Seven budgeting factors for integrating an AI virtual assistant into your organisation
August 14, 2018
It might seem obvious that the biggest advantages of introducing an AI-powered virtual assistant into your organisation would be a reduction in costs and an increase in effectiveness. An Artificial Intelligent VA has the potential to perform tasks more quickly and efficiently saving time and money.
But there’s more to it than that. VAs don’t come ready-formed: they need training and attention like any regular employee. And just as with any other addition to your organisation’s systems, you have to construct robust use cases to identify costs and benefits.
To help you plan, here are 7 budgeting factors you should consider when integrating an AI virtual assistant into your organisation.
AI is a hot topic, but you need to think carefully before adopting it. As with any other critical decision you need to identify your organisation’s business needs and explore how an AI virtual assistant can help you. Consider which aspects of your organisation would benefit most from having a VA agent. Which areas can be effectively automated? Ask yourself how it will fit into your existing systems and processes. What impact will it have on employees working with a VA? What additional business benefits will it bring? You can’t begin to budget for an AI assistant until you’ve addressed these questions.
As with other IT applications, there are tools to build your own AI VA. These bots need little coding and can be integrated into existing applications. They can also be customised at the back end to fit in with your organisation’s systems. These simple bots can be ‘hired’ on a monthly or yearly subscription.
From this basic level, you can move to more sophisticated forms of integration. These include full integration with your own services and resources and, at further cost, integration with third-party services and applications.
Once you introduce the AI component you’re reaching another level of complexity and therefore cost. But greater investment also has rewards: AI-driven VAs can recognise and process natural language and the most technically sophisticated actually learn to improve from their interactions.
Clearly, any considerations about the way forward for your organisation depends on your own IT and L&D resources.
Not all VAs or chatbots are alike. The simplest ones work according to a set of rules. They respond in a limited way to specific, programmed commands. The more advanced chatbots use machine learning, employing AI. This means they learn from each interaction or conversation, detecting preferences and making recommendations based on past requests for information.
The greater the sophistication, the bigger the budgetary outlay. The intelligent bots need to be trained, not simply installed and that means assigning L&D and HR resources above and beyond your basic IT support. It also means dedicating IT staff to producing scripts for the bots. This area will grow as more AI VAs are deployed.
Whether these measures are cost-effective for your organisation depends on how you intend to use your VA.
Let’s look at a couple of common deployments of AI Virtual Assistants.
Firstly, you can have a VA act as a PA. This can mean providing assistance in anything from daily time management to weekly project management, to support during the onboarding process for new employees. In these instances, your AI virtual assistant needs to be fully integrated with your HR, training and workflow systems and have access to their resources.
The second common role for an AI VA is in customer support. Today many online customer support engagements are with chatbots rather than human agents. Bots can make customer service more efficient, increasing the number of queries resolved and reducing the time taken to resolve them.
To provide that customer support role, your bot not only needs the resources to respond but also the intelligence to deal with queries and if necessary hand anything too complex off to a human agent. You need not only to train your bot but also your customer support staff who will work alongside it.
As the customer support example suggests, you need to think of your VA not simply as a piece of software, but also more like a regular employee. You need to onboard it too and teach it what it needs to know.
But just like an employee, with training, sophisticated AI-powered VAs perform better and more efficiently over time. And VAs don’t need breaks, holidays, pay rises, don’t go looking for other jobs, and don’t get bored or frustrated with repetitive work.
In that sense, they’re model employees, but by deploying a VA in your workforce to do the uninteresting, repetitive tasks you can also raise morale in other employees and free them up to tackle challenging, more rewarding tasks.
The benefits of an AI virtual assistant need not be just about cost and efficiency savings, but, in a more positive sense, about increased performance and productivity, greater job satisfaction and better staff retention.
With AI-powered VAs training and learning, development can have a greater impact. Given access to training content, VAs have the potential to be teachers and mentors at the point of need. VAs placed in the workflow can be accessed when information or training is really required – while people are working and doing their regular jobs.
It’s like training the trainer, but while the human trainer might forget, the VA never does. Intelligent VAs are also learning from delivering learning. And one of the key things they are learning is what employees really need to know and when they need to know it. L&D can analyse this feedback from the VA and fill any gaps in training. In this way, training becomes more effective, because it’s addressing real, quantifiable needs.
AI VAs can perform many tasks that humans do and perform them quicker at a lower cost. But you should hesitate to think that that means you automatically require fewer resources. An AI virtual assistant also needs training and support, so this means the involvement, reallocation, and potentially re-skilling or up-skilling of HR and L&D resources as well as those in IT.
The deployment of virtual assistants and the expansion of their roles are creating a raft of new jobs such as scripting, data analysis, content design for bots, analysis of human-AI agent interaction and so on. Integrating an AI VA involves crossing departmental boundaries and calls for the establishment of cross-functional teams.
And if you want to get the most out of your VA you might extend its role which means integrating it into different aspects of your organisation’s work.
You also need to consider what to do with the data the AI virtual assistant feeds back. Someone needs to analyse it and react to it. If there are queries that the VA can’t respond to, you need to coach it. The VA’s inability to answer may point to a critical gap in your processes or training which needs improving.
These measures might seem to add, rather than reduce costs, but properly integrated and implemented, AI VAs have the potential to streamline work practices and standardise processes leading to efficiency savings and improved performance.You might start out thinking an AI virtual assistant primarily represents cost savings. But looking at these budgetary factors, you’ll come to realise it also involves the reallocation of existing budget as well as investment in the future.
You need to integrate an AI virtual assistant from a more general, organisational perspective. From that vantage point, it’s about enhancing your organisation’s capability and performance by enabling better communication, greater standardisation, improved efficiency – all translating into the real prize of enhanced performance.
Got a learning problem to solve?
Get in touch to discover how we can help