- Choosing your first AI project is essential as digital organizations evolve.
- The correct use of AI is helping brands with insights with precision, accuracy and scale.
- The benefits are measurable and real if you know what you want to achieve.
- AI is here to help and complement marketers’ skills, help them with laborious tasks and unleash more creativity and elevate personal and professional performance.
- A look at five simple steps organizations can take to start their AI journey.
Today, organizations of all types and sizes are implementing AI; according to Gartner, 70% of organizations will integrate AI in the workplace to assist and increase the productivity of their employees by 2021.
Artificial intelligence offers far more utility than simply automating processes, though. In digital marketing, AI is helping brands achieve the personalization and engagement it takes to drive revenue and loyalty in the current consumer environment.
From personalized content and product recommendations to dynamic pricing, intelligent chatbots, intent-driven search optimization and even data-informed content creation, AI is changing the game for marketers.
The benefits are measurable and real, with recent research showing that AI enables marketers to:
- increase sales (52%)
- increase in customer retention (51%)
- and succeed at new product launches (49%)
In another study by McKinsey, 44% of companies reduced operational costs and increased business revenue by using AI in marketing.
How can you put AI to work within your own marketing organization? There is no singular path to AI success. Work through these questions to determine your level of AI-readiness, then check out five steps below to help you get started on your first AI project.
What is the current state of our data assets?
Algorithms are hungry for data, but like the food we feed our bodies the data you feed your AI tools is key. The problem is, most organizations are falling down on data governance.
A recent study showed that just 3% of organizations analyzed had Data Quality Scores considered acceptable by researchers.
What is our organization’s comfort level with the concept of AI?
If your marketing team and C-suite aren’t even talking about it yet, you may have some serious work to do in laying the groundwork for budget allocation and adoption. Your first step is advocating that decision-makers gain a basic understanding of AI and machine learning.
They don’t need an academic understanding of the internal workings of artificial intelligence, but should have a grasp of different AI techniques, models, approaches to data collections, and applications.
Artificial intelligence and machine learning: What are the opportunities for search marketers? offers a good primer on the opportunities and from there, there are plenty of free courses to work into your organization’s ongoing training.
How all-in do we want to go?
Maybe you want to streamline entire business or marketing processes. Or maybe you just want to dip your toe in the water.
Getting started in AI a lot easier than it was even a few years ago. The tools you already use may even be incorporating AI capabilities already—we’ll talk about that shortly.
What do we want our AI solutions to achieve?
Be specific. AI might be used to personalize content to customers’ needs and expectation, or to optimize aspects of the customer journey. You might use it in ad targeting or content promotion to get in front of more motivated audiences, or to create hyper-targeted campaigns that span networks and publishing platforms.
L’Oréal for example used BrightEdge (my company) AI based insights to drive 30% growth for Kiehl’s with quick answers. eBay uses AI to generate thousands of personalized email newsletter subject lines in minutes, for example.
It’s a very specific task with a clear benefit and measurable outcome that enables the marketing team to connect at a scale it would take an exceptional (and ultimately too costly) amount of human copywriting talent to achieve.
Don’t choose your AI solution first and then try to make it do what you need. Determine your desired outcomes so you can approach different tool providers with a solid idea of what you need your solution to do.
Do we have the skills and resources in-house to implement?
If you are looking for a solution to digitally transform the business, start by making sure you have the right talent in-house to help guide the solution to successful implementation.
You might need programmers with specific language skills to make it happen—Python, R, Proof, Lisp or Java are commonly used in AI applications. You’ll need managers with equal parts creative and analytical skills to encourage adoption, troubleshoot, and analyze performance, as well.
Are we adding to or replacing elements of our stack?
Will your AI solution integrate with all other elements of your stack, or might you inadvertently be creating new (time consuming) workarounds?
CMOs need to manage the performance of the stack as a whole and ensure that any new introduction to it serves a necessary purpose without taking away from any other function. If there are challenges in integration, can your choice of vendor help you find a solution?
In some cases, integration is less a concern, particularly when the output is immediately actionable and augments the human team’s performance.
In my role I have seen customers with massive gains from the use of real-time research and automation such as 2X increase in conversions and 28% improvement in ad quality score.
AI led research can help enable brand’s content teams to quickly identify the most lucrative content gaps and create content optimized to meet those exact needs.
Getting started with AI: Five tips for selecting your first project
In a recent HBR column, Coursera founder Andrew Ng recommends that brands start small and use your first AI projects to build organizational confidence in the technology. “The purpose of your one or two pilot projects is only partly to create value,” he said. “More importantly, the success of these first projects will help convince stakeholders to invest in building up your company’s AI capabilities.”
Keep that in mind as you establish your success metrics, and use these 5 tips to get you started:
- See if your current vendors have AI. Maybe they don’t today, but is it on the roadmap? You may have AI features and functionalities coming to your stack via the solutions your teams are already using.
- Pick a project based on low-hanging fruit. Personalization is perhaps one of the most pressing needs in marketing. How can you better connect with consumers and meet their needs for tailored content? The American Marketing Association used AI with NLP (Natural Language Processing) to personalize its email newsletter content and boosted engagement by 42% in the process.
- Document use case. What steps or actions are needed to ensure your AI solution is able to achieve the goals you’ve set out for it? Make sure each member of the team understands their role and any ways they are expected to support the project.
- Empower staff to work alongside and maximize the value of your AI. Ideally, your team and technology complement and augment one another’s performance; your employees should never be made to feel as though they’re being replaced or competing against tools for budget. Support your team with a robust AI training program and nurture a culture of failing fast and moving on.
- Measure performance and analyze outcomes. Again, for your first project you aren’t necessarily looking for outcomes as massive as gains in continental market share, as Coca-Cola was with its Trax Retail Execution AI. Your KPIs for your first project will depend on your solution but should include multiple metrics that give you a clear view of actual performance. This might include email opens, content engagement, time saved, sales driven, etc.
Want to learn more? Read Finding intelligence to act on from big data: a five step approach and read more about some key influencers to follow here.
The post How to set your first AI project up for success appeared first on ClickZ.