Platforms versus point solutions

This winter break has been a big one for Shovels. Despite having my kids at home, trekking out to Hawaii, and getting swept up in the Southwest chaos getting to and from Disneyland, I put a lot of deep thought into Shovels. This exercise brought out some gut-wrenching paranoia, deep frustration and confusion, and then a bit of euphoria coming out the other side.

It was a critical exercise, and I don’t do this nearly enough, perhaps because it hurts and takes a lot of energy.

Here are the insights from my platform exercise.

We are not alone

First of all, I learned once again that good ideas have company. I was quite certain two months ago that nobody else was scraping building permits. That notion came crashing down hard over the last couple of weeks. We are not alone.

  • BuildZoom: nationwide permit database but no API
  • HazardHub: nationwide permit database with some key dates we need and an API
  • ConstructionMonitor: nationwide permit database
  • Builty: nationwide permit database with an API and a data feed AND INSPECTIONS

BuildZoom was the first permit database I came across. I was confused at first because they appear to be a marketplace for contractors. Their data product is hard to find on their website. It’s as if they intentionally buried this link on their homepage. The more I thought about that, the more it bothered me. In newspaper parlance this is called burying the lede. Why would they hide the very product that I was planning to build a business around? WHY? What was I missing??

That thought nagged at me as I dove further into BuildZoom’s public data. I looked up addresses in cities that had no electronic permit system. They had permit data. Lots of it. I looked up addresses in Contra Costa County, where I had our own data to compare. They had more records than us. I looked up an address in Livermore, where there is an electronic permit system but the permit details are very sparse. They had more data than the city’s own electronic permit system. This was a miracle. It made no sense. And yet they were actively trying NOT to advertise their data product. Whyyyyyy?

Luka pointed me towards HazardHub. I started an email thread with the founders. I was very open about Shovels and what I’m doing and they shared that they built custom scrapers, thousands of them, to collect building permits from every city and county that issues them. They gave me access to their API. It’s fast and thorough. I was impressed once again. We are not alone.

The ultimate find was Builty. I didn’t find them in all of my Googling. They came to me in a conversation with the founder of PermitFlow, a permit processing startup. He listed off the building permit databases he looked at, encouraging us to try to build something better. I didn’t recognize the name when he said Builty. I asked him to spell it, and then I took a note to look into them later. I’m so glad I did.

Builty, as it turns out, is Shovels. They’re doing exactly what we said we’re going to do. It’s uncanny. There’s not a whole lot on their website, but it was enough to give me the shivers. I reached out through their contact form and asked if they have inspections. The CEO wrote back to me right away and she said they did have inspections. They grabbed them whenever they’re available.

She put a phone number in her email signature so I called it. We talked for 30 minutes. At the end of our conversation I was equal parts elated and defeated. Elated because it seemed we don’t need to do the permit scrape ourselves. Defeated because I was still holding onto the notion that we had an original idea.

Fortunately, in retrospect, I think both are true.

Platforms versus point solutions

Before I get into why Builty could be such a meaningful relationship, I want to set some context around platforms versus point solutions.

A point solution is a product that does one thing very well. I like point solutions. I’ve built and sold several of them, and they make some good money for a solo entrepreneur. Toofr, my first point solution, made about $250K per year at its peak. MightySignal was a point solution that topped out at around $1M in revenue per year. AppMonsta was also a point solution, and it made around half that much. All of these businesses stopped growing and were acquired. There’s nothing wrong with building and selling a small business. But that’s not what I’m trying to do now.

Platforms, however, are like a bunch of point solutions tied together. I’ve seen that platforms are worth more than the sum of their parts. There are synergies (yes, that word), and network effects that make the platform exponentially more valuable as new features get added. In the mobile data universe of MightySignal and AppMonsta, the platforms were App Annie (now, Apptopia, and 42matters. Each of them was orders of magnitude greater in revenue, number of customers, and valuation. They all did what MightySignal and AppMonsta did, but they also did a bunch of other stuff. They had other data and mixed them all together to make a killer app that was far more useful than each of the individual point solutions combined.

I’ve seen this elsewhere in my career. I remember talking to the founders of Sendbloom, an email marketing software, in their office many years ago. They were growing slowly and watching a couple of competitors, Outreach and Salesloft, take on ambitious product roadmaps. Sendbloom decided to focus on a single feature, drip campaigns. They were confident that Outreach and Salesloft would implode under their ambition, that they couldn’t possibly execute drip campaigns, and contact management, and prospecting, and more. In short, Sendbloom chose to be a point solution rather than a platform.

Here’s what happened: Outreach and Salesloft raised TONS of money, over $700M between them. Ambition attracts capital, and capital attracts talent. They hired great engineers and product managers, and they both succeeded. They became colossal tech companies with multi-billion dollar valuations. Sendbloom was sold to LinkedIn for an undisclosed amount. I figure it was enough to make the founders some money but only a fraction of the valuation that Outreach and Salesloft are now.

Point solutions don’t grow like platforms. They don’t accrue value and become big companies. I don’t want Shovels to be a point solution!

Shovels as a platform

Builty is a point solution, and from my conversation with the CEO, she’s happy to remain that way. She doesn’t want investors. She wants to build the best building permit scrape and enable us and others to create products (platforms) on top of it. When I asked her why she said what I’d quietly feared as Shovels moved down this path: the scrape is hard enough. They don’t have the bandwidth to figure out what to do with the data. Making the data accessible is all they can manage.

If Shovels wants to be a platform, it must avoid being a point solution. It should not be Builty. When my mind took me here yesterday, I had to sit back and laugh. All that mental struggle over the last few weeks, all that googling and searching and frustration that other companies were already scraping building permits, all of it was a blessing. We don’t need to do that. We can be, Apptopia, Salesloft, and Outreach rather than MightySignal, AppMonsta, and Sendbloom.

This means we can start building the platform… NOW.

Here’s the data we need to mash up together to become the most interesting real estate and construction data platform:

  • MLS/Zillow for property detail and sales history
  • Builty for permit and inspection detail
  • Census for demographic information

With this data we can build products like:

  • Public contractor profiles with ratings
  • Time to permit and permit fees by permit type and city or county
  • Property valuation and square footage adjustment using detail from permits
  • A new green home rating using detail from permits
  • CRM for contractors pre-populated with their permitted jobs and leads based on permit history

And with other data that I’m not aware of yet, we’ll be able to build even more. This is the beauty of being a platform.

The clouds part when we don’t have to do the scrape ourselves. Now we can be creative!

The platform exercise

This got me thinking: could I have landed here sooner? Is there a lesson here about thinking beyond the initial problem?

The platform exercise could be like this:

  • State the current problem and solution you want to build
  • Now assume that the solution already exists and the original problem is solved
  • What’s the next big problem? <– This is probably where the money is!

Point solutions solve problems for platforms. Platforms solve problems for the market.

I want my next company to be big, so Shovels should be a platform.

Leave a Reply