Ignoring MRR for the Self-Serve Switch, “Dogfood upon Dogfood upon Dogfood,” and other Lessons in Breaking out of Constraints with Lately AI’s Co-Founder, Kate Bradley Chernis


Honestly, we’ve been eating glass as a group for a long, long time now,” admits Lately AI’s co-founder and CEO, Kate Bradley Chernis (@Kately). In that and other keenly affecting sentences she makes plain that amid downturns or in general, startups are innately, inescapably hard.

But what also pulses through her detailed, frank, and feisty account is that one can, with skill and verve, carve out great gains — a 45% conversion rate with self-serve, for instance — from debilitating constraints.

How ignoring enterprise MRR and prioritizing retention has defined Lately AI’s self-serve transition
Facing the ‘great resignation’ and learning a whole new discipline as a founder
Fashioning the product to focus on what they’re good at and dogfooding to build a “self-fueling marketing machine”
How, in the last 3 years, not one day has passed where they’ve not received a public mention


“It’s an actual hockeystick”

Why decide on this category? Well, no. 1, we go where the fish are jumping. Because that’s smart. Secondly, it’s all about paying attention to our customers.

When we launched Lately, it was designed to be software that replicated a spreadsheet system that I had built for Walmart. Each feature, essentially, was one of my spreadsheets.

As we watched the early customers use it, we noticed that they were all really dialing into one particular feature: the AI content generator.

So we realized that we needed to reconfigure the product, the pitch, how we sold, and everything else — to focus on what people truly liked.

It was clear that writing was the biggest pain point. But we had thought organization was the pain point. Which is about as sexy as a carrot. Or peanut butter (although I guess peanut butter might be sexy to some)!

But it was not only writing – it was also the fact that once you have a piece of longform content, an article you’ve written or a podcast you’ve recorded, what, exactly, do you do with it? Yet another blocker. And, at that time, we only had text generation within Lately.

So, the next steppingstone that resulted from paying attention to our users was around video. What we were doing with AI and text was amazing. But it wasn’t amazing enough.

People were like, “oh, that’s cool.” And we were like, “are you kidding me, this is really hard to do.” They wanted more. Video had already become such a huge medium. So we learned to apply AI to video and snip that up.

The next thing we saw was that very large companies started to have similar demands as very small companies.

Which goes against what most investors will ever tell you: “You can’t market to everybody.” But we discovered that, in our case, this wasn’t true.

And we kept discovering this again and again. In fact, we proved that we could market to small, medium and enterprise the exact same way – constantly in the face of people saying, “you can’t do this.”

But I was convinced that we could. People are people are people. No matter the kind of company you’re selling to. So this year, we made a crazy decision and again did something you’re not “supposed” to do.

Typically, you sell upmarket and then you’re able to sell down the market, after. Salesforce is a good example. Get your enterprise set going and then create a self-service product.

But we realized that in order to secure the enterprise, we had to have self-service in place, first. And the reason is — not too different from Zoom and similar software — people want to pass our product around and collaborate with others. We needed to give them the ability to do so.

The idea is that these first cohorts of users in the company become the social proof to eventually sell Lately to the CMO or CRO. This kind of bottom-up adoption was already happening for us. But it was breaking somewhere along the process. We had an enterprise product in the hands of individuals and they were like, “shit, where’s my team? I can’t use this without more help.” Or, “I need more training.”

Also, on the enterprise side, our $99/mo price point was actually too much – users still had to go through months of paperwork and shenanigans with the company accounting team.

This prompted us to make another fundamental, not-advised switch in our strategy: ignore our MRR. Which was a big deal, as everything we had done up until that point was laser-focused on that very metric.

In fact, our focus on MRR was obscuring our focus on customer retention and on KPIs that actually signaled when and where users were making progress. This was a mind-blowing “aha” for us.

The catalyst behind the insight was a guy named Mark Roberge, the former CRO of HubSpot, who took them from seed to IPO. Mark’s insight unlocked a mindset switch which gave us the permission to return to building what we had initially set out to build – a self-service marketing and sales product designed for any army of one to be an actual army.

After all, the long tail of the self-serve, those are my people! Having been a rock n roll DJ, broadcasting to 20 million listeners for XM and elsewhere, I knew all about marketing and selling to the longtail. The key is: when you sell to businesses like they are individuals, not only do you make the sale but you make evangelists. Exponential win.

Thus we decided to more or less ignore our enterprise product. We placed a “do not fix” rule on bugs unless they were 5-alarm. And we repositioned resources, eliminating our sales team and doubling down on engineering.

One of the biggest questions was: can we market the self-service product the same way we were marketing enterprise? With no paid ads, no cold calls, and no cold emails? With just dogfooding Lately (I’ll unpack this below); all organic? Because our enterprise product has a 98% sales conversion – not a typo.

The answer is “yes.” We have a 45% conversion on my self-service product (industry-standard is 25%-33%).

The key is paying acute attention. Every Monday/Tuesday, me and my core team meet, and we do the hard work. That means Fullstorying customers, every single one of them, by hand.

We take extensive notes on what they have in common, where they are getting stuck, where they are succeeding and we communicate our lessons from those observations to the engineering team.

All of this happens in one helluva spreadsheet, of course! And once we have the data, we are able to examine which conversion knobs we can turn. It’s tough because there’s just so many. The trick is not to fix too much at once because if you break it, it’s almost impossible to know what caused what.

Constant questions include: is it in front of the funnel? Middle of the funnel? Retention? Customer service? Internal workflow? Results? Choosing what to address and when is EVERYTHING. Our resources are limited and we need to decide which will have the biggest bang for our buck at any given moment, to get us where we need to go.

One thing that’s driving me crazy is that I now have the knowledge from my self-service product to put in my enterprise product, but I’m not going to do it. I have to wait. Stay the course.

Like I said, it’s a crazy thing we are doing. We’re flipping the switch. The idea is to transfer all the MRR that has come from enterprise for years and put that onus now on self-service. So we have to wait for those scales to tip.

It’s happening. But my special gift, fellow founders, is always seeing the glass half-empty :-)! In fact my CTO, Brian, is often helping me gain perspective – he’s much more positive. As he put it, “imagine we just launched the company with this self-service product – you did the impossible – in January, when we launched, we did $600 in ARR. In February, $11k ARR. In March, $22k ARR. In April, $44k ARR. It’s an actual hockey stick, you did it!”

Brian continued, “Stop talking about the enterprise, talk about this! This is why we’re all working overtime and on weekends again, because we’re all excited about this. We are doing what almost no company can ever do. Let go of the past.”

I’ll try.

A great reckoning

God is good. Shit happens for a reason. We’ve had an unlimited number of things happen to us that were definitely out of our control. And some that were. All at once. And I’m grateful for every single bit of it.

One thing out of our control was the great resignation, that happened just as we were beginning to think through the self-serve transition. We lost some salespeople, which turned out to be a gift. I was going to have to let them go anyways, now that the self-service product would be taking over their jobs.

This then made it easy to flip resources from sales to engineering. (Our engineering team is made up of my two co-founders, a VP, and our AI lead, and I haven’t been able to pay my co-founders for ages.)

Honestly, we’ve been eating glass as a group for a long, long time now. We are exceptionally good at it. But, it gets old.

The other thing that happened was that I lost a co-founder in the middle of this as well. While we weren’t really paying him that much, the mental investment that was going towards this person was debilitating for years. I didn’t want to see how bad it truly was because I was scared. I genuinely thought I couldn’t live without this person on my team. Turns out, not only could I, but we were incredibly better off for it.

Much of the work he was doing was around accounting, payroll – back office stuff. Fundamentals. So we worked with our advisors and educated ourselves. I’m so proud of my COO, Lauren. She stepped up to this very big, intimidating plate and knocked the living shit out of that ball. We’ve learned so much – it’s been beyond empowering.

Now, I can actually see — this sounds so obvious and it’s embarrassing to say it — but I’m not the only founder, by far, who suffers from this…. Now, I can see – to the dollar, to the dime – every fucking second, where we are, what we’re spending, how to predict and project. I am in control.

I now have the knowledge – where before I was relying upon someone else to have that knowledge. Granted, it’s knowledge that either keeps me up all night or lets me sleep. And I don’t have to ask anyone for it. I don’t have to wait for anyone to answer me. I can just know, instantly.

Talk about a bit of a reckoning…. A Great Reckoning.

Now, me and my team are doing the thing we were built to be doing…. That we’ve been wanting to do for eight years but we didn’t have all the pieces quite in place. Now, we’re finally doing it.

One beautiful aspect is that this happens to all be going on with what’s become our “core” team – other people come and go but these guys and gals are steady, loyal, unflappable. They’ve been with me the longest. They are the easiest team to manage because there’s almost no management required. They know how to run, as a coordinated fleet, on their own. An autonomous symphony.

I love this team. I love them as people. I still don’t know why, to this day, but they’re pathologically loyal! They never seem to waiver. They believe in Lately. They believe in me. Money doesn’t fucking motivate them, which is amazing to me.

Everybody likes money, of course, but it is not a core driver. It’s the fun, the problem solving…. We all want to win. We want to play to win. We want to be on this team that wins.

Making delicious meals out of (product and marketing) offal

I’m constantly considering and reconsidering our collective strengths. Because the entire team is almost always resource constrained. So what can we do with the resources we have? This question played a huge role in our recent choice to make the flip to self-service.

After all, we haven’t really raised since 2019. Instead, I’ve cobbled together a drip-feed of angel money in a couple of loans, which has meant a string of 3 to 6 month runways – as opposed to a single substantial chunk.

Other considerations that drove the road to self-service included the massive strain on customer service and engineering from dealing with Facebook and Twitter, and their APIs.

I mean, if Facebook sneezes, it ruins my day. So the thought was, can we offset the functionality of these platforms to someone else? Someone who does it way better than we do? Like HubSpot Social or Hootsuite?

The answer was yes. With this change, we could focus even more on what we do best – high-performing AI content repurposing. As they say, scarcity fuels prosperity; we’ve been living this for eight years.

With marketing, for instance, there is no marketing team, per se, at Lately.

Instead, we’ve pieced it together with interns, contractors, sales and even engineering. Everyone plays a part. I’m where the buck stops, obviously. But I’ve got all these other jobs as CEO. Hence, there’s no true marketing lead. There’s no one who has the bandwidth to really run the show, to always have the ball. So we do it together.

Again, the thought has been, what can we do with the resources we have? That will have the biggest bang for the buck? Writing, for example, even for me, is hard and time-consuming. To sit down and write a blog or newsletter, I need four uninterrupted hours – nearly impossible. And I’d essentially need to crank that kind of content out daily – wildly impossible.

But I’m great on-air. It’s easy for me. No sweat. I don’t have to prepare for anything, I don’t even have to think about it. My biggest concern before I hop on an interview or a podcast is my hair (which is always a hopeless disaster).

Right around the beginning of COVID, I had done a few interviews, here and there. But the requests started coming in more often – I’m a great guest, I’m a woman, I’ve got a background as a rock ‘n’ roll deejay and I’ve got my hands in AI. I am in demand.

That was the aha. I recognized that I could use this “earned media” the same way I was thinking about “owned media” and repurpose it through our own AI. I created a self-fueling marketing machine that costs zero dollars and almost no effort.

This was my answer to the resource question on marketing: I can certainly spend half an hour, or an hour a day, almost once a day, every day, giving a workshop, having an introspective conversation, and reaching shared audiences.

To take it a step further, we not only broadcast what comes out of the AI on our own brand channels but also then syndicate it across all of our employees’ own social channels. Because I built that feature into Lately. Double dog food!

This is the thing that so many entrepreneurs forget. If you don’t have a huge network, glom on to the others in your network. This is why the culture we have internally at Lately is so essential – not just a team, a fandom. I built a team that’s willing to promote the team’s work. Because I can’t possibly do it by myself. Together, we’re stronger.

What’s more is that everyone on the team is a social beast. That’s their nature. They are already doing it. It’s effortless. It’s part of the DNA of how they exist in the world. In fact, I’ve learned to hire that way, having made a few mistakes in that arena.

If a salesperson I was about to hire had a LinkedIn page without a custom URL, someone on our team now flags it, “dude, you know this person is not going to sing for you, what the hell are you doing?” That’s now part of how we inherently think.

We also have a Slack channel called #sharingiscaring. Every time anyone writes about us on social media in any way, we take the link and we pop it in there. And the entire team is required (and wants to) share it, retweet it, comment on it. This is how we organically boost - an essential component to our lead generation.

In addition, our social media contractor keeps an eye out, like a hawk, on every new customer coming in on Stripe, and she finds them on LinkedIn or Twitter or whatever platform they’re on, and starts to reshare their content. We also try to participate in whatever they publish.

We take links that they post, put them through Lately, and then share them with the team and ask, “hey guys, our friend, Jen McFarland was featured in an amazing book this week. Hop on. Let’s congratulate her.” We ensure they’re aware of who we are and that we’re on their team. Again… Together, we’re stronger.

It’s dogfood upon dogfood upon dogfood.

All the way through.

All of that together is how we get that 98% sales conversion on the enterprise side. Once we kick off promotions, we watch to see who likes, comments, and shares – as these people are warm leads, because there’s some recognition or social proof around that activity they did.

Then by the time we get them to a demo, they’re essentially hot.

As I mentioned above, the big question was, will this work on the self-service side? Because obviously there’s no demo there.

You already know the answer :-).

“Don’t make a sale, make a fan”

So many founders are afraid of putting themselves on social media because they say, “I don’t want to muddle the brand.” Well, duh. You are the brand. There’s no escaping it. Investors invest in you and not the product, for a long time in the beginning. You have no choice. You’re on the cover of your own magazine and that’s just the deal.

I also constantly think about how to put other people on the cover. And it’s hard. I try to get interviews for all our co-founders and senior leaders. We also try to put Lately’s customers on the cover of that magazine as often as we can, too. So letting them run our weekly live webinars, for example. After all, we’re all in this together. Sound familiar?

Think of it like this: when you’re in a room and you’re networking, the way to get the most bang out of the buck from that, is to not be the person who is the brightest light. Instead, you want to be the person that makes other people feel like they’re the light.

Lift them up. That’s how you create evangelists. It’s subtle but very powerful. It is also the Long Tail, which I’m a huge fan of, again thanks to my former career in radio. Back then the idea was: “Don’t play a song, create a vibe.” At Lately, it’s: “Don’t make a sale, make a fan.”

One of the unexpected results has been that in the last three years, not one day has passed where somebody hasn’t said something nice about Lately publicly on social media. We even enjoy regular public praise from those who’ve churned.

We also drive a thousand leads to our website every week with this method. The FREE method. It’s a lot of work. I know that. But the residual, exponential bang-for-the-buck will carry us forever.


Related reading from the Relay archives:

GoSquared’s co-founder, James Gill, on the wins that accrue (and the challenges that surface) when a team uses its own product
Lou’s co-founder, Rachel Pardue, on how adopting the self-serve path brought in their first 1000 customers


This account rings with perseverance, @kately!

One would think that there’s a limit to what founders can make of constraints but reading about such journeys is a great reminder that there isn’t.

A question I often ask any founder, but especially those who’re in that phase where they’re almost doing everything, is, what helps? What are some personal routines and practices that help you stay sane with all the anxieties of scaling through such phases?


Tequila helps.