Council Post: Three Key Points To Consider When Creating An AI Marketing Strategy

By Jon Reilly, Co-Founder and COO at Akkio, working to level the playing field for AI in business.

Artificial intelligence (AI) is the next big thing in marketing. Why? Consumers are more demanding than ever, and tomorrow they’ll want even more. Manual teams simply can’t keep up with the demands for instant-response chatbots, personalized digital experiences and more. Especially at scale. Automation is the only way. And AI is the answer.

AI can do things at scale and at speeds that humans cannot. AI is especially good at what humans hate doing the most: manual, repetitive tasks like scoring sales leads based on their likelihood of conversion.

Suppose that your sales team handled one hundred incoming leads a day. Naturally, not all leads are the same, and only a minority will convert. Even fewer will become high lifetime value customers. With AI, the conversion probability for each lead can be automatically calculated, letting your salespeople focus on the leads that matter most.

Making AI accessible for marketing teams is hard. Here, I’ll try to help you get started with the basics: three key points to consider when creating an AI marketing strategy that actually works.

1. AI’s Complexity

For years, businesses have been trying to strengthen their growth strategies using AI tools designed for technical experts only. No one looked out for the marketing teams themselves. These powerful new tools came out of marketing’s budget, but the AI wizards held marketing hostage to their own priorities. 

The constraint of an AI solution that has too much technical load can be counterproductive for both businesses and their customers. Try to express your marketing objectives in code (much less understand its output) when you’re not a technical expert yourself. 

Traditionally, AI models were built by PhDs in fields like computer science and mathematics, using complex programming languages and new public cloud technologies like Google Cloud Platform.

Now, marketing teams have the benefit of no-code AI platforms (Disclaimer: including my own company). These are visual, point-and-click tools that anyone can use to build AI models, whether it’s to score leads, predict customer churn or even detect affiliate fraud.

2. AI’s Cost

Any marketing team looking at AI has to be mindful of budget. AI is expensive and new. 

Marketing teams need to decide if a platform’s cost is worth it for their marketing needs. When evaluating cost, look beyond the initial sticker price. Have you also factored in the cost for AI model training and how quickly you can get up and running? Time is money, too.

Some platforms offer a free trial period so interested marketers can evaluate the offering for themselves before purchasing. This helps weed out potential problems or concerns on both sides before committing any money or resources.

Pricing varies widely on different platforms. It’s early days in this market without standard pricing, and some platforms more than double in cost as you add essential “options” to make it work in a production environment. For example, one provider may not include ongoing data storage and model training costs, while another does.

So list all these variable costs to see what they really cost over time instead of just looking at upfront expenses. With my company, for instance, all plans include free model training and data storage, unlimited users and unlimited AI deployments.

3. Real-World Usability

If your marketing team is considering leveraging AI to help make more informed decisions, you’ll need something that is easy for your team to deploy and use. Real-world deployment of your models into marketing tools, such as Salesforce, will be critical for success. After all, to deliver value (and justify that line item in your budget), AI needs to clearly demonstrate value. 

Workflow is critical. It trumps many features and other bells and whistles on AI platforms. It’s critical that you understand how a given platform fits into your marketing workflow. Your teams will need to deploy AI models where they work, whether it’s Salesforce, Google Sheets, a live web app or somewhere else entirely.

The tool selected should have flexible deployment options so your models can be deployed anywhere. One technical option is with API deployment. Teams can easily serve AI predictions anywhere that can integrate with code. A popular non-technical deployment option is Zapier, which integrates seamlessly with thousands of apps.

So understand three things well before you embark on your AI marketing journey: 

1. Real total cost, including ease of use and what comes standard in the upfront price;

2. Ease of use in your current workflow (it’s very expensive to make lots of different teams change how they work to incorporate your new wonder AI tool); 

3. Technical skill sets required to use in production.