More Flexibility through API
Process Automation in Customer Communication
How do companies stay agile enough in their document and output management to meet increasing customer expectations for speed and quality?
The following article highlights the various aspects of modern omnichannel customer communication and explains the growing role of APIs in this context.
Is the world of batch-driven document processing soon over? A world in which customer inquiries are collected as letters or e-mails and processed over several days or weeks? Analyses have shown that the average "life cycle" of a business transaction of this type is three weeks. Are such long processing times even accepted today?
The fact is that corporate customer communication is currently undergoing rapid change; not only because there is a diversity of analogue and digital channels, but also because the diversity in document and output management is increasing. Typically, for example, information about one and the same business transaction is exchanged via different media. Today, the consumer or customer wants to have the choice of which channel to use to contact their insurer, bank, telecommunications provider and other business partners. If, for example, he is on the road and wants to read the message on his iPhone or tablet - then the corresponding document is required in HTML format for optimal display. On the other hand, if the document issensitive or has a valuable tactile value, such as the purchase contract or insurance policy, then pressure still plays a role. One would like to file a valuable document in the main folder, sometimes this is still required by law.
Communication Today Is No Longer One-Dimensional
A high degree of "suppleness" in customer communication is therefore required in order to be able to switch quickly from one medium to another if necessary or to be able to "serve" several channels simultaneously. In general, expectations of quality and speed are rising - the latter is certainly also the result of experience in online trading, where it is now almost taken for granted that the product ordered will be shipped the same day. The time factor is playing an increasingly important role in customer communication, because short response times ultimately also mean a piece of quality in general. Of course, there are gradual gradations in the individual communication media in terms of processing time: e-mail is certainly more accommodating than video and audio messages. And of course, various socio-cultural factors such as age, education, IT affinity, language, etc. also influence communication behaviour.
Added to this is the "fragmentation" in communication. For example, in order to repair damage to a building, you need a craftsman's workshop. The communication required for this could be as follows: First, the responsible insurer must approve the settlement. To do this, it would be sufficient for the injured party to photograph the object with his smartphone and send the image file together with important data (insurance number, time of damage, address etc.) to the insurer via a messenger service (WhatsApp, SMS etc.). The following maxim applies: The earlier, the better, because once the basic approval has been given by the insurer, the injured party can already take care of commissioning the repair company.
The details of the settlement could then be clarified at a later date between the insurance company and the injured party - whether by electronic means (e-mail with attachment, web portal, etc.) or by traditional mail. In either case, the injured party would benefit from a quick settlement of the claim and ultimately the insurer (image).
This example illustrates two fundamental aspects of modern communication: firstly, quality plays a decisive role; secondly, despite all digitisation, both analogue (paper-based) and electronic channels will retain their authorisation for some time to come and exist in parallel; even the "good old" fax will not disappear any time soon. The challenge is therefore to take into account the omnichannel character in both input and output for all conceivable application scenarios.
More "Smoothness" with API Services
The question remains how to achieve this agility - and without a lot of effort. A basic requirement is that the IT systems used in a company have so-called open programming interfaces so that applications, both internal and external, can operate efficiently and smoothly with each other, i.e. exchange data, can be easily combined (for example, via a cloud) and can be easily integrated with each other. In principle, this is all about "collaboration" within the so-called API economy, which offers companies undreamt-of opportunities to flexibly adapt or expand their business areas to market requirements without having to invest heavily in the development of new software solutions every time a change is made. This ability to integrate new applications, software solutions and services into existing IT structures at will will will help companies add value, especially in their customer communication. Suddenly, completely new approaches and scenarios are possible in document and output management.
Just think of the integration of a translation service available on the web (DeepL, Google Translator etc.). If, for example, a Spanish customer asks his German insurance company for a new quote for his life insurance, he can send his message in his native language - whether as a WORD document, e-mail attachment in PDF format or as a WhatsApp message: the text or audio document runs through a translation app used by the insurance company (whether developed in-house or an external API service) and is then automatically sent to the clerk in German.
Or take the topic of accessibility: Since autumn 2018, authorities and public sector organisations have been required under EU law to offer all their communications barrier-free. According to this law, the content must be "perceptible, operable, comprehensible and robust". Users with disabilities have to be able to navigate through the pages and perceive and understand the information provided. Against this background, how nice it would be if an API service could be used to automatically prepare the content as an audio file (for recipients with visual impairments) instead of just making it available as a PDF or HTML file?
Identification by Telephone through Chatbots
Some scenarios are not so new and have become part of everyday business life. In private health insurance, for example, apps for recording doctor's bills have now become widely accepted: Instead of collecting the individual paper receipts and sending them to the insurer by post at the end of the month (including adding a cover sheet and manual signature), the insured simply photographs each receipt immediately and sends it to the insurer via the app.
In countries such as Denmark, they are already further ahead - no wonder, since the Scandinavians have always been among the pioneers in terms of digitalisation. Here, insurance companies are already using chatbots in their call centers, which are able to request the necessary information from the caller to identify the customer's request, for example: Are you already a customer? What is the policy number? The chatbot thus carries out a telephone authentication and ideally, in the ideal case, i.e. routine tasks such as the request for a new offer, automatically triggers a response e-mail with the offer as a PDF attachment immediately to the caller. In the meantime, chat bots have become so intelligent that today 70 to 80 percent of all processes typical for an insurance company can be handled in this way (changes of address, submission of damage reports, etc.) These self-learning AI systems can even deal with customer-specific features such as unusual intonation, special expressions and voice pitches, etc.
Clearly: The processes in customer communication today are so complex that the applications and software systems used are increasingly made up of components from different manufacturers and communicate with each other via web services. Incidentally, this is also a reason why many companies are increasingly turning to the cloud.
Whether one wants to admit it or not, the business world is moving towards the API economy, as "rigid" batch processing is increasingly being displaced by agile or transactional communication. Ultimately, this means that case processing time is becoming an important indicator of customer service.
APIs Are Business Critical Factors
So in order to achieve short "turnaround times" without neglecting quality, companies need a way to parallelize applications with a high volume of communication so that they can ensure transactional work in the required Service Level Agreements (SLA). In other words, it is all about application scalability and the intelligent use of resources. Against this background, APIs play a decisive role, because they enable companies to respond to changes in communication behaviour (increasing data volumes, new channels, higher customer expectations in terms of processing time and quality) quickly, satisfactorily and with the best possible utilisation of the existing infrastructure in document and output management.
A look at the current figures underlines the importance of APIs. For example, the www.programmableweb.com platform currently lists around 23,000 Web API offerings, around 80 percent of which are based on the widely used REST architecture. According to the web directory, there has been a rapid increase especially between 2010 and 2018. This certainly has to do with the role of big players such as Google, Amazon & Co., which have been "pushing" the issue since 2012. Platforms like Netflix, PayPal or eBay, which have made a lot of progress in the cloud, are now also among the leading API providers. The "API Integration Report 2019" concludes that the majority of companies and organizations consider APIs to be a business-critical factor.
According to the report, API services are used by almost 55% of them to develop new B2B solutions, followed by products for mobile communication (approx. 36%) and consumer goods production (B2C, approx. 28%). Every year, 2,000 new services are added. The pioneer in the development and use of API services is the financial industry. What is certain is that the business world, and with it communication itself, has become more agile, which is reflected not least in a growing demand for microservices. The fact that companies are increasingly switching to the cloud also "fires up" the API issue.
Background on APIs
API Economy is the business activity based on automated, multilateral, dynamic and comparatively anonymous competence networks of highly specialized partners.
An API (API = Application Programming Interface) is first of all a software interface that enables another application to use a certain functionality. For example, flight bookings can be made by other software in the application that provides the booking functionality. APIs are therefore the "digital glue" that connects services, applications, and systems. If several applications are combined in this way to form a complete system, the result is an API ecosystem. In order for it to work, a common description language for the interfaces and coordination of the overall system is required.
To the extent that the partners in an API ecosystem are autonomous companies and organisations, each with their own economic interests, the whole thing is called the "API Economy". In order to understand them, it is therefore necessary to consider the technical aspects of the API, coordination issues and the economic interests of suppliers and users of the parties involved.
As a rule, those providers of API services will prevail in the long term whose solutions help the requesting companies to achieve real added value, for example in customer communication. At the same time, the API Economy offers users of API services the opportunity to better focus on their core competencies, as they do not have to worry about developing applications that are not part of their core business. Overall, the demand for external expertise is growing, as it is becoming less and less worthwhile to build up in-house expertise for services that can be obtained much faster, cheaper and often functionally better via an API ecosystem and can be easily integrated into one's own IT infrastructure, taking into account applicable legal guidelines and framework conditions.
Meanwhile, it is almost mandatory for companies in almost every industry to adapt to the all-encompassing digitalization of economy and society. This is all reinforced by the rising expectations of customers and users, for example with regard to response times and user comfort in communication. Today, it is common practice to authenticate yourself via your personal social media account or to use the online payment tool of your choice. All this only works through the intelligent use of APIs. Technologically not a novelty, but in recent years the way they are used has developed in an innovative way. Whereas APIs used to function mainly as programming interfaces between two systems within a company (i.e. the data only moved internally), they are now becoming the "enabler" of new business models involving several parties with different interests. There is a change in perspective - from an exclusively internal interaction to an "interaction with the outside world", i.e. with third-party applications and systems.
Examples of Cloud-based API Services in Customer Communications for Companies, Organizations and Government Agencies
The Google Cloud Vision API transfers the concept of machine learning for the first time to images. With this API service, the content of images can be automatically recognized, interpreted and assigned - not only everyday objects, sights, portraits etc., but also embedded text, product logos, colors and even speech elements (automatic speech recognition). Even emotions behind a certain facial expression on photos should be able to be interpreted with this service.
Amazon Comprehend is a Natural Language Processing (NLP) service that uses machine learning to uncover contextual relationships in unstructured data or content. Various analysis tools of the cloud-based service extract "key statements", recognize the tonality of a text with regard to the "mood" of a communication (e.g. e-mails, support tickets, product reviews, social media, advertising texts, damage reports) and filter out certain names, addresses and locations within the framework of "entity" recognition (identification of the same characteristics/attributes of information objects such as persons/objects). The service identifies the language of the text, extracts keywords, places, persons, brands and events, recognizes how positive or negative the content is, analyzes it using tokenization (segmentation of a text into units of paragraph, sentence and word level) and parts-of-speech tagging (PoS = assignment of words and punctuation marks to word types) and automatically organizes the collection of text files by topic.
Users can also use AutoML capabilities in Amazon Comprehend to create a custom set of information objects (entities) or text classification models to meet the unique needs of an organization. This service is a fully managed service, which means there is no need to provide dedicated servers or develop, train, and make available machine learning models to use it. The user only pays for the actual use, there are no minimum fees or prepayments.
Amazon Comprehend Medical is a specialized tool that allows you to extract medical information such as clinical conditions, medications, dosages, strengths, and frequencies from a variety of sources (including physician letters, clinical study reports, patient records, and more). The application also identifies the relationship between the above-mentioned drugs as well as test, treatment and procedural information with the aim of facilitating analysis - for example, a specific dosage, strength or frequency related to a particular drug from unstructured clinical notes.
- Analysis of customer interactions in terms of satisfaction with the aim of improving products and services (e.g. call center analysis);
- Search more accurately by enabling search engines to register key phrases, information objects (entities) and moods (e.g. indexing and searching product reviews)
- Knowledge management: Documents are classified or categorized by topic to provide content recommendations for further articles on the same topic (e.g. personalizing content on a website);
- Automated categorization of incoming customer support documents (online feedback forms, support tickets, forum posts, product reviews) based on their content
- Recruitment of selected risk groups for dedicated clinical trials based on previously analysed medical information (e.g. patient files, physician's letters, entries in health-specific registers)
are an architectural and organizational approach to software development, where software consists of small independent services that communicate via carefully defined APIs.
Microservices-based architectures simplify scalability and reduce application development time, enable innovation and shorten time to market for new features.
Source: Amazon AWS
Representational State Transfer (REST or also ReST) describes a programming paradigm for distributed systems, especially for web services. REST is an abstraction of the structure and behavior of the World Wide Web. REST aims to create an architectural style that better represents the requirements of the modern Web. REST differs from architectural approaches mainly in the demand for a uniform interface.
REST focuses on machine-to-machine communication and represents an alternative to similar procedures such as SOAP and WSDL and the related procedure RPC. Unlike similar architectures, REST does not encode method information in the URI (Uniform Resource Identifier). because the URI specifies the location and name of the resource, but not the functionality that the Web service provides for the resource.
The advantage of REST is that a large part of the infrastructure necessary for REST (e.g. web and application servers, http-capable clients, HTML and XML parsers, security mechanisms) is already available in the World Wide Web and many web services are REST-compliant per se. A resource can be represented by different media types, also called resource representation. For example, an online service that only offers unchanged page contents according to the Internet standard http is already REST-compliant. Dynamically generated pages, on the other hand, often do not follow this principle. For example, news pages offer constantly changing information in both different formats and content that is difficult to process automatically. If the format remained unchanged, an important REST property would be fulfilled.