Tools to Leverage Artificial Intelligence in Marketing and WordPress

There was a time when the mention of artificial intelligence was met with fear (“Oh no! The machines are gonna take over!”) and apprehension (“Agh! They’re gonna put me out of a job!”) Business owners have quickly discovered that there isn’t much to fear, at least when it comes to leveraging artificial intelligence like machine learning, deep learning, and natural language processing for marketing.

A Business Insider Intelligence report paints an impressive picture of AI technology in marketing today:

BI Intelligence - AI Adoption
Look at that YoY Growth

The only problem businesses are having with AI right now is in the preparedness factor.

Conductor’s 2018 marketing and technology report showed that more executives feel unprepared for artificial intelligence than any other marketing technology:

eMarketer - unprepared for AI
So who’s going to define all this new tech for marketers?

Having acknowledged that, is the adoption of AI in marketing even realistic for most businesses?

Whether it’s inside of WordPress or in your efforts to promote the business outside of the CMS, there are a number of ways in which you can leverage artificial intelligence in marketing. You might be surprised by how many ways there are and how effective they can be.

How to Leverage Artificial Intelligence in Marketing and WordPress

TUNE asked over 340 CEOs, marketers, and influencers the following question:

“What technology or set of technologies will impact marketing the most next year?”

Tune Marketing impact
The vast majority voted for AI

The overwhelming response was that AI and its associated technologies (chatbots, voice search assistants, etc.) was anticipated to be the most impactful technology this year.

Rather than wait until AI has become commonplace and has given your competitors enough of an edge over your website, let’s brush up on the benefits and ways you can leverage AI in marketing now.

More Efficient Data Handling

In general, artificial intelligence gives you the ability to make smarter decisions when faced with greater amounts of data. Think about how much more efficient and accurate your audience segmentation and user journey profiling would be if you could mass analyze all that data your website and other marketing channels receive.

Prediction Capabilities

Add deep learning, machine learning, and predictive analysis, and you can go even further with AI. Through these learning techniques, “machines” grow and adapt as more information is added to the system. The machines then make inferences about user patterns, behaviors, and triggers to come up with real-time recommendations that help prevent user churn, cart abandonment, and so on. They also suggest optimal campaign conditions and platforms to help you see the best results from your marketing efforts.

Amazon currently uses machine learning to optimize its supply chain and maintain the right inventory levels. It also uses deep learning to provide personalized recommendations to users.

Enhanced Customer Service Through Automation

Chatbots have been part of the public lexicon for a couple years now, thanks to Facebook’s Messenger service. But it’s in 2018 when we should expect to see this website trend really shine. With access to more data and tools to now process it, chatbots don’t have to be relegated to answering overly simplified questions.

Enhanced Customer Service Through Data

Customer service is getting better now, too, thanks to what human support representatives can do with properly labeled data. Take, for example, Udacity, that uses a supervised learning system to influence responses its support and sales team give to users.

In essence, it works like this:

  • The co-founder and assistant of Udacity noticed that certain support center replies were more effective.
  • Using a reinforcement learning system, they labeled chat room logs based on what they deemed were successful interactions and what they deemed failures.
  • The data was then sent through the supervised learning system and used to make predictions about what kinds of responses were best for certain queries.
  • The sales team was given these “successful” messages and asked to use them when those kinds of queries arose.

As a result of the extensive data analysis, labeling, and team training, Udacity boosted effectiveness rates by 54%. And, because of the increase in successful conversations, salespeople were able to handle twice as many customers.

Quicker Searches and Responses with Natural Language Processing

We have already seen ways in which voice users might interact with a WordPress site and what you should do to reach them in search. With natural language processing technologies, AI systems (like chatbots, social media bots, voice-activated devices, and so on) can receive spoken queries and respond in a natural-sounding manner.

As these devices interact more with users and learn what kinds of responses are the most well-received, this technology will become incredibly important for improving the user experience in search.

Optimized Content Production

As of right now, AI-powered bots are really only capable of writing short news posts comprised of short textual summaries. So, it looks like you’re going to have to continue employing that freelance writer for a little while longer (ahem).

That said, there are other areas of content production that can be automated and optimized by artificial intelligence.

Dynamic email marketing is one such way. Basically, the AI reviews your website and your newsletter subscriber’s behavior when they’re on the site. It then compiles custom and personalized email marketing messages from the data.

Email marketing, in general, can also be automated by allowing an AI to curate content from your website or from other RSS feeds.

Other aspects of your marketing strategy can be optimized, too. For instance, you can use AI algorithms and predictive analysis to choose the right platform and timing to send messages on social media. You can also use this to improve publication of blog content so that it’s created in the right form (e.g. written, video, etc.), goes out at the right time, and is received by the right audience.

Improved Customer Acquisition

One of the frustrating things about the customer acquisition process is how long it can take. Sometimes it’s because you’re too busy to nurture those leads and build a relationship with prospects. And sometimes it’s because you don’t have the right software to segment your data, which would consequently give you a clearer path to conversion.

But AI-powered technologies can take away those pains for you by mass-handling the data, segmenting your list, and automating your lead generation efforts.

Sentiment Analysis

This is an AI and machine learning method that’s still currently in the works, but it’s one worth mentioning. Basically, sentiment analysis is a system marketers will be able to use to sweep platforms like social media and forums to get a sense for what the general public opinion is towards a brand.

Right now, we’re able to use tools like customer ratings and testimonials to gauge something like that. However, that is really only just the tip of the iceberg of reviewing public opinion. With sentiment analysis on one’s side, a marketing or development team could quickly scan large amounts of text to gain a general consensus.

Artificial Intelligence Tools for Marketing and WordPress

Clearly, there are a lot of use cases for AI in marketing these days. Even if your business and website haven’t had a chance to accumulate massive amounts of data yet, you can make use of other kinds of artificial intelligence methods like chatbot customer service and content production automation.

Here are some of the artificial intelligence tools to leverage for marketing both inside and outside of WordPress.

  • Akismet

    You already know about Akismet. It’s the de facto anti-spam plugin we all use to keep comment boards, forums, and contact forms safe from spam. While this isn’t the most advanced form of AI, it is AI all the same. Not only does the plugin know to look for verified spam, but it learns quickly what kinds of content you deem as spammy (even if they don’t necessarily come from malicious sources).

  • Albert

    Albert is an AI tool that optimizes cross-channel marketing campaigns. That means you no longer have to handle each part of your marketing strategy in isolation. With Albert, you create or curate content and it will direct you towards the ideal campaign execution for it. It also conducts autonomous testing and analysis in order to improve future marketing campaigns.

    Dole was able to use this software to determine that mobile and video were the optimal platforms for reaching customers. With Albert’s help, it increased sales by 87%.

  • Dynamic Yield

    Dynamic Yield is another example of an omnichannel marketing campaign optimization tool. This one focuses on helping users more effectively segment their customer base and then create personalized messages and marketing campaigns to target them. Specifically, Dynamic Yield targets e-commerce, media, travel, and gaming websites–especially ones trying to make a splash on mobile and email.

    As an example, Dynamic Yield helped online retailer Lamoda create over 160 customer segments and then deliver highly targeted messages and personalized offers to each.

    Interested in Dynamic Yield?

  • Grammarly

    Grammarly is a tool I generally recommend to boost productivity and streamline the content creation process. This online writing assistant is easy to use. Simply install it in your browser and allow it to mark up your on-page communications and content for errors. As you provide it with feedback on what actually constitutes an error on your part, it will provide better recommendations in the future.

  • IBM Watson Assistant

    IBM Watson Assistant is more than just a live chat tool. It’s a bot-powered interface that automates customer support while also helping connect your visitors to real people (when needed). The great thing about this plugin is that you can program it to respond exactly as you would, which means less friction for the user as they’re encountered by a more natural-sounding support representative.

    Interested in IBM Watson Assistant?

  • MyCurator Content Creation

    While your WordPress blog should put out custom content regularly, sometimes it’s hard finding time to sit down and craft something completely new.

    One way to supplement your blog writing process is to curate content from other sources. With the MyCurator plugin, you’ll receive recommendations on third-party content worth sharing. As you review the recommendations and up- or down-vote them, the plugin learns what your tastes are and provides more accurate suggestions in the future.

    Interested in MyCurator Content Creation?

  • Related Posts for WordPress

    Internal link building is a marketing strategy that allows visitors to find more and more related content within your site to peruse. When it comes to blog posts, there’s another opportunity for internal linking; only, this one takes place at the bottom of the post when you suggest related content for them to read.

    This plugin analyzes all of the content on your site and creates natural connections between them. Then, as it automatically makes recommendations after each post, it monitors user response to “learn” which ones work best.

    Interested in Related Posts for WordPress?

  • WordLift

    If you want assistance improving the overall quality and impact your content has on visitors and search engines, look no further. WordLift is an AI SEO tool that carefully dissects your content and optimizes it where needed.

    For the user experience, it provides suggestions on information and visuals that it knows your end users want to see and that would enrich their experience overall. For search, it adds semantic markup for better structuring and readability. You can also use this tool to review various distribution platforms and publish at the right place and the right time.

Wrapping Up

The one thing I would suggest is to be careful in how much artificial intelligence you adopt and how much you depend on it to control your marketing. That’s not to say you shouldn’t leverage artificial intelligence, just be smart about what you use.

As explained by Harvard Business Review, there are three risks to be mindful of with AI:

  1. Machine learning biases can unintentionally stem from the original human source if you’re not careful.
  2. AI technologies focus on statistical truths and not literal ones. Machines predict outcomes based on what patterns and histories show to be the case the majority of the time. However, the best course of action might not be the statistically proven method and may need to be evaluated on a case-by-case basis.
  3. When an error is made, diagnosis in deep learning or machine learning can be near impossible.

So, leverage artificial intelligence for what you can, use it to improve your business, but remember that these systems should complement your marketing efforts; not replace them altogether.

Brenda Barron
Over to you: What’s the most common request you receive from clients when it comes to AI technologies? Are you leveraging AI already? Tell us how!

8 Responses

  • Good article, thanks. I think it’s important to recognize the audiences for providing and consuming AI.

    When I see the stats on companies using AI or IoT and those who plan to use them, I’m a bit skeptical that the survey was applied to a select audience of large companies and/or those that are already very leading-edge. The average small-to-mid sized business still has no idea what AI is or how to apply it to their sales of widgets or local services.

    This is a good time to get involved with AI, frankly because few people really know what it is. Right now everyone is translating “chatbot” to “robot” and that goes to “evil automatons that will take over the world”. Look at the opening paragraph of this article, which is similar to the opening paragraph of all articles on AI these days. That’s just ridiculous, but for better or worse there is opportunity in confusion.

    WordPress developers may not need to take part in the AI revolution at all. Many are freelancers or work in small groups, and perceive AI as something that only large companies might use for analyzing large volumes of data. I suggest though that all of us need to consider what AI means for our potential clients more than ourselves.

    Simply offering AI-based tooling will put a website developer into a separate class from those who don’t, at least until the novelty wears off and “everyone does AI”. Remember that cell phones in the 80’s were only for those who afford such things, but look at where we are now. Providing scalable solutions with AI can shift a provider’s prospective audience from local stores needing a single-page site to larger companies that need a Lot more. In other words, the earnings potential goes from a few bucks per site way up to building a “real” business. Carpe Diem.

    I’m sure many reading this article will say “yeah but what can I actually do with it?” While this is new and interest is high, I suggest that some people should simply focus on one thing very strongly. Individuals can learn the facts from the FUD, consult for a fee, and create new offerings in this chaotic space that offer true value. Become the go-to expert for just one plugin or vertical market (retail, medical, geo/mapping, travel…) so that companies that just want to do one thing can easily find someone who knows how to do it. You don’t need to write the code, just find a plugin where the author seems to know what they are doing, learn it well, participate in the plugin support channel, and make it a core part of your own offering. Build a business around it – and remember to give back generously to those who make it possible for you to do so, or they will go away, and so will your business.

    Probably the first step in that process is to look at the examples cited in this article, and ask yourself “what is ‘intelligent’ about this tool?”. When you understand what they do you’ll see that it’s all a variation on the same theme. What’s that theme? Ahh, now it’s time to look closer into how AI works. We provide multiple samples to the software and tell it how to categorize them. Then we give it new samples and see if it can tell us how to categorize those based on what it knows already. The better able it is to do that, the more we say it’s “intelligent”. That’s all there is to it. So seek out the real AI engines, not just the software that has a lot of If/Else or Switch statements with a claim of intelligence. And then think about how different kinds of categorization like that can be of value to you, your clients, consumers, Facebook users, or any other audience.

    • This is, by far, one of the best comments I’ve read in a long time and some very solid advice Tony G!

      I’ve labored in the OSS software arena for nearly 2 decades and have built a very good career doing so. Your advice succinctly conveys the advice I would give newcomers in this space. Learn it, love it, leverage it.

      Thanks for the comments! :-)

      James Morris

    • Tony G said it much more eloquently than I could possibly endure towards. I would say that the splatter fest of weak AI in everything we encounter on a daily basis, it’s getting old… Fake chat is Fake, it’s such a weak attempt at using AI that it discourages people to reach for the real goals of AI being helpful.
      Who enjoys their interactions with the automated phone systems that are all but impossible to get past unless you define your inquiry according to the established set of variables?
      How about jumping on a chat that can only answer this or that and has absolutely no interaction with what you are are actually saying.

      Fake AI is Fake, we are not there yet, and all those encouraging the use of Fake AI are damaging the potential of actually encouraging development that is real. encountering an automated chat or phone service is the equivalent of a 1980’s visit to the DMV, worse than useless, it’s like, I could of just Googled it faster…

      • Hi Matthew!

        I think you’re right. “Fake” is a strong word but it’s what we all (most likely) deal quite often with. Like Tony G already said “seek out the real AI engines, not just the software that has a lot of If/Else or Switch statements with a claim of intelligence” and that’s true. Most of so called “AI solutions” (especially chatbots) nowadays are in fact just more complex/advanced “conditional logic implementations” :) I’d say that the “AI” term became “fashionable” so many overuse it just to attempt to stay “on top of the market”.

        However, that doesn’t mean that there is no AI tools/solutions/implementations around. On the contrary, it’s a “blooming” industry now and I am a strong believer here, I’m pretty sure will see that “exploding” over the web (but also in “offline” areas of our life) in these upcoming years. Which is, in my humble opinion, a cool thing :)

        If I was to add up something to what you both wrote here, I’d say that there’s one more thing we all desperately need regarding AI: education. My point is: the “hi-tek industry” tend to make an assumption that “if we understand it, everybody understand that too”. In other words, “if I know what AI is and I understand how it works, I can put it on market and everyone will be happy”. And that’s not quite true. Following-up Tony’s comment above: an “average Joe” would probably be scared because he doesn’t understand what’s happening “behind the curtain” and if he doesn’t understand something, it’s plainly dangerous. But also – if the same “Joe” deals with “intelligent” chat bot that’s not even capable of the basic interaction, he’ll probably consider it not intelligent at all and “project” that image on everything that’s labelled with “AI” mark :) So, more effort should be put into explaining what AI actually is, how it works, how to recognize AI from “just if/else logic” and how could that server a regular user :)

        Thanks for taking part in the discussion!

        Adam

        • “I’d say that there’s one more thing we all desperately need regarding AI: education.”
          Truer words have never been input… That really defines so many aspects of progression and the acceptance of new technology in the whole.

          I use Siri and Google Assistant on a daily basis, repeatedly, once the two start playing well together things have gotten so much easier. People in general, Tony G’s average Joe, do not understand the implications. Standing at the the counter at the doctor’s office, given the next appointment, without moving my hands or taking out my phone, “Okay Siri, add an appointment at Dr. Dean’s office for August 7th at 9:30am to reprogram my cochlear implants” then from my pocket, “How long will the meeting last, Matthew”, “A long time Siri, a long time”, then from my pocket, “Okay, Matthew, meeting scheduled with Dr Dean at 1435 Rosedale Avenue, Fort Worth, between 9:30am and 12:30pm on August 7th to program cochlear implants.” then “Hey Siri, make sure Google knows too” from my pocket, “Google calendar updated, Matthew”

          I didn’t tell Siri, how long the appointment would be ‘exactly’, I was just being a smartass with my pocket assistant, and the AI figured out all the needed details, even the snarky reply about sharing it with Google, as if she wouldn’t have already. I’ve found the more I use both services the adaptations and ‘logic’ is getting far more versatile. While Siri is better with personalized interaction, and Google quite cold, Google’s conditional logic is not set in stone, you can just throw random things at the GA and it figures out what’s needed.

          That, I feel is AI, the kind of AI we need to encourage, I would sing praises to the AI we find in Google Adwords with incredibly expensive suggestions… While the AI we find in our Google Analytics console always seems to know what we should be looking at even if it’s not what we set out to look at, pulling up stats that prove useful.

          I wouldn’t know how much y’all use either Siri or the Google Assistant, I was myself hesitant at first. Now I find myself talking to one or the other several times a day, in all kinds of different circumstances, discussions in the conference room, “Well let’s see, Okay Google, what’s the average cost of the keyword ’18-wheeler accident lawyer’ in North Texas, show results on the conference room LG display”, and it happens, that’s AI and it always earns strange looks from the uninitiated.

          • Matthew, that is indeed excellent AI. It’s amazing that this subthread hit a nail on the head – Just yesterday I sent a comment to an app author and got a response the next day that smelled of Fake AI, as does a lot of email spam these days. I commend them for trying to hop on the leading edge, while at the same time feeling highly insulted that while I took the time to explain an issue in their software, they dismissed it without human intervention – and did so while pretending that the response was coming from a person. I’m certain that we will see as much of this as we have of telephone automated attendants – some companies get it spot on but most get it awfully wrong.

            On topic with leveraging AI and WordPress: A couple of the offerings that Brenda cited in this article are freemium enticements toward SaaS subscriptions. I don’t think there are many other AI-based WP offerings out there … yet. There’s no doubt that this wave is being monetized in a way that others are not or cannot. There are lots of ways for people to catch and ride this new wave and get rewarded for the effort. Do it poorly and the ride will probably be short. Personally, I’ve always had an eye on this area, and with a newfound enthusiasm for WordPress plugins and the broad audiences using WP-based sites, I will certainly be looking to create a few WP offerings.

            Catch that wave dude!

  • …yeah, I was thinking about ‘ai’ in the WP space awhile back and had a look around for some good domains… have to mention: https://wp.ai/

    …seems they’re running good ‘ol Coming Soon from SeedProd ;)
    …and I really, really hope that they are working on something amazing!

    personally, have wondered about some possibilities leveraging XWP’s Stream data; any case, great article, thanks much :)

    Cheers, Max

Comments are closed.