How to start and scale a full stack startup

Jason Freedman's Avatar By Jason Freedman in Startups on

cepal

When we started 42Floors, I didn’t know it was going to be a full stack startup.  Naively, I thought it was going to be a commercial real estate marketplace centered around a great product and great engineering – one of these places where supply and demand pull each other together.

And the way we built our company reflected this.  Our first ten hires, including the founding team, were all great product engineers.   Before we had a single listing secured in our database, we had already iterated many times over the interface.  It’s a preference that many of us in the startup world have.  Our talent is for making beautiful, simple products.  And so that’s how we choose to attack each problem.

Six months ago, we knew we knew we had a problem at our company.  We were attacking every operations challenge with a product solution and it was getting us into trouble.  It was Chris Dixon  (who coined the term full stack startups) who really opened my eyes.  After he invested in 42Floors, we walked through our plans for automating the acquisition of listings.  And he told me plainly that I needed to scale a manual process. There may be a day when brokerages and landlords have all of their listings automatically exported into 42Floors, but that’s not happening anytime soon.

Today, we have a tremendous operation that collects commercial real estate data twenty different ways so that our users can view available office spaces on a simple app.  We manually process email blasts. We manually check for website updates.  We make thousands of phone calls.  We have photographers in every city.  We upload each individual photo and use our own team to categorize each amenity.  And that’s just to name a few.  And once someone starts looking for office space on our site that kicks off a whole additional stream of individual manual activities we have to perform to help them find their space.

 

***

 

This year is going to be the year of the full stack startup.  The seeds were definitely sown in 2013, but we really have only just seen the beginning.  Uber is definitely one of the leaders with a very simple mobile product on the front end and a very sophisticated operations engine on the backend.  We have startups like Instacart with a wonderfully simple mobile app attached to a tremendously complicated supply chain machine.  You can throw in the whole wave of delivery apps in here:  Postmates, GrubHub, Door Dash, and more.  This list could go on and on and there’s going to be so many more in 2014.

The prevalence of smartphones is definitely a catalyst for this new wave of startups.  Cheap ubiquitous smartphones means that everyone interacting in your operations has access to custom built apps.

But perhaps the real motivating factor is simply that we entrepreneurs became less afraid.  I’m willing to admit it. I spent the first year at 42Floors afraid that it would become some boring operations company.  Our competitors rely on updated manual process, so I figured we should be focused on building true technology.  We were just blind to possibilities.

But now I get that a technology company can become really good at scaling simple operations.  And the benefits are so massive because we you control more of the user experience, you get to solve problems much more holistically.   Who would you rather be Uber or Taximagic?

Along the way we’ve learned a few things about what it means to be an full stack startup.  Hopefully, some of this will be helpful to you.

 

How to start and scale a full stack startup

Start with manual processes

This actually goes for all companies.  Early on, whenever you can replace code with a manual process, you should; if for no other reason than it can help you to iterate faster.  We do it religiously at 42Floors.  Everything starts manually.  Save your precious engineering cycles for the times when you actually need it.

 

Fix what’s broken and not until then

The inevitable complaint you will receive from everyone on your team is some manual operation could be done better, and they’ll be intent on building out the scalable version.  You’ve got to resist this as much as you possibly can.  Focus on the bottlenecks, the things that are most broken in your company, and fix those first.

 

Become a power Google spreadsheets user

Google spreadsheets can get you so incredibly far that it’s worth thinking of it not simply as a quick temporary measure but actually as your short and medium term strategy.  Because every time you use Google spreadsheets, you’re starting with a piece of software that everyone understands how to use and is pretty scalable.  Sure you’re always going to be missing the features of a custom built solution for whatever you’re doing. But the hassle of, say, moving your sales team to Salesforce is simply not worth it early on.  It’s also worth learning how to use the Google spreadsheets API so you can both read and write from your database to your spreadsheets.

 

Hire specifically for biz ops

Biz ops people always have a little bit of sales to them in that they like tracking throughput and conversion rates.  They’re accustomed to monitoring pipelines and judging themselves purely on results.  The big key is you want biz ops people that come from the startup world.  The last thing in the world you want is a big company perspective too early. When in doubt choose the scrappy operator who can duct tape anything together over the sophisticated guy with credentials.

 

Measure your operations

We’re all really good at measuring user acquisition funnels, performance, SEM campaigns, and we have a bundle of out-of-the-box SaaS tools that help us with measuring all of these standard startup metrics.  But your company’s business operations are going to be much more challenging to use an out-of-the-box tool for.

But it’s even more important that you get those metrics up on your dashboards.  We use Chart.io in our office which allows us to draw custom graphs based on anything we have in our database. Even still, it’s challenging. But if your company is going to become good at operations you have to be analytics oriented.

 

Hire people who will grow with you

You’re going to have this tension in hiring your various operators. You need people who are able to do stuff, and often a lot of those things are repetitive.  If your company were to never grow or change , you would simply want someone who could adequately do the job, take a low salary, and be comfortable doing it for many years.  But your company is going to grow and scaling the operations is going to be one of your ongoing challenges that never go away.  When in doubt, hire the person who will help you figure out to change things, even if they’ll be bored with their job earlier than you would prefer.

 

 

I’d love to hear your thoughts.  This is going to be big year for full stack startups. Don’t be afraid to put a little elbow grease into your vision.

 

 

Discuss on Hacker News.