation features can help mortgage companies move an automation solution into production within as little as four to eight weeks.
The Right References
Whenever you evaluate an automation vendor, ask to see references from some of their other customers. Reach out to those customers and find out how long it took to get their solution to full production.
Also ask how many of the customer’s processes were automated. Different solutions provide different levels of automation, resulting in widely varying ROI numbers. For example, a simple OCR solution can be launched fairly quickly but it includes no automation capabilities.
High Data Pass-Through Rates
To better understand how well a product could serve your needs, ask the vendor about its pass-through rates for documents. High pass-through rates mean that users don’t often need to stop the data verification process to review data accuracy and completeness.
Advanced solutions can offer 90% pass-through rates, meaning that bots can handle nine of every 10 data verification instances in the workflow. More typically, older solutions offer pass-through rates of just 60%. If humans need to check 40% of all data to review accuracy and completeness, that’s only nominally better than manually entering all of the information.
Commitment to Automation Innovation
Choosing the best solution to automate your mortgage processes requires your time and effort. It can be tempting to go with the vendor that promises the easy path, but you’re unlikely to get the results that will win respect from your COO.
Instead, take the time to look for a vendor committed to performing ongoing research and developing new technology solutions that will keep you ahead of your competition. It’s a good sign when a vendor is a leader in its industry – having its executives chair technical conferences or publishing research, for example.
Does the vendor seek new ways to use advanced technologies such as AI to solve business problems? Or is it just building fancier apps around traditional OCR technology? Does it have the technical expertise to use AI-based automation methods to understand your problem? Can it extrapolate existing AI solutions to your specific needs? That’s the best way to help you get an automation solution into production in the shortest possible time.
A provider with these characteristics can create solutions that will help you deploy mortgage automation quickly and productively – before your competitors do. That’s not hype – it’s a realistic goal for your business.