This post originally appeared on the Marketing Artificial Intelligence Institute's blog. The Marketing AI Institute was created and is powered by PR 20/20.
“How do I get started with AI?” This is one of the top questions we get from marketers. A lot of professionals want to understand and apply AI to their work. But they don't know where or how to start.
It's understandable. There is a lot of content out there on AI. And there's a ton of hype.
So I want to walk marketers through how I would approach this process, using mock examples.
The goal? Show you how a brand could actually go from conception to reality with marketing AI.
Let's say I'm a marketing manager at a large manufacturing company. Our business is largely B2B, selling widgets for machinery used in the oil and gas industry.
I’ve been tasked by our CMO to investigate the potential of AI to improve our marketing operations. We're not the most sophisticated marketing department around. But we're doing OK:
We use marketing automation software to manage our contact database. We have a content marketing program. And we use some basic lead scoring and email marketing to qualify leads.
But, we still invest a lot in trade shows. We sell a lot based on relationships and referrals. And the other departments we work with don't really get the whole point of "online" marketing. Our team is talented, but still learning advanced content marketing and inbound marketing.
So, we're still working to bring our marketing into the modern world.
Now, we're hearing all about artificial intelligence and how it's the future.
What exactly is AI? Is it right for us? How do we even go about applying it?
We know we need to do something here. We just don’t know exactly what.
Step 1: Evaluate My Use Cases
In fact, I would start by looking at my actual use cases. Notice I didn’t say “use cases for artificial intelligence.” (That comes later.)
I want to define my problems, then see how AI might help. That way, I don't overfit AI to a use case it might not apply to.
How would I do that? I'd list out all the repetitive tasks my marketing team does every day. Then, I'd break them down into repeatable steps.
Use your best judgment, but you probably have a pretty good idea of what activities take up the most time. For my made-up marketing department, these tasks look like the following:
- Create and edit content for our blog and for social media.
- Create premium content to generate leads and support our sales team.
- Promote content of all types using social posts and paid promotion.
- Strategize and launch formal marketing campaigns in our marketing automation software.
- Send marketing emails to our database of contacts.
- Pull performance reports on various campaigns and marketing’s overall success against KPIs.
Now, I’m sure there are plenty of other things we do. But these take up 90% of my team’s average day.
Step 2: Pick a single use case.
All these use cases might be useful to intelligently automate with AI. But if I can use AI on even one of them, I'd be in good shape. I'd free up tons of time and improve team performance.
To keep it simple, pick a single use case from the list to start exploring.
In this scenario, I’m going to start with something that takes up a ton of my team’s time:
Step 3: Break down the use case into repeatable steps.
Something like “creating content” is a pretty broad use case. So I want to break it down as much as possible. There are plenty of repeat tasks that we do to create content:
- Brainstorm topics.
- Pick topics that will perform well.
- Build an editorial calendar or plan of those topics.
- Outline each topic.
- Write posts on each topic.
- Edit posts.
- Upload posts to marketing automation software.
- Publish posts to blog.
With these repeatable steps, I have a starting point to research AI applications.
Step 4: Initial Research
I’d begin my initial research by conducting some basic searches for my use case and its repeatable steps. For the above, consider some of the following types and formats of searches:
- “AI for content marketing”
- “Create content with AI”
- “AI for blogging”
- “AI for content topics”
- “AI for content performance”
- “Build content calendars with AI”
- “AI for writing”
- “AI for blog writing”
- “AI for editing”
- “AI in marketing automation software”
I'll pull my most promising links into a common repository. As I read and research, I'll also note names of interesting solutions that come up.
(Hint: You'll find a lot of this on the Marketing AI Institute blog, including specific solutions.)
Step 5: Narrow Down Potential Solutions
By now, I have a better idea of what's out there. I also have some names that jumped out during my research:
- CONCURED, a content strategy platform powered by AI
- Crayon, a competitive intelligence platform that uses machine learning
- BrightEdge, an AI-powered SEO platform
- HubSpot, the marketing automation software we already use, which may be using some type of AI
I don’t know a huge amount about each of these solutions just yet. So the first thing I’d do is check out each of their websites to learn more about them.
I would also look for articles and reviews about them.
Step 6: Demo and Test
It looks like all of these tools could potentially benefit my content creation in some way. But to be sure, every marketer should demo or test drive tools for themselves.
Because AI tools have pretty specific applications. You need to make sure your brand actually has the data required to make the tool work. And you want to confirm it has the capabilities the vendor says it does.
There's no substitute for this step. AI has a lot of hype and buzz: it's easy to get sucked in by bold claims about capabilities.
This is why having a specific use case you can bring to vendors is essential. You'll be able to tell readily if the solution does what you need it to do.
Many vendors offer free trials and demos. Take advantage of them.
Step 7: Build an AI Strategy
Let’s say you’ve vetted some tools and like what you see. You’re ready to secure budget and start rolling some of these out.
You need an AI strategy. A formal one that's documented and vetted by key stakeholders. This strategy is going to formalize how you rollout these tools. It's going to streamline implementation. And it's going to help you understand where your team is going with all this.
Now, you don't need this document before you test drive any AI tool. You could even start piloting one or two tools while you create your strategy. But the document is a must earlier rather than later in the process. It'll take some time to create, too, so the sooner you start, the better.
Want even more advice on how to get started with AI? Select below.