How Benn Stancil’s newsletter became the cure for data biz ‘thought leadership’

The backstory is a little bit funny.

In the very early days of Mode, there were three of us: our CEO, who was off talking to investors, our CTO, who was chained to a desk building the product, and me. My job was going to be PMing and working with customers, but we didn’t have any customers at first because we didn’t have a product yet. So the thing that I started doing was writing blog posts for Mode’s corporate blog.

The very first blog post was a blog post about Miley Cyrus, one of those “data-driven looks at pop culture” things, kind of like what FiveThirtyEight did for a while. I did posts like that for seven to nine months, and I remember making a distinct choice to not write at all about data tools or give our “5 things you should learn for your first data hire” or whatever because that kind of content was out there already. So I wanted to write stuff that was fun for data people, stuff that wasn’t going to tell them what to think, instead, hopefully, it’s just entertaining.

When I started moving into different roles at Mode, I left the blog to become more of a typical place for product releases and other corporate and brand content. About eight years later, our company had grown and we added people to fill the jobs I had been handling, so I decided to devote time again to these kinds of pop culturey data blogs. But since the Mode blog no longer felt like the right place, I started putting them on a Substack instead.

I think the Substack eventually became more about the data world itself because I wound up having a long backlog of various data world things I had thought about while I was away from writing. It wasn’t really my intention to have the newsletter focus on those, but over the course of the first handful of posts, I wrote a couple of things that were kind of like that, and those tended to be the things that people read.

So, yes, I was originally going to talk about Miley Cyrus, but then I found myself talking about data pipeline tools.

I actually took math classes in college because I didn’t want to have to write papers. I just loathed writing. So it’s a little bit of a weird place that I’ve ended up.

But I did gain some experience at it in my first job after college, which was at a think tank. I had to ghostwrite economic policy papers. So that’s where I would’ve learned it. I never did any writing in Silicon Valley until the Mode stuff.

“I think the silly blog posts [we did at the beginning of Mode] actually were more helpful than we probably gave credit for because they were some differentiation. They helped us establish a brand, and brand matters.”

After a bit, I did find more of my style. Like, there’s going to be a bunch of footnotes and random links to YouTube videos from the 90s. What emerged was just what worked in my mind and what entertained me.

When I first started writing at the Mode blog, it was meant to be lead generation—kind of. 

We wanted to be participants in the conversations that we thought data people might like to have. Like, “We think this is fun. If we all think it’s fun, let’s hang out and talk about it together.” So it was a brand-building exercise, and we found that there were some legs to that from a marketing perspective.

I think that we should have leaned on that even more. As we got customers and I started doing customer support, the company and I had to pivot to focus on those things—the real business stuff as opposed to silly blog posts. But I think the silly blog posts actually were more helpful than we probably gave credit for because they were some differentiation. They helped us establish a brand, and brand matters.

With the Substack, we thought maybe we could revive it. Even though the newsletter is not fully associated with Mode, when you’re a founder, you can’t fully dissociate.

Because the topics now are more industry-related, I think the newsletter has generated more conversation in the niche Silicon Valley data world. So now Mode and ThoughtSpot can be a part of those conversations, which probably has a business value to it.

But we never tried to directly monetize or put conversion metrics on any of this.

The posts broadly that seem to do better are the ones about specific tools or companies. So a take about data company X tends to do a little bit better than some kind of vibey, softer thoughts on data best practice. And that kind of makes sense. The stuff about specific companies has sharper edges and it feels a little bit controversial or whatever.

But I do think you have to be a little careful about chasing those numbers and paying too much attention to how much people read one post from another. I think if you become too much of a slave to that, you end up chasing engagement and not chasing quality ideas.

Yeah, there’s this faux thought leadership stuff of, “Here’s how we solved this problem. Oh, you’re struggling with cybersecurity issues, too? By the way, we happen to sell a cyber security solution.” I’m sure it works, and nobody’s going to blame you for running an ad. And if you call it an ad, I think that’s fine.

But in terms of saying things that are interesting, you run out of gas pretty quickly if everything you’re doing is like, “Oh, it needs to be a funnel to this other thing.” You can only say so many interesting things about why someone should buy your product. At some point, you’re just recycling a bunch of stuff that’s not useful. So you’ve got to allow yourself to step off the path and talk about whatever you want if you want to actually make it interesting for any sustained period of time.

“If you become too much of a slave to [numbers], you end up chasing engagement and not chasing quality ideas.”

As for why more people don’t do that, I guess I just spent the last several minutes trying to justify why it’s useful for me to do it. And so companies and people would very reasonably look at it and be like, “Why would I do that? Why do I need to be interesting? Why not write the posts that we know work as ads and generate leads?” There’s a very clear cost to writing these blogs, which is it takes some time—it’s work. The benefits can be good, but it’s not directly generating leads or whatever.

On the corporate side, people have read my newsletter and asked me to speak at conferences, which is good for the profiles of Mode and ThoughtSpot. I’ve also had prominent customers and prospective customers request to meet with me. “Hey, come and talk to our team about this stuff. They read your thing and they want your thoughts.” So there’s just a sort of reputational element I can use for us, whether to help close a deal or to extend our companies’ brand reach.

And the conversations the newsletter has opened the door for me to be a part of are two-way. They’ve helped me learn a lot of things that I’ve brought back into trying to make a better business and a better product.

On the personal side, most of the stuff I write about are things I was going to sit around and think about anyway. My job has been to think about BI tools for forever. You end up having a bunch of thoughts about it. I’ve seen someone say if you have an idea and you write something about it, you can finally be allowed to forget that idea and move on to the next. I think that is sort of true. Writing can sort of clear the brain space of the nagging itch of whatever is on my mind.

Maybe not just my writing alone, but there is a blogosphere, or whatever, that can help drive conventional wisdom and where things are going to go. It’s not complete groupthink: I just explained in one of my recent posts that I think everybody in data is sort of working in a rough direction, but it’s still somewhat scattered. Over time, though, the conversations and disagreements people have will probably start to turn into some established way of thinking or generally accepted architectural decisions.

For example, at Mode, we were involved in that when we decided to write a lot about SQL. SQL has become the standard language for how we’re doing things with data now, but at the time Mode was getting started, Hadoop and MapReduce were supposed to be the next big things. We sold a tool that was SQL-based, so we advocated for it because it was good for Mode, but the truth was that we created Mode because we thought SQL was just the right direction to go. And so I think our advocacy, among other leaders, helped make it so everybody eventually ends up at, “Yeah, I think we’re going to agree that SQL is pretty good.”

You have to give yourself a schedule and stick to it. One of the things that people say holds them back is that they don’t know what they’re going to say or write. In my experience, you’re not going to have grand epiphanies very often where it’s like, “Oh god, I have to write this thing.” Most people perpetually have a bunch of half-baked ideas. When you have a deadline, you force yourself to take those half-baked ideas and turn them into something, and it’s in the process of baking those ideas that they become good.

“The conversations the newsletter has opened the door for me to be a part of are two-way. They’ve helped me learn a lot of things that I’ve brought back into trying to make a better business and a better product.”

But like I was getting at earlier, all this is very hard to do with an agenda. If the idea-baking process takes you in weird directions, landing on things unrelated to what you set out to talk about, you have to go with it. So if you know you have to talk about lessons of running a sales team, and that’s the agenda, then that’s the only thing you’re ever going to talk about, and pretty quickly you’ll run out of things to say.

Alternatively, sharing your lessons learned from running a sales team will get readers, and might actually get you more leads, too. But you can just ask ChatGPT to write that and it’ll be fine.

This interview was edited for brevity and clarity.