SPECIAL NOTE: For this post, I am turning over my blog to another one of my fascinating podcast guests. He writes about shifting leadership capabilities with different business models. I am sure you will find this piece interesting. Let us know what you think . – Todd
By: Dennis R. Mortensen, CEO and Founder of x.ai
For the two decades leading up to the founding of x.ai, I ran (with many capable people) several B2B companies. We were fortunate to have three good exits, so if we learned anything, it was how to keep enterprise customers happy ;-). However, in conceiving of x.ai, we understood immediately this would be a B2C2B (Business-to-consumer-to-business) SaaS business, with a price point around $20-40 per month, per user.
What goes almost without saying is that most B2B companies deploy a top down selling strategy; you identify the target enterprises and accompanying individual decision makers upfront, and you try sell to them over many months, if not years, using elaborate sales funnels that help you move forward. Typically there only so many potential customers you can attack, sometimes only hundreds (or maybe thousands), but rarely millions of these.
B2C2B companies are bottoms up; if you succeed, you get individual customers to find your product, onboard themselves, and become evangelists within their organization and essentially sell in for you. (Slack has executed this strategy brilliantly.) And you’re aiming for millions of customers. That seems efficient and highly desirable, but B2C2B businesses come with all sorts of other hazards; the main and most obvious one is that it’s much harder to smooth customers’ feathers when things go wrong.
How to recover customer trust when the **** hits the fan
The day I realized that I would have to adapt to a new order was June 16th, 2014. For context, x.ai makes an AI assistant—Amy or Andrew Ingram—who schedules meetings for you. You just cc firstname.lastname@example.org, and she’ll take over the email ping pong to get a meeting on your calendar. Up until then, we had a very small closed beta, and all of the beta users were friends. Like enterprise customers, I had a unique and personal relationship with every single one of them.
But on that fateful Monday in June, we sent out about 300 beta invites to people I didn’t know. It didn’t go well, and it was entirely our own fault.
The same day we invited people into the beta off our waitlist, we were joined by Laney (our first AI trainer), and we launched v1.0 of our annotation console, which is the proprietary software AI trainers use to annotate data extracted from scheduling related emails. So we were essentially training a new employee and launching new software, with zero testing, and under these conditions, we still had to make sure our AI assistants scheduled strangers’ meetings.
As you might imagine, my inbox blew up with complaints, which I knew I couldn’t respond to fast enough. This is rarely, if ever, a B2B problem, since in those businesses you sign few customer at a steady rate.
I had to quickly figure out how to communicate to many people at the same time, intimately, which sounds like an oxymoron because it is.
Transmuting individual customer love into mass kindness
At all of our prior ventures, I knew all of our customers personally. I knew a lot about each one—names of their wives or partners and kids, if they had them, career history, who they knew at other organizations—and I didn’t think twice about entertaining them at my home. But starting on June 16th, we didn’t even have a name for many of our customers, just an email address. For x.ai to succeed, we had to figure out which customer care tactics to translate, and we had to be super vigilant because doing so poorly would mean unhappy customers and/or slowing us down too much.
And yet, good things come from having to love all of your customers. It means I don’t just think of our customers as user IDs, but rather as real individuals, even if I can’t get to know them personally. If someone named Kenneth doesn’t like Amy, I can now assume there are 10K more of him, with their own idiosyncrasies, and now it’s on us to figure how we can address all of them to change their minds.
With a little bit of trial and error and a lot at stake, we quickly figured out how to do this. The main thing I personally did was ramp up my activity on social media, Twitter especially. I found that if I was personable and accessible, rather than just tossing out a FAQ link, it had an obvious positive effect on frustrated beta customers’ attitudes, in addition to helping them solve their specific problem. And to this day, I remain very active on social media, interacting with easily a dozen or more customers every day.
Team structure changes too
Besides having to relearn how to interact with customers, we obviously had to think about the organizational structure very differently. We had a traditional enterprise sales team at our most recent prior venture, Visual Revenue, focused on 172 named accounts. That’s it. Conversely, here at x.ai, we see all 90 million US knowledge workers as potential customers, and that’s not counting any international markets.
So we’ve had to build out a Customer Experience team focused on acquisition, retention, and customer support, all of which have to be done at scale. The team members and mix of skills reflect that. (We will also end up building a sales organization but we rely on the CE team to generate all leads.)
And of course, our current acquisition tactics belie the differences between a B2B and B2C2B company. At Visual Revenue, events drove most of the activity at the top of our sales funnel, and at the top five conferences each year we’d be able to meet with representatives of nearly all of our customers. Not so today. Events are nice, and certainly spread the word, but at any given event we touch only a fraction of a percent of our potential customer base. Instead, we are focused on product marketing, referrals, paid channels, and content.
More customers means more awareness
On the flip side, there are ways that the product itself makes this shift a lot easier to execute. In the past, we had to hope that one of our customers, say a front page editor at the NY Daily News, had such a good experience that she would recommend Visual Revenue to her counterpart at the NY Post. We had very little control over that word of mouth, aside from delivering a great product.
Today, our customers expose our product to others just by having Amy or Andrew schedule their meetings. It’s impossible NOT to interact with Amy or Andrew if you want to meet with one of our customers (unless you already are a customer). That’s rare and powerful, and it certainly makes it exponentially easier to ramp up awareness of the product.
We’ve seen the impact in both organic growth and earned media. At Visual Revenue, our entire customer base was made up of media companies (including every major US newspaper and many familiar tech publications). Over the three years we ran the company before being acquired, we received 15 maybe 20 credible mentions in the press. And we fought hard for every single one of those.
In less than three years of operation, x.ai has received an order of magnitude more mentions in the press, with countless stories in major outlets. I believe this level of exposure has come from three sources:
- a genuinely viral product
- being on trend; there’s a ton of curiosity about AI right now and few products in market that regular people can interact with (or at least know that they are doing so, since much AI is behind the scenes)
- the care we give to every single customer, blogger, fan, and reporter who shows some interest in writing about us
For me, it’s been incredibly rewarding, if occasionally nerve wracking, to move from the closed and intimate setting of a B2B company to the wilds of B2C2B. I’ve learned that showing your customers that you care, even when that means hanging out on Twitter at 11PM on a Saturday night trouble-shooting, is well worth the investment.
Dennis is the CEO and Founder of x.ai. He’s a pioneer and expert in leveraging Data and a serial entrepreneur who has successfully delivered a number of company exits on that theme. He’s also a frequent speaker on the subject of AI, intelligent agents, and the future of work. A native of Denmark, Mortensen currently calls New York City his home.
x.ai makes an artificial intelligence personal assistant who schedules meetings for you. There’s no sign-in, no password, no download. All you do is CC email@example.com into your email conversation and she’ll take over the tedious email ping pong that comes along with scheduling a meeting.