Forrester highlights CSS Corp and its AI-based text analytics platform, among the overview of 30 providers, that improves customer insights, acquisition, service, and retention. Customers can manage governance risk control (GRC) and leverage RPA for efficiencies. With numerous providers varying in size, functionality, geography, and market focus, how do customers choose the right service provider?
The Now Tech: AI-Based Text Analytics Platforms, Q1 2020 is a ready-reckoner in helping customers understand the value from the platform provider; the report guides buyers to select a vendor based on size and functionality.
Customers’ business insights (across operational, financial, risk management) such as strategic, tactical, and operational can come a full circle by tapping information hidden in unstructured data.
Leverage Unstructured and Semi-Structured Data
Unstructured data can range from voice-of-the-customer (VOC) communications, voice of employees (VOE), emails, and social media. It’s not just unstructured data but also the semi-structured data stored in documents, reports, and websites. Other than customer insights, all enterprise processes, such as supply chain and GRC, rely on these data types. And rightfully so, enterprises are investing in text analytics to garner insights from customer and employee behavior and also unearth efficiencies from process automation and managed GRC.
Various Functionality Segments
Before delving into where CSS Corp. stands, let’s look at the market, which comprises three segments with different capabilities.
The first segment involves platforms that support people-focused applications for marketers, researchers, CX, and HR experts. The concentration is on VOE, VOC, brand management, and market research, and they require out-of-the-box connectivity to data sources such as survey platforms. Natural Language Understanding (NLU) is a crucial capability with these platforms as they unearth sentiment, emotion, effort, and intent from unstructured data.
Forrester recommends buying instead of building platforms unless customers have specific or one-time requirements like embedding customer sentiment scores into an existing analytics application. Or, in use cases that lack ontologies.
The second segment entails platforms that support document-focused use cases. This segment involves document classification by customers or topics and segregation by risk for security and compliance. Other use cases include e-Discovery and contract analytics. A unique capability involves document clustering and summarization; some providers are introducing basic image recognition.
The third category, general purpose, is a broad one involving a combination of the above 2 segments and covers multiple use cases. These platforms are an attractive proposition for customers who want to realize efficiencies across numerous use cases.
The report rates (Scale: None, low, moderate, high) the functionality of the 3 segments across various capabilities. It includes NLU, supervised machine learning, deep learning, streaming analytics, speech analytics, taxonomy creation, and the like.
Where Does CSS Corp Fit In?
The Forrester highlights CSS Corp among mid-size providers (revenue-wise) for the general purpose segment with ontologies across retail, cloud infrastructure, and tech-support services. The report noted the general purpose segment’s high functionality in natural language understanding, linguistic rules in multiple languages, and supervised machine learning. With knowledge of the 3 categories and how they fare on various capabilities. Business insights pros need to make decisions based on their enterprise needs.
Considering the entirety of a comprehensive functional text-analytics solution, NLP is just a part of it. Business insights pros need to think hard. Think twice before going for an open-source NLP engine, and before building capabilities on their own.
Decide Based on Requirements
The report clearly states that unless customers have some highly specialized needs or one-off requirements like extracting customer names from a text column during an ETL process. They need to opt for a turnkey solution or platform. They need not devote time to build unique capabilities. Forrester recommends that customers request proof for out-of-the-box capabilities. It includes multilinguistic features, source data connectors, and domain-specific ontologies.