Aren’t chat bots overhyped at the moment?

Here is the audio file for this answer >>>

Premise – The hype of chatbots.

  • Are they just a fun new toy, or are they a serious weapon for CS, Ads, Recruiting…?

Use cases:

  1. Customer service
  2. Lead generation
  3. Recruiting
  4. Serving content

Main differences in use cases:

  1. Fully-chat-bot powered 24/7
  2. Vs. A conversation starter…
  3. Vs. After-hours only with an option to schedule something with a human the next day or continue with the bot experience.

A few big tools to checkout:

  1. – Cognitive App Platform – for creating and deploying custom bots into other applications (like Intercom)
  2. Manychat / Chatfuel – for Facebook Messenger chatbots.
  3. Drift – Conversational Marketing Platform – for converting web traffic.
  4. | Build Conversational Websites – for creating bot-based landing pages, or embedding bots into content.

My view – The latter two use cases are the ones I suggest.

  • Cost/benefit discussion is important to have as well when you put thousands into a chatbot just to have it cause more harm than good.
  • I believe they are useful for certain use cases where you are having to either pre-screen or provide some low-level assistance.
  • But, in the circumstances where questions can and will arise, and/or the product you are offering to this new audience is expensive, I always recommend having a human during business hours at the very least.
  • Now, that leaves us with off-site bots for things like FB ads… which is why Nolan is here to discuss.

I herewith Nolan because he is much more experienced with chatbots deployed for lead capture on Facebook. He recently wrote an article (found here) on how chatbots will shape marketing in 2018.

Nolan –

“Hello again Marketing Automation listeners, I’m Nolan…. , and I am experienced using chatbots for both B2B & B2C cases….

Bots may be overhyped in the sense that they are the “next big thing”, but part of that lies in the sense of Silicon Valley trends – they’ve learned not to overlook new technologies, and tend to be trend-setters, in some cases too early.

Marketers like Alex and I, as well as the product developers, are the ones who will ultimately create the value and increase the effectiveness of chatbots. They have to meet those expectations in order to stand up to the hype, and we’ll go over some use cases and other ways that they are delivering later in the episode.

Nolan’s view – At the very least, use them a way to collect emails/phone numbers if you are bearish on bots, or otherwise use the bot as an augmentation of your existing marketing channels.”


What are some good opportunities for chatbots to add value in B2B applications?

Here is the audio file for this answer >>>

Use cases:

  1. Customer service (i.e. in-app chatbots during off hours, post conversion initial dialogue before a real person comes to the chat)
  2. Lead generation (i.e. frontend visitor bots, in-content bots, bots in FB ads)
  3. Recruiting (i.e. on an application page, instead of having traffic fill out long forms, you have them chat through a bot)
  4. Serving content – My former heads of tech and product at Curiosity Makes You Smarter created a great chatbot that essentially acted as a content serving machine – ask
  5. Internally – for speeding up or better managing processes. We won’t talk much about this, but you can imagine the Slack bots.

But… because this is the marketing automation podcast, we will be focusing on the led gen use cases.

Section 1 = Frontend Site Chatbots

Intro to this section – This is where I have used chatbots the most. Tools like Drift have come about for this very use case – how to create a better experience for users on your site using chat and chatbots.

Use case A – You have a blog with a ton of traffic, you want to bring more of that traffic to a sales/pricing/demo page, but you do not have enough sales support bandwidth to power a chat. Answer: Drift – Conversational Marketing Platform

“Nolan’s comments on this use case: From here, you can at least generate leads that can then be nurtured through automated means, so that any human energy exerted is well-spent.”

Use case B – You require visitors to fill out a lot of fields in order to convert. This could be an application, like a lease application. You can set up a bot to gather the same fields you need from a long form but in an engaging and fun conversational manner. For this, I suggest – Cognitive App Platform + Intercom

“Nolan’s comments on this use case: “Studies have shown that more input fields will result in lower conversion rates, so getting the info you need this way is much more fluid.”

Use case C – Recruiting off-hours or for low-level jobs. is doing interesting things for this use case.

Use case D – Serving content. This can be done via Intercom + – Cognitive App Platform.

“Nolan’s comments on this use case: Cross-channel is key for content marketing, so anyway that you have to publish to a user’s preferred channel will increase your overall yield for your campaigns as a whole.”

Section 2 = Chatbots starting in Facebook

Intro – Chatbots which deploy in Facebook… These can begin when someone lands on your business page, join a messenger newsletter of sorts or click on one of your messenger ads. We will go into more detail, but first, what is your high-level state-of-the-union on messenger ads with a bot deployed for B2B purposes Nolan?

Nolan – “Quick high-level what’s happening with chatbots deployed in messenger ads for B2B uses:

  • Basketball Analogy (quick back and forth, high CTR & Open rates)
  • See exactly where prospects are dropping off since most of the marketing funnel can reside in a chat flow.


(24+1) rule – dribble rule in basketball

subscriptions (not promotional)

Sponsored messaging (also ad inventory in the messenger inbox) – no restrictions”

Use case A – A FB Messenger ad to a webinar – where a bot helps convert clicks to webinar registrations in the messenger bot – the user will register for the webinar (even submit payment if you choose) right inside the messenger bot. Bots act as a tesseract to skip steps (visit website, check email inbox) and increase conversions.

  • Example: Service businesses

Use case B – Local businesses offering a service to a B2B audience who are active on Facebook to promote their services. This could be anything focused on any brick and mortar business, but the best target audiences for this approach typically work in the restaurant industry, real estate, anything health-related…

  • Example: Merchant services targeting restaurant owners/GMs.
    1. Deploy messenger ad which pre-screens and gauges interest, then the click-through includes a parameter which tells intercom to continue the conversation on the website.
    2. Requires: Intercom and a bot running in Intercom (I mentioned I like – Cognitive App Platform)
  • Example: Give away contest (gift cards, etc.) using the comment-to-message feature.
    1. Print shop giving away business cards.
    2. Restaurant with gift cards:
    3. Collect emails in order to enter contest – follow-up with a message to place an order after entry, and add entrants to email marketing
    4. Tool – Manychat

Use case C – Content Marketing via subscription messaging – don’t include anything promotional, and be sure to elicit some form of interaction.

  • Example: Marketing agency wanting to share guides/ebooks on a particular subject.

Final thoughts:

Bearish view = Chatbots can do more harm than good if not deployed correctly and for the right use cases. I believe their will continue to be a place for them in the funnel, but a human should intervene when (a) the conversations could become complicated, and (b) when the stakes are are high – like an expensive product/service which relies heavily on CS.

Bullish view = If email is the most powerful channel, and if chatbots can outperform email, then Facebook chat bots may become a great channel to use for B2B.

It mainly depends on how fluid the user experience is; they will ultimately side on the path of least resistance, which is something that applies for your product/service as a whole, so be sure to take the time to really test out your bot and make any necessary adjustments.