AI IRL Podcast Episode 28: Conversational AI That Doesn’t Pretend to Be Human
It’s usually completely obvious when you’re speaking to an AI, right?
Actually, that’s exactly the way it ought to be.
Within standard AI, Arjun is building a B2B conversational marketing platform that will enable marketing and sales to engage one-to-one with their prospects and give them a very personalized real-time experience supported by AI.
“There are several things that we are doing that can be so much better with the right use of AI,” Arjun said.
“The first thing that a company should look at is, how can I make my data better?” –Arjun Pillai
What to Do with All This Data
On average, 2% to 3% of the data in CRM get outdated every month.
Which means 25% of all your CRM data is getting outdated every year.
“When was the last time you updated the data?” Arjun asked.
In the last decade, the amount of data that marketers and salespeople have at their fingertips has exponentially increased.
You 10 years ago: I need more data about my customers and prospects.
You now: Seriously, I have a ton of data. What do I do with it all?
We may all be drowning in data, but at the same time, there isn’t a drop to drink. In the sense of getting insight out of the data (which is where AI comes to the rescue).
“Databases are just containers if you think about it. It’s only as good as the data that goes into it,” Arun said.
“The first thing that a company should look at is, how can I make my data better?” he said.
“The world has gone 100% real time. The only conversation still stuck on asynchronous methods is company-to-customer conversations.” –Arjun Pillai
How to Leverage Data with AI
So, two individuals speak on SMS and on Slack and WeChat or Messenger.
“The world has gone 100% real time,” Arjun said.
“The only conversation still stuck on asynchronous methods is company-to-customer conversations,” he said.
Basically, B2B marketing is still stuck on emails.
You go to a website and sign up on a form. They put you into a nurturing campaign. They bombard you with emails.
But people are on your website in real time, and you’re just pushing them away.
“That’s like asking somebody to walk into your store and saying, ‘We don’t have time for you now. Sign up on that form, we’ll get in touch with you on a call five days after,’” Arjun said.
Nobody in their right mind would put up with that.
AI is a good middle line here.
- Assess the prospect
- Understand the prospect
- Qualify the prospect
Using internal data or third-party data, AI can take over 50% of the conversations and send the relevant ones to marketers and salespeople.
So that marketing and sales can have those one-to-one conversations that allow them to attend their prospects in a personalized way.
“A bot can come in and make sure the conversation keeps moving down the funnel.” –Arjun Pillai
A Human-First Approach
A conversational AI that will overcome the problem of being mainly rules-based is going to be designed with a human-first approach.
“We assess the prospects when they are on your website based on a lot of the signals that we have,” Arjun said, such as internal systems, reverse IP, browsing pattern, etc.
“Then we essentially get an alert to the salespeople saying, ‘Hey, Amy from SAP is browsing on your website and she just jumped on the pricing page. You might want to start a proactive conversation,’” Arjun said.
Another good strategy. If a live salesperson is not available at that time, the bot takes over.
And it’s plain that the bot is a bot.
“We are not trying to camouflage the bot as a human,” Arjun said.
The bot’s role is to tell Amy that Aaron the account executive isn’t available at the moment. I can help you set up a call with him.
“That is where a bot can come in and make sure the conversation keeps moving down the funnel,” Arjun said.
If you don’t use iTunes, you can also find every episode here.