In 2016, a Stanford undergrad named Joshua Browder, who had recently gotten his driver’s license, incurred several parking tickets. While trying to contest them, he found that the process to fight a ticket was “formulaic.” Drivers had to compose a short letter containing a fixed set of details, like where the violation occurred, and send it to their city government.
Since every letter required the same set of details to be effective, Joshua realized that he could create a standard template for a letter which could be reused by aggrieved citizens like himself to contest parking tickets. He created a web-based chatbot called DoNotPay, which asked users for their details and automatically plugged them into a letter template which could then be sent to the local government. By his own admission, Joshua did not think that his website was all that special. He figured that, at most, it would help a handful of family and friends who were facing a similar problem. In the next few months, however, DoNotPay would go on to help fight over 200,000 tickets across America and the UK, and would be touted by leading news outlets and tech publications as a chatbot lawyer set to disrupt legal systems the world over.
Chatbots are red hot in the digital marketing space. Amidst all this buzz, it is worth taking a step back to understand what all the hype is about. As someone who has written about and built chatbots for the the last 3 years, I am acutely aware that much of this fervor is driven by FOMO, with marketers not wanting to miss out on a new trend.
To use the technology correctly, however, one needs to understand what the benefits of the technology are and, perhaps more importantly, what hurdles the technology faces. Let’s take a look at how you can use chatbots to boost your PPC conversion rate and increase your lead quality.
On the surface, there is nothing wrong with a landing page like the one above. It has colors that pop, a prominent lead-capture form, and a clear explanation of the company’s value proposition. The process of interacting with this page, however, is similar to the process of filling out a letter template. Prospects are expected to trawl through the page to find the information that they need and then fill out a lead capture form all on their own. In an age where marketers are competing with Instagram and Snapchat for attention, even these basic tasks (reading and form-filling) can feel tedious.
Now imagine if the information from the same landing page was reorganized and presented as a chatbot.
What's the Catch?
The picture I painted about chatbots may seem idyllic. Chatbots generate more high quality leads at a cheaper price. So what is the catch? As with any new technology there are some key hurdles that need to be addressed. Perhaps the most important thing to evaluate is Artificial Intelligence.
Working in the chatbot industry I have found that just the mention of the word chatbot elicits ideas of hyper-intelligent computers which are capable of pretty much anything. In fact, I am willing to bet that when you started reading this article, you probably thought that it was going to be about AI. This instinct is natural. Over the past few decades, pop culture has created a mythos around the sci-fi, talking computer which is all-knowing and all-capable (like Jarvis from Iron Man). This has resulted in the emergence of a seemingly inseparable link between conversational software and the AI that often underlies that software.
In coming years, AI will play an important role in the way we run our marketing campaigns. As I suggested in the previous section, the ultimate goal with chatbots is for marketers to be able to recreate their best human sales rep’s knowledge and skills as a virtual assistant who can guide prospects through their buying journey. To provide such a personalized experience, chatbots will have to be adaptive to the myriad queries, and levels of awareness that prospects might have. This adaptivity is only possible if the software underlying the chatbot is intelligent enough to understand what it is being told. Apple store experiences are awesome because the sales rep in front of you listens to what you are saying and responds accordingly.
The issue, however, is that the state of natural language processing today is such that unless you are Google, Facebook, or Amazon you will have to settle for an AI which is not quite ready for prime time.
Don’t get me wrong, AI has been advancing in leaps and bounds and what we have today is pretty good. Unfortunately, it is undeniable that today’s AIs simply do not have human-level understanding so prospects might feel a sense of disappointment when the chatbot they are speaking with doesn’t understand their query.
This is why I urge businesses to temper their expectations. Personally, the vast majority of bots which I have built do not use AI, relying instead on quick reply buttons and a set of custom input UIs to enable interaction. Even DoNotPay used the bare minimum amount of Natural Language Processing if any, relying instead on quick reply options to interact with users:
Chatbots have the potential to be the ideal lead generation tool for marketers. They help reduce costs while increasing PPC conversion rate and lead quality. Of course, to see these benefits, marketers must be wary of not over-promising the capabilities of their chatbots. In the years to come, AI will play an increasing role in marketing campaigns. The promise of an ultra-intelligent sales rep, who can answer every single query, is still years away. While it might be the trend for businesses to implement AI-powered virtual agents into their websites and landing pages, marketers hoping to see real gains from their chatbots today are better off focusing on how the chat interface can help them increase user engagement and supercharge their lead generation process.