Practical Uses of Artificial Intelligence for Midmarket Marketers

In Technology by Tom PickLeave a Comment

Practical Uses of Artificial Intelligence for Midmarket Marketers

Artificial intelligence (AI) is being embraced by enterprise marketers. But a combination of skepticism, fear, and confusion appears to be holding back adoption in the midmarket. Nevertheless, there are practical applications of AI that marketers in small to midsized companies can adopt today, without significant risk to or disruption of their current operations.

AI use is picking up rapidly in large firms. According to Forbes, 38 percent of enterprises are already using AI in some form, and that figure is projected to reach 62 percent by the end of 2018. Another studypredicts the size of the AI market, $8 billion in 2016, will grow to $31.5 billion by 2025.

But the story is different in smaller firms. For example, according to Gartner, “By 2020, 30 percent of all companies will employ AI to augment at least one of their primary sales processes.” That means 70 percent still won’t be applying AI to sales even three years from now.

Additionally, recent research by Demandbase found that while “80 percent of all marketing executives believe AI will revolutionize marketing over the next 5 years… only 26 percent are very confident they understand how AI is used in marketing and only 10 percent of marketers are currently using AI today.”

And AI doesn’t make the list of the top marketing or technology challenges faced by small to midsized businesses today.

AI Skepticism in the Midmarket

Though marketers in small to midsize firms are intrigued by the possible applications of AI, many remain wary, for a number of reasons including:

Implementation challenges. Chatbots have been called the “gateway drug” for AI; the simplest way to get started with this technology, and where most small and midsized companies are likely to go first. Yet, even proponents of chatbots acknowledge such efforts are not “plug and play.”

Further, VentureBeat notes fostering engagement with chatbots requires human efforts to “Ask (for) feedback from your users, make changes, ask (for) more feedback, and improve.” And even then, properly assessing engagement is difficult.

AI-washing. Similar to greenwashing (in which companies exaggerate the environmental-friendliness of their operational practices for business benefit), marketing technology vendors may now be applying the “AI label” a bit too promiscuously.

As a recent post from Intellyx explains: “There’s no question we’re in the AI-washing phase of the AI revolution now. It seems that every vendor, from IT operations management to business intelligence to digital marketing, is now using AI under the covers…The problem is that there are many different types of AI, from simple machine learning to much more complex deep learning and various types of cognitive computing. Today’s vendors are in large part doing the easy stuff.”

An article from Martech Advisor adds, “AI is now nothing more than a marketing term used for any software application displaying even the most rudimentary intelligence…AI (has been) transformed from an almost unreachable goal into to a cute acronym designed to sell to the Fortune 500 and anyone with deep enough pockets.”

And per Philippe Botteri, a partner at global venture firm Accel, “A lot of…companies have been doing machine learning and (that work) has been rebranded as AI. It doesn’t mean it’s fake. It’s just that the term ‘machine learning’ is not as sexy to put on your web page.”

Confusing complexity. Related to the point above, Information Age recently discovered that confusing vendor messages and a lack of understanding were core problems midmarket AI use, stating “AI adoption in marketing is being hindered by marketer’s understanding of AI, with 40 percent thinking they are already using the technology.”

Furthermore, factors such as over-reliance on outside agencies for marketing strategy, and difficulty integrating different tools and data, add complexity to any potential AI implementation.

Talent shortage. Beyond its simplest applications, exploiting the capabilities of AI requires specialized talent. Business2Community reports, based on research from A.T. Kearney, that individuals with AI experience and skills are hard to find: “Two-thirds of companies answered that they can’t hire enough people who can generate insights from corporate data.”

Smaller firms will likely have an even more difficult time than their Fortune 2000 counterparts in identifying people with the right skills—or being able to afford that talent if they do find it.

It’s hard. Beyond the simplest uses of AI, as in chatbots and certain features of Google’s search engine, the deeper and more valuable applications of AI are proving to be difficult and slow.

Even a company with the AI prowess and deep pockets of IBM admits to struggling. The New York Times reports that after its Watson supercomputer won the quiz show “Jeopardy!” in 2011, expectations were high for new conquests such as curing cancer.

Optimism ran ahead of practical reality. As the Times notes, “Medicine proved far more difficult than (IBM executives) anticipated. Costs and frustration mounted on Watson’s early projects. They were scaled back, refocused, and occasionally shelved.”

Experts from universities, startups, and IBM itself note that while significant progress has been made in areas like voice recognition and machine vision, tackling higher-level tasks has been more difficult than anticipated. AI may yet prove capable of performing some of the wondrous feats predicted, but such developments are a still a decade—or several—away.

Proving ROI. Even for less sophisticated tools, marketers in smaller firms report they often have difficulty getting budget approved for technology purchases because of the difficulty in demonstrating the return on investment.

The Business2Community piece also points out, “It’s difficult for a business to measure and predict the returns on investment in (AI)…results for improved customer satisfaction and revenue would not be visible immediately.”

Job killer. Seeing headlines like Robots will destroy our jobs – and we’re not ready for it, marketers are understandably nervous. For now, AI and robotics are replacing humans mostly in manual, repetitive jobs. But how long until their capabilities move upstream?

The panic may be overstated. Small Biz Daily, among other sources, contends that robots are likely to make marketing jobs more interesting rather than take them over. But whatever the net impact on job numbers, there’s no question AI will change the roles of marketers and demand new skills.

Collateral damage. Finally, there are more peripheral reasons for skepticism. Self-driving cars aren’t directly related to the types of AI systems marketers utilize, of course. But they are one of the most high-profile examples of AI in use. So problems with autonomous vehicle technology can raise concerns for other applications of AI.

A reasonably well-trained 15-year-old human can drive a car safely in most situations and conditions. But even after seven years of research and development, driverless cars still crashComputerWorld even recently questioned whether cars will ever really be self-driving.

Practical Applications for AI in the Midmarket

Despite the very real concerns and challenges, however, marketers remain optimistic about the potential for AI. According to destinationCRM, “80 percent of marketing executives believe artificial intelligence will revolutionize marketing by 2020,” bringing benefits including richer customer insights, better campaigns, and greater efficiency.

Many of the concerns noted above, on closer inspection, are based on (justifiable) skepticism about some of the wilder claims about AI. Pragmatic marketers in small to midsized firms understand that while today’s entry-level AI technology isn’t quite as revolutionary as its hype, it does offer practical areas for exploration, such as:

Interactive voice recognition (IVR). Thought chatbots are often described, as noted above, as the “gateway drug” for AI, that label might better apply to IVR. This technology, based on speech recognition, has been in use for years and is widely deployed (as anyone who’s called a customer support line can attest).

However, the technology is now spreading to mobile applications, where, as a recent Forrester report notes, “a hands-free interaction is often preferred.” These applications are still in their early stages, but may offer marketers a way to apply IVR technology in a way that delights rather than annoys customers.

Per Forrester, vendors doing interesting work in this space include Nuance Communicationsand Calabrio.

Chatbots. Chatbots are spreading rapidly in applications like first-line customer service and website visitor greeting (answering simple questions and providing recommendations).

This is a practical way to introduce AI in a small-to-midsized organization, and the savings in customer service costs alone may be sufficient to build the business case.

Marketers just need to choose vendors carefully and understand the limitations as well as the potential benefits of chatbots. As noted on, developing a high-quality bot is difficult, and “Despite the advances in AI technology, today’s bot systems are still not able demonstrate human-like understanding of open-ended language.”

Chatbots are best deployed as a first level of contact, with the ability to answer simple questions but also a quick and clear escalation path to a human when needed.

Reporting. Early iterations of automated reporting tools often failed to live up to their promise, doing little more than converting Google Analytics data into (somewhat) more digestible charts and graphs.

But newer tools like PaveAI take the next step, analyzing millions of possible data correlations in just minutes. The results still require human interpretation and judgment, but they give marketers much more to work with than just raw visitor and session data.

Sales operations: Applications of AI targeting sales effectiveness are primarily used at the enterprise level. But a few such tools may be useful in midsized firms as well.

For example, Econsultancy has suggested products like (for analyzing recorded sales call conversations); the Emarsys Marketing Cloud (ecommerce discount personalization and optimization); and LeadGenius (B2B lead generation) may be worth evaluating.

What Marketers Need to Know

Given that product features are often easy to duplicate, and the Internet has made pricing transparent, customer experience is the ultimate battleground for marketers.

AI technologies can provide value to small and midsized businesses to the extent it can help improve the customer experience; for example, by personalizing content and enabling customers to interact more naturally, through voice, with websites and apps.

But those technologies can have the opposite effect, as Carla Johnson explains, if not chosen and implemented carefully:  “79 percent of marketers said they believe customers are ready for AI and are either ‘excited or very excited’ about chatbots. Half of customers, however, say they’re ‘very unexcited or somewhat unexcited’ about chatbots.”

Internally facing AI platforms should improve customer insights and measurably reduce costs. Externally facing systems should be chosen and rolled out to provide an experience like Siri or Alexa rather than an endless automated phone tree.

As summed up in Forbes, “Today, more and more companies are realizing that true competitive advantage lies in creating an engaging customer experience—one that is personal, fast, easy, and useful.”

Applying AI to customer data can enable marketers to respond to customers in a more personalized, and faster, way than humanly possible. But only if it’s used that way.

Photo Credit: Diogo Saraiva Flickr via Compfight cc

This article was first published on V3Broadsuite.

Tom Pick is a digital marketing consultant, working with Kinetic Data, a provider of enterprise service request management, workflow automation, and collaboration software. He writes about content and social media marketing topics on the Webbiquity blog.

Leave a Comment