There is no shortage of interest and discussion around artificial intelligence (AI), language modelling and other exciting new technologies in the context of business communications.
However, the best way to see these technologies is that they are tools to achieve an end result, writes David Meintjes, CEO of Telviva.
That end result should be to aim at a better customer experience, digital when the customer wants it, but human when they need it.
Automation of customer communication is not new. In a business to business (B2B) context, Electronic Data Interchange (EDI) was available in the early 90s led by Shoprite, historically placing a strong emphasis on supply chain efficiency and cost reduction.
In the business to consumer (B2C) market, the increasing popularity of the internet meant that email became a viable and cost-effective channel for businesses to reach consumers directly.
In the mid- to late-90s this laid the foundation for automated email campaigns, like newsletters and promotional offers, though these were often more rudimentary than today’s highly personalised automated flows.
Thereafter the rise of e-commerce also necessitated more automated communication for order confirmations, shipping updates, and customer service inquiries.
The current technologies like speech-to-text (and vice versa), as well as the large language models, open the door for further automation to step into the real-time engagement or self-service dimensions of interfacing with customers.
The next frontier of automation is no longer just about automating repetitive tasks; it’s about creating intelligent, adaptive, and autonomous systems that can learn, reason, and make decisions, often without direct human intervention.
This evolution is largely driven by advancements in artificial intelligence (AI) and its integration with other cutting-edge technologies. It is a present-day necessity for businesses seeking to remain competitive and efficient.
The first step for any business is to look at why they are pursuing automation and how it fits into the business model. The path to successful automation is neither simple nor uniform. It demands more than the adoption of the latest AI or other tech tools. It requires foundational readiness, careful planning, and unwavering business ownership of the customer automation journey.
The first, and potentially the most critical, step in any automation journey is ensuring that the business systems are ready. Without integrated, robust, and reliable systems, there is little to automate because automation cannot succeed in a fragmented environment. The foundational layer – this is systems and data infrastructure – must be in place before any meaningful automation can occur. This readiness is not optional; it is the bedrock upon which all automation efforts are built.
Once this has been addressed, the business needs the self-control to walk before it runs. In other words, when it comes to automation, ambition is important, but so too is pragmatism. This means starting with a clear plan, identifying low-complexity, high-volume processes as initial candidates for automation, and iterating from there.
Rushing into automation without a solid plan often leads to wasted resources and suboptimal outcomes. Careful planning also involves mapping out customer journeys and business processes to identify where automation will deliver the most value. There is no substitute for meticulous planning from the as-is state to the desired to-be.
Any successful automation initiative is always underpinned by a well-defined strategy. This strategy should articulate not only the “what” and “how” of automation, but critically it must address the “why.”
In addition to this, businesses must take the time to fully understand the costs – both direct and indirect – associated with automation. This includes technology investments, integration expenses, change management, and ongoing maintenance. Only with a clear-eyed view of these factors can businesses make informed decisions and set realistic expectations.
Perhaps the most important principle is that ownership of automation must reside within the business itself. While expert partners can, and should, offer expertise, direction and support, the ultimate responsibility for outcomes, good or bad, rests with the business.
What does this mean? It means the organisation must take charge of process design, data quality, and ongoing optimisation. To understand why this is so important, the following must be reiterated: Automation is not a set-and-forget solution; it is an ongoing journey that requires active stewardship, and the best steward is the business leadership itself.
Once businesses understand all of this, they can look at actual potential benefits and use cases. When everything is set up properly, the potential positive benefits are certainly appealing. Some, but certainly not all, include:
- Scalability without proportional increases in cost;
- 24/7 availability and improved customer self-help options;
- Consistency in service delivery and customer experience;
- The ability to redeploy human resources to higher-value tasks; and
- Enhanced personalisation as data and systems mature.
Practical use cases include automated support and ticket routing, proactive notifications, feedback collection, personalised self-service, sales process automation, HR onboarding, and internal knowledge management. The key, when designing the right use cases for each organisation, is to focus on repetitive, routine, and time-sensitive tasks of low complexity as starting points.
Automation certainly carries risks, though not all risks will manifest in all scenarios. Some risks include the loss of personal touch which can be a particularly annoying bugbear for customers, an inability to handle complex or novel issues, technical glitches, and the potential for bias or “hallucinations” in AI-driven systems. Biases are largely negated when the source data is the business’s own, clean and well-managed, dataset.
Potential risks brought about by automation can be mitigated by:
- Focusing on low-complexity, high-volume processes first;
- Mapping and optimising customer journeys before automating;
- Testing thoroughly with real user groups (both of the previous two points go a long way towards mitigating against hallucinations);
- Ensuring a human fallback is always available at the outset; and
- Maintaining business ownership of processes and data.
Knowledge management: the logical starting point
Automation is not a destination but an ongoing journey of iteration, maintenance and management. It requires starting and then building from there.
For many businesses, the most accessible entry point into automation is knowledge management. Automating responses to frequently asked questions and common customer inquiries can deliver immediate value, improve customer experience, and free up staff for more complex tasks. As systems and data maturity increase, businesses can then expand automation into more sophisticated areas of the business.
By starting small, focusing on foundational processes, and building on early successes, businesses can unlock the transformative potential of automation while avoiding the pitfalls that have derailed many initiatives. The best partners are those who guide appropriately while understanding that the future belongs to those who prepare, plan, and take ownership of their automation journey.