Please enter your username or the email address associated with the account so we can help you reset your password.
This is the second post in our “AI for CX & Marketing” series – check out the first post here.
If you’re trying to build AI into your martech stack, you may hit some internal resistance, whether you’re trying to convince your coworkers to get on board… or asking yourself if AI is really a smart investment.
It’s no surprise – straddling the line between science fiction and high-tech reality, AI comes with a lot of unknowns.
So to make sure you’re prepared to face up to these concerns, we’ve compiled a list of 5 common fears about the use of AI in marketing – and how to address them.
Why they’re worried: Automation – especially powered by AI – can certainly pose a risk to jobs, and even entire industries. As AI gets smarter (read: better at completing human tasks), many of your coworkers may be wondering if algorithms could replace their roles, too.
The reality: With marketing teams and budgets frequently under-resourced, a lot of teams end up bogged down in repetitive tactical work, with little time for strategic thinking.
AI has the potential to free up the time (and head-space) that marketing automation platforms were supposed to. And, thanks to its knack for data analysis and picking out patterns, AI can also deliver insights that go much deeper – and are far more valuable.
AI tools that help you better understand the effects of your marketing efforts, the preferences and behaviors of your customers, and so on will enable you to make smarter, more strategic decisions.
That means more insight and the ability to act on it faster – something all marketing teams could use more of.
How to address the concern: Start by looking at areas that cause pain for your customers and your team alike – these are the use cases to target first.
Look for ways that AI can make these tasks less of a headache for your team – if you can show how it’ll take the load off or make things more efficient, it will be easier to get buy-in.
Why they’re worried: There’s a reason the EU included “transparency” in its recent set of AI ethics guidelines.
AI isn’t infallible – it learns from the data it’s given. But that data might be inaccurate, biased or simply lacking the context for accurate decision-making. It’s important to know how a decision has been made, so you know what to fix if there is an issue.
There’s also business value in understanding an algorithm’s decisions.
Without an understanding of, say, why a customer is predicted to purchase a particular product, it’s hard to get any actionable insight that you can apply to your wider marketing strategy.
The reality: Understanding why an algorithm makes the decisions and predictions it does can be challenging. Often thousands of data points have been processed and analysed to reach a conclusion.
Fortunately, AI vendors are beginning to realise that the algorithms they offer need to be “explainable”. That might not mean diving into every piece of data that the tool or platform processed, but it does mean that platforms should be able to give you some indication of why a decision or prediction was made.
How to address the concern: Look out for tools or platforms with plenty of visibility and control built in. These should give you deeper insight – beyond surface-level predictions – and allow you to approve or tweak actions if needed, before going ahead.
Why they’re worried: Let’s be honest – marketers have been known to chase after the newest and most exciting tech, often before its usefulness to the general public has really been validated.
If AI is going to be worth the investment, it needs to create long-term value.
The reality: While it’s fair to ask whether AI is just the next marketing buzzword, it’s also pretty clear that AI is much more than hype. It’s a decades-old discipline that is making some significant advancements – advancements that have already had a huge impact on our lives.
Machine learning (a subcategory of AI) drives all kinds of predictive personalization – from your Google search suggestions to your latest Netflix recommendations.
Then there are AI-powered assistants like Alexa, Google Assistant and Siri, who are gradually earning a place inside more and more homes (and pockets).
Additionally, McKinsey predicts that AI applied to marketing and sales functions will create up to $2.6 trillion in value across industries.
How to address the concern: Show that AI has staying power, and that it creates value for businesses like yours. Spend time researching current applications in your sector and potential use cases that might benefit your team.
Look for tools that are geared towards solving a few specific business problems – it will likely be easier to measure their value. Some vendors may also be able to help you with custom ROI predictions.
Why they’re worried: When we talk about CX these days, it’s primarily about “human” things – emotion, connection and personalization. So to bring in artificial intelligence and machine learning can seem counter-intuitive.
When we talk about machines being involved in the customer experience, lots of people think of automated phone systems that send callers in frustrating circles.
The reality: Counterintuitive as it may seem, AI is actually essential to achieving true one-to-one personalization.
After all, it’s unlikely that humans are manually creating personalized experiences for your customers right now – that would take a lot of team members (or a really small customer base).
So whatever level of personalization you’re offering already relies on software to some extent – probably rule- or segment-based marketing automation.
AI simply takes this existing software to the next level.
By analyzing millions of customer records and data points, an algorithm can understand each customer’s needs and predict the next “best action” for every individual. That means experiences and interactions can be more personalized than ever.
How to address the concern: Ensuring that AI-powered customer interactions still feel “human” – that is, not too cold or impersonal – is mostly down to the quality of the tools you choose.
Look for software that gives you plenty of insight, integrates with all your main data sources, and offers customization at set up, plus on-going flexibility.
Why they’re worried: AI has traditionally been a very technical discipline. To make it a reality, most companies have built solutions in-house by employing teams of data scientists – and that means a big budget is a must.
The reality: More recently, “productized” solutions have reduced the technical requirement and made AI more readily available. Rather than having to build and train algorithms in-house, companies can roll out pre-built solutions, that are then customized to their needs.
It’s easier than doing it yourself, but it can still be time-consuming, requiring input and approval from multiple stakeholders.
There are also self-serve options gradually coming to market, which significantly reduce the costs and internal development needed. These tend to be focused around a specific use case – but often that’s a good thing when you’re getting started.
How to address the concern: Do plenty of research beforehand into potential solutions. It’s not just cost – find out how easy it will be to implement, the integrations that are available and how much support you’ll need from IT.
And think long-term – once you’ve rolled out the initial project, how much control will you have? Will you need additional support (and budget) to make changes or can you manage things yourself.
SaaS platforms are good for this as they tend to be hands-on – just make sure you get something user-friendly.
View Original Article