Introducing Mattermark, the Deal Intelligence Company

Part of my job as an early stage CEO is to protect my company from being put into a box too early. Leaving us unsorted makes investors, customers and even employees a bit uncomfortable. It’s human nature to categorize things, to crave the efficiency of communicating, “Mattermark is XYZ for ABC.” It feels safe and socially correct, intellectually digestible, rational. But those motivations aren’t what drive category defining innovation, and making decisions for the sake of ease is always a big red flag to me.

Don’t get me wrong, there will be a time for having an extremely clear vision that we can lock in on, market against, hire against, build against. In fact, I think we might be very close to that moment of clarity. But before we go there, I’d like to share the journey of the past year after announcing the shutdown of our previous startup about a year ago on March 10, 2013.

Prologue

In March of 2013 my previous startup Referly was down to just 3 of us, after reaching a peak company size of 10 people. With $350K left of the $1 Million raised during Y Combinator, we had enough money to get another project off the ground. Kevin, Andy and I cut our salaries to nothing and headed to my parent’s house in Washington State for a couple weeks. We committed to nursing our pride, reading books, drinking wine, cooking, reconnecting as friends and thinking about to do next outside the fishbowl of Silicon Valley.

I was still a little pissed off at Paul Graham, who months earlier had said:

“You’re just too plausible. It’s as if someone was writing a sitcom about startups and you needed a believable idea.”

Every time I reflected on this conversation I would get pissed off all over again, but in a state of defeat it seemed reasonable to ask myself the previously forbidden question, “what if PG was right?”

If PG is right, I reasoned, then I need to stop doing such plausible things.

Moving to Seattle — We could have stopped here.

Before Mattermark, There Was Cursive

“Mattermark Wasn’t a Pivot”, but before there was Mattermark we almost did pivot the existing technology of Referly into a different idea, which we code named Cursive. It was clear the posts I’d been doing about startups were the biggest drivers of traffic for Referly, and we had a beautiful custom CMS. Medium and Svbtle were still much earlier in their growth curves, and appeared to validate the model so I thought maybe I would start blogging prolifically on my own platform and then invite others to join me, like a community-written TechCrunch.

I shared this with PG during office hours and he said something like:

“If you’re shutting down Referly why not start over with something completely different? If you want to kill TechCrunch then do that.”

Having some direction was exciting, but I also felt the foreboding sense that running a media company had the potential to be a soul sucking grind, with little software innovation and even less margin as a business. I’d have to really love the day-to-day to make it my startup. To test the waters I went on a mission to publish at least one post every day, for 30 days. 3 that stick out in my memory:

1. On April 3rd I ranked the growth of Andreessen Horowitz’s portfolio companies and Marc Andreessen emailed me along with several General Partners, saying: Thanks Danielle! We should do this ourselves :-).
2. On April 5th I published “Zombie VCs” using Crunchbase data to generate a list of firms who appeared to be inactive. This post electricified the VC industry and business media for a full week, intriguing and infuriating investors while revealing the lack of publicly available data. Awash in corrections and new relationships after hundreds of calls and emails, we’d found something important.
3. A scoop on Tumblr’s revenue shortly before the Yahoo acquisition earned a link from Kara Swisher, outreach from someone at the WSJ about a potential job, and a meeting with Michael Arrington. I’ll never forget, Michael sat down looking very grumpy and read the entire 1,200 word piece in silence before looking up to proclaim, “you can write”. I just laughed out loud, my blogger hero just said I could write! Holy shit.

At first I imagined building a “TechCrunch-killer” media business. I imagined a future where I would write about under-appreciated startups, analyze deals, and report on the dynamics of Silicon Valley and other startup ecosystems worldwide. I estimated TechCrunch made ~$20M in revenue a year, and that the bulk of this came from events. Much like the early days of Twilio, I rationalized the small market by telling myself the writing would initially appeal to hobbyists but that we could build a broader appeal from there, and learn as we went. We’d find a bigger market somewhere.

Building a tech startup oriented media company was a much narrower vision than Referly (my previous startup), where I felt we never settled on a single crisp sentence to describe ourselves. After more than a month adrift the clarity of focus in our new direction was appealing and 90 posts later my personal blog was squarely in Alexa’s top 10,000 websites, comments and emails of appreciation had started coming in, and we were even getting tips and sources for more original reporting.

We could have stopped there.

Fun lifestyle business, but what about building a software scale company?

The Research Lab

Byproducts of work are a gold mine. In the process of writing articles I created hundreds of spreadsheets to research markets, compare companies, and come up with unique angles. I published raw spreadsheets in many of my posts, and received a lot of requests to download them. As programmers, we were inspired by things we had heard about BuzzFeed’s assignment desk technology. We don’t know exactly how it worked, but the idea that they had built tools and processes to detect the most buzzworthy topics at any given time across the web was fascinating. We imagined we might do the same thing, but with a focus narrowed in on tech companies.

To do this, we knew we need to detect buzz about companies and spreadsheets weren’t going be enough. We compiled all our spreadsheets into a database, wired it up to a super-simple javascript table plugin and showed it to our investors. Not everyone got why this was so powerful, but for the handful of people who did the excitement was palpable, “can we use this right now?” they asked. We didn’t even have a login system, it was just an HTACCESS username and password on the project we called Cursive.

They told us about the money they spent commissioning custom research from GLG consultations, 451 Research Group and other expensive resources. They showed us their own spreadsheets and internal apps. They asked us if we’d consider coming on board full time to build an internal research team and make investments based on our data and methodology.

We could have stopped there.

If this data is so valuable, and you really have a unique angle, why aren’t you investing with it?

The Venture Capital Firm

Once it was clear we would add significant value and differentiation to someone else’s VC firm, the logical next step was to consider starting one ourselves. Could we raise 10, 20 or even 50 Million to test out various data-driven approaches to investing in startups? We spoke with several established investment professionals who could have come on board to help us run the thing, and it seemed promising. But we knew if we decided to run a fund we would not be able to be objective, share all our data and views as openly, or remain a trusted source charging people for our research.

I knew if I became a VC I wouldn’t be satisfied with small checks and a small role in the lives of my investments. I’d want to lead deals and dedicate the next 10+ years of my life to becoming part of the top 10%, both in results and in value add. Wait! My subconscious screamed.

We could have stopped there.

Are we going to be satisfied building software for a dozen people?

Bloomberg for Startup Investors

Through April and May of 2013, and could feel some sort of groundswell thinking around “Quantitative VC” which saw various firms positioning themselves based on their sourcing and research tools. Google Ventures, Greylock, Correlation Ventures, Andreessen Horowitz and Sequoia were widely known to have some kind of data-driven aspect to their operations, although no one shared any details.

Leena Rao’s trend-piece in TechCrunch on June 1st was the moment things began to come clear. It was like someone started the play clock, and I wasn’t even through reading the piece before I started drafting an email to pitch Mattermark to her. I remember it was a Friday, and so we went out for our launch on a Monday — moving into our new shared apartment, sick as a dog, after coding furiously for 48 hours to productize the damn thing.

We made people sign up on a wait list, because we hadn’t really optimized anything. It was ugly. It was slow. But it was the start of relieving pain for a set of customers who had been neglected for far too long. In the next three weeks I’d spent more time in Menlo Park than I ever had in my life, cruising up and down Sand Hill Road meeting with partnership after partnership. Except this time there was no pitch, there was no angle other than: let me show you this thing we made, let me teach you how to use it. I scribbled more than 100 pages of furiously scrawled notes.

I didn’t truly know what customer development meant before these four intense weeks. I thought I had done it with Referly, but I realized had been lazy and unfocused by comparison.

A few short weeks after launch, Union Square Ventures became our first paying customer and Albert Wenger penned a post endorsing our product to the community of investors and we were off to the racing signing up VCs.

We could have stopped there.

The Deal Intelligence Company

Our customers are venture capitalists, so they think like VCs and often ask:

But isn’t venture capital a small market?

Yes, and we love it! Small is a relative term. The NVCA says there were 462 “active” VC firms in the US in 2010, and when you include all firms who have raised money since 2002 that number climbs to 791. Using the AngelList API, we found 6,961 users with the word “Partner” in their bio.Based on Mattermark’s pricing of $499/month per user, or $50,000/year per firm, the annual revenue opportunity for this initial market is somewhere in the $20M — $40M range.

As a founder who is getting these initial customers to use her product and say “wow” on a regular basis, this is glorious! It’s not the swing-for-the-fences multi-billlion dollar sized market that venture capitalists are looking for, but when you’re not fundraising that doesn’t really matter.

Let me repeat that: when you’re not fundraising, it doesn’t really matter whether your immediate customer base is swing-for-the-fences large.

What a liberating realization!

This knowledge, combined with our wonderful seed investors and very healthy annual revenue growth (averaging 40% MoM since we started charging in July 2013) bought us time to explore adjacent market spaces to our initial set of customers. The wonderful thing about speaking frequently to happy customers is that they tell us all sorts of useful things about the world they live in, and as those relationships deepen we have opportunities to continue solving more and more of their problems.

Once we began to demonstrate that we could build quickly and be responsive to feedback and feature requests, our customers started introducing us to their friends from adjacent social circles and professional ecosystems. It turned out these folks had similar problems Mattermark could solve, and this is how we discovered that Mattermark was useful for much more than venture capital deal sourcing.

VCs hang out with angel investors, investment banks, limited partners, growth and private equity professionals, wealth managers, venture bankers and of course meet with hundreds of startup founders each week. As we began talking to all these constituents we learned more about their problems and it became clear Mattermark could be useful to each of these people in different ways, often with only very small improvements that also benefited all our existing customers.

It’s funny looking back, because at first most of these requests were actually kind of annoying, and felt like one-off opportunities that were getting in the way of our existing product roadmap. It was tempting to brush them away as distractions and stay the course with our core product and core customer base, but on our mission to be “revenue first” this time (and also running on very little cash before our bridge round) we were easily persuaded by money. For example, even before we started charging for our SaaS product a customer offered to sign a $50,000 contract to build new functionality that was way down our list, but would ultimately be hugely beneficial to our customers as well as strategic to opening up future markets, so we built it.

With our first taste of revenue it was only a matter of time (about 3 weeks) before we would begin charging for monthly subscriptions to the suite of Mattermark data and tools.