26 September 2018

Is A Chatbot Right For My Business?

Tim Parrack
Written by Tim Parrack

Tim Parrack is Regional Director for The Marketing Centre and specialises in working with small and mid-size businesses. He has over 30 years’ experience working with technology, comms, healthcare and construction sectors, focussing on UK as well as global markets.

Driven by headlines about self-driving cars, robot-run factories and fully automated call centres, businesses are spending an awful lot on AI - the sector is expected to be worth $1.2 trillion by 2020. It’s easy for business leaders to be swept up in the hype and think, “we need to get in on this”.

Is that actually true? Maybe. AI is definitely disrupting industries across the board. In agriculture, its ability to track events and make predictions is helping to address labour shortages by directing personnel to where they’re needed; in retail and tech support, AI powers chatbots which automate frequently asked questions and data collection, saving human man-hours for where they’re needed.

According to Stanford University’s AI Index, top-performing companies are more than twice as likely than their industry counterparts to use AI for marketing. Away from the top tier, 84% of enterprises believe investing in AI will lead to greater competitive advantages. But is there any substance behind the statistics - and is AI right for your business?

Before we get started: explaining AI

Be honest: did you think “robots taking our jobs”?

We can’t say that’s never going to happen, but it’s a long way off yet. AI is a little more nuanced than that, and it’s important to know exactly what you may or may not be investing in.

Artificial Intelligence is the general term for “making computers think like people”. AI mimics the way humans make decisions and behave, but harnesses the enormous processing power of computers to make those decisions faster and more accurately.

It works through ‘machine learning’ - which means predicting patterns based on large amounts of data. What it can’t do is understand the real world context of that data.

Consider Amazon’s Alexa. It doesn’t understand words as such. It learns particular sounds - literally the patterns of the waveforms - and it’s programmed to react to them in particular ways. So it can play ‘Despacito’ when you ask it to, but there’s nothing there that can grasp why you want the song played.

A website chatbot is limited in the same way. It can learn, over time, that a particular facial expression means it should use a particular script, but it doesn’t know what anger is or why a particular tone of voice is reassuring.

Now that you understand the limits of AI, it’s time to tackle the crucial question: to invest, or not to invest?

When to invest...

When it will significantly enhance your customer experience

Your customers should be at the centre of all business decisions, and adopting new technologies is no different. If AI isn’t going to significantly improve your customers’ experience with your business, then why would you use it?Artificial intelligence

Consider habito, whose AI Digital Mortgage Advisor (DMA) automates the mortgage management process for consumers, with no brokerage required. The DMA tracks the customer’s financial situation - employment, salary and personal life plans - and cross-references them with real-time market mortgage rates to calculate payments. These calculations are nothing a human broker can’t do, but the AI can compare hundreds of products in a fraction of the time. There’s a palpable payoff there in terms of how quickly the customer gets what they want.

If it will give you a competitive advantage or USP

If the benefits that AI brings will give your business a competitive advantage, such as pricing or efficiency, or substantially differentiate your offering and give your business a USP that’s appealing to your audience - then it could be a good idea.

For instance, if your business is tying up work hours and productivity in routine tasks that could be automated - and according to Sage’s bot and AI specialist Kriti Sharma, most businesses spend 120 days a year on admin alone - then AI could streamline your operations by letting your people focus their energy on more important operations.

When not to invest...

Because everyone else is

Your business is its own entity, with its own strategy and its own goals. Rushing off to invest in AI because you want to look cutting-edge and on-trend is not going to pay off.

If it looks like your whole industry’s jumping on the AI bandwagon, it’s tempting to either stick your head in the sand and ignore it, or throw money at it and hope some ROI shows up down the line. Neither of these approaches works.

A successful AI investment isn’t based on chasing the trends - it’s based on a full, frank review of your business and its operations. Unless there’s a process or point in your customer journey where things slow down and AI can speed them up again, you don’t have a strong case for investing - you have a fear of being left behind.

If the results aren't proven

EDF were understandably reluctant to adopt AI on the energy generation side of their business: as their digital innovation leader David Ferguson put it, “‘move fast and break things’ doesn’t really work in a nuclear power station”. Their adoption has been cautious, controlled, and focused on monitoring situations rather than making decisions, ensuring that AI has a limited purpose for which it’s extremely fit.

Early adoption - in other words jumping on new technology as soon as it’s available - doesn’t pay off. Think about smartphones. They’ve been around since IBM released the Simon in 1992, but the tech didn’t mature until Apple perfected it decades later. Virtual reality hit the marketplace at the same time, but couldn’t deliver on its promises until computers’ processing power caught up, well into the twenty-first century.

Some technologies don’t take off first time around. Until tech has proven itself useful to your sector, there’s no case for investment.

If you have no data

This is the big one. AI needs data. Machine learning - the process of developing an AI so that it can actually do the job that needs doing - relies on it having data to process, analyse, and sort into trends and segments.

Exactly how much data you need depends on the complexity of the data and the situation at hand, but you need enough for trends and tendencies to be apparent.

If you don’t have data relating to the specific problem you’re trying to solve, AI can’t solve it. It’s like a train without tracks or a car without roads; it’ll just sit there, being expensive and doing precisely nothing.

If it's not part of your long-term business strategy

It can be tempting to be attracted to the latest shiny thing. However, if it’s not part of your business strategy then don’t be distracted by it. Remember that a lot of the hype you read comes from people in AI development who are marketing their own wares - not providing measured, practical advice. Wait, assess, and take up something once it’s been proven to work.


AI is predicted to have a huge impact on businesses in a wide variety of sectors - but most of the buzz we’re seeing is still predictions. Business AI is in its early stages, and any significant change to your business offering shouldn’t be taken lightly. Consider how AI will impact your customers’ experience with your business, and how it fits with your wider business strategy - and if you can see an obvious practical gain, it’s time to invest.

If you’re struggling to understand your customers, or get visitors to your site, get in touch - or to see how your marketing efforts are working, use our free Marketing 360 Healthcheck.

 

 New Call-to-action

Share

Ready to take
your marketing seriously?

Since 2010, we’ve helped business owners make sense of marketing.
To make smarter decisions and make the most of their investment.

So if you’re tired of switching from one thing to the next hoping one will stick, maybe it’s time to try a different approach.