SparkFun is building an 80,000 sq ft building with construction starting in May of 2013. It’s going to be awesome.
tl;dr: SparkFun is building an 80,000 sq ft building with construction starting in May of 2013. It’s going to be awesome.
In January of 2012 SparkFun started to get worried. We had a few open offices and a fair bit of open space on the inventory, production and shipping floors but we were growing. Knowing how long it takes to get a new office space lined up, we started looking for a new home back in March of 2012. This is the story of how we started by looking for a new place to rent and ended up starting a new company, enlisting some of the greatest engineers in the area, getting through a fair bit of politics, and starting a construction project beyond my wildest dreams.
Boulder proper is in the lower left. Red flag is our current location. The red pin is the new land.
We currently live in 52,000 sq ft (4800 sq meters) just outside of the city center. Unfortunately–here in Boulder, CO–there are not many buildings that are in the 70 to 100,000 sq ft size and have the odd mix of manufacturing, warehouse, and offices that we need. With a very low commercial vacancy rate of 7%, we couldn’t find what we needed. The 7% vacancy number comes from this awesome 168-page appraisal (warning: 10MB download).
On May 8th of 2012 a local father/son team asked to meet with me. They have built dozens of commercial buildings in the area over dozens of years. They had a piece of land about 1.5 miles north of our current location that would work very well. It was a bit too large but could be split to a smaller size.
7 acres just north of Boulder city limits
Price of the land: The land was originally listed at $14 per sq ft. That made the full 7 acre plot worth about $4,268,880. YIKES. We immediately decided this was way more than we could afford. They, however, were not dissuaded. What if they split the land and offered to donate it instead of sell it?
In May, the deal looked something like this: the father/son (Andrew and Don Unkefer) donate the land (4.3 acres at $8 per sq ft) and SparkFun would contribute $1.5M in cash and a new company would be formed to construct and own the building of 80,000 sq feet. The new company would then rent the building back to SparkFun at $5.75 per sq ft per year + NNN. The landlord would in turn loan SparkFun $1.6M for tenant improvements (to build office walls, a break room, etc).
Enter Flywheel: I own the lion’s share of SparkFun and at this point it became obvious I should start a new company to deal with building stuff and leave SparkFun to concentrate on manufacturing and teaching electronics. Flywheel Industries LLC was created in May to be the property holding company owned by myself and members of my family. This was done for a few reasons but the big one is separation of assets: if SparkFun ever gets sued by Samsung or Apple (or SPARC International) they shouldn’t be able to go after Flywheel or the building. SparkFun would distribute profits to shareholders (me and a few family members and employees of SparkFun), I would in turn invest in Flywheel.
June of 2012: Mkay so Flywheel is going to invest some money in the new building, Andrew is going to put in the land. It turns out Andrew owned only 50% of the land. The other 50% was owned by another person. We spent about two months hammering on what’s called an operating agreement to try to get Flywheel (50%), and the two other parties (25/25%) to form yet another company named One BTC. There would be a 20 year lease (yikes) to be signed between SparkFun and One BTC.
The rub was this: One BTC would construct the building and provide it core and shell to SparkFun to rent. Core and shell means the HVAC is in place, there are lights on and windows in place, but that’s all. SparkFun would be wholly responsible for all the tenant finish or tenant improvements (TI). Tenant improvement is always spend and forget: when you move into a new building you have to paint the walls, maybe move a door, maybe put in a new sink. You spend this money and never see it again. Often a landlord will give you some of this money as a way to sweeten the lease. We are admittedly a very frugal company but in all the commercial spaces SparkFun has moved into, we never spent more than a dollar or two per sq foot to finish a space ($10 to $15,000).
However, the TI for this new building was very different from normal TI. It was a blank canvas; SparkFun was going to have to build everything inside the box! And while that can be very beneficial to some companies I saw it as unfair. There was a lot of construction that would need to be done before SparkFun could really begin operating out of the building. About $2M worth and SparkFun would need to pay for it all and never see anything for their investment. It didn’t make sense to me to be putting in all the cash ($1.5M plus another $2M = 70% of the project), the other partners putting in land (worth $1.5M = 30% of the project) but I only own 50% of the project.
July of 2012: I made my discomfort known and one of the partners let us know they also had varying degrees of heartburn. So we met and he offered something wild: if I would buy his 50% ownership of the 7 acres he would sell it at $7.00. He would be out of the picture and I could do whatever the heck I wanted to do. Remember the original price for this land was $14 per sq ft so $7.00 was really good. If Flywheel could cough up the dough, it would own 75% of the project.
August of 2012: A partnership is like a marriage - it’s not something to do lightly. When we were talking about partnership between 50/25/25, alarm bells were ringing. I didn’t like being 50% owner because I felt it would certainly lead to disagreements. However, if Flywheel had 75% of the project, we had control and it felt good. The smaller position (the Unkefers would be at 25%) was understandably disconcerting, especially with a complete amateur (me) running the show. So they came up with a very interesting proposition. The plot was 7 acres. SparkFun’s 80,000 building only needed 4.3 acres. If I was willing to do a land swap then I could get 100% ownership of the land we needed. Here’s what it looked like:
Now the land swap:
So when the dust settled, Flywheel would own 4.3 acres outright for about $1.4M. The Unkefers would be out of the deal (but would stay on to develop the project), the other partner would be out of the deal, and Flywheel would be 100% owner of land that was $14 psf but would cost $7.75 psf. Wow. Ok. Let’s do this!
Hey buddy, can I borrow some money? The deal had been verbally agreed upon, now it was time to find out if we could get a bank to loan us the money. I began talking with our current bank (JP Morgan Chase) as well as reaching out to other banks to see if they want to help out with this project. Construction projects are notoriously risky but once the monolith is built, you have a hard asset and banks love to loan money on hard assets. Throughout the months of Sept/Oct/Nov Trevor (SparkFun’s COO), Andrew, and myself interviewed 5 banks’ offers. Grueling, but very interesting all the same. It turns out there are not a lot of banks that willing to, or are capable of, taking on an ~$8.5M project. Interest rates are currently ridiculously low. We sorted through them all and our current bank (Chase) presented the best offer:
I’ve never built a building before so you hire what’s called a developer. Someone who knows the local building codes, laws, politics, contractors, and generally what needs to happen when. I had a good relationship with Andrew Unkefer so Flywheel retained him as the developer through a developer services agreement.
Realize every time there is paperwork passed back and forth it is days of review, meetings, counter offers, revisions, and legal reviews. We’ve gone through multiple offers, operating agreements, leases, and contracts just to get to this point. And we haven’t even figured out what’s going into this 80,000 sq ft box. Time to to hire an architect! We select RVP architecture to be the main architect on the project. Bob Van Pelt is extraordinarily easy to work with. Yet another contract is reviewed and signed.
Yay bubble diagrams!
Professionals taking our gobbledygook and making it legit
In September the initial discussions for the building design begin. The first step in the design process is bubble diagrams. All the department heads of SparkFun are asked how much room then need now and how much they may grow in the next 5 years. These square footages and a few brainstorming meetings about building setup give us the basic beginnings of layout. We ask for the moon knowing that we may have to give it up if it costs too much. We want: windows that open, a deck on the roof for lunch, open work space, closed/quiet work space, a basement for storage, an awesome classroom for teaching and meetings, … The list grows quickly.
Floor plans taking shape
From September through December there are many revisions to the layout as ideas and layouts are hammered out.
In December we start regular Monday morning meetings with the architect, developer, Trevor and myself. A general contractor (Wyatt Construction) joins the team, is vetted and a contract is worked out. From 8:30 a.m. to around 11 a.m. every week we try to vocalize the needs of SparkFun and balance that against what is physically possible and financially sensible. We want to build a very energy efficient building that helps us do our job better but we’re scrappy - we’re not going to be building a glass doughnut. What blows me away is the ability for us to poorly describe a room or a process requirement and a few days later Bob and his team have designed something really amazing.
The final numbers from the appraisal
As the building begins to form, Chase is actively working on the loan. Part of the process is to get an appraisal for the project to be sure the bank does not loan more money than what the land and building will be actually worth once it’s complete. CBRE is chosen as the independent 3rd party and after a few meetings and a few weeks, they come back with a 168 page report that does a very accurate job of capturing the Boulder market. It’s a really interesting read and has some great photos of the plot and the plan.
This process continues until early January when Boulder county passes new code requirements requiring the use of the International Green Construction Code (IgCC). This 110 page book, costing slightly more than a dollar per page, describes all sorts of really interesting requirements for construction. If you’ve heard of LEED certified buildings then IgCC can be viewed (very simplistically) as the lower levels (silver and gold) of LEED certification. There is however a small difference; where LEED was optional and mostly the realm of very expensive government buildings, IgCC is now required for every commercial building over 20,000 sq ft in Boulder county. I found it exciting and inline with what we were attempting to accomplish but all the other parties in the room went fairly pale at this new twist.
The SparkFun project would be the first project in Boulder county to use the more stringent IgCC. We are proud to be the first but it meant most of the engineers, county officials and banking representatives were going to have to learn a lot over the coming months. Everything was going to cost more and while we had the ability to make decisions in 2012, the new code for 2013 was going to force us to a level we were not yet comfortable.
Oh no! Pallets are going to hit the sprinkler heads!
On top of all the new code changes, Boulder city and county have had some very strict building height limits in place for decades leading to very few buildings over two stories. From the beginning, the new SparkFun building was going to be a 33' tall, two story building. Jordan, our director of inventory was very concerned with the usable height of the warehouse. She’s seen our use of pallets greatly increase over the past 12 months and predicted that going into the future we would need much, much more storage for incoming inventory. At roughly 6’ tall, in order to stack pallets on shelves this meant we needed 15 to 16’ of unobstructed airspace. 16 feet out of 33 sounds like more than enough but when you take out the foundation, roof, interstitial spaces for HVAC, plumbing, and electrical wiring there’s actually only about 13 feet of usable space.
Have you seen similar signs? This one has our name on it!
Participating in a county commissioner’s hearing was not something I signed up for when I graduated with a EE degree. Because of the size and complexity of our project we were scheduled to go in front of the county commissioners for a ‘special use’ hearing in Be. Early in the January Trevor and I floated an idea with Andrew to increase the building height from 33' to 38'. The extra 5 feet would allow us to stack pallets in the warehouse and would, in the next 10 to 20 years, reduce the need to building yet another building. Part of the developer’s job is to work with the local governmental bodies to get the required permits and paperwork. We originally thought that extra height was out of the question. Thanks in part to Andrew and the county commissioners, in late February we were able to state our case and persuade the county that the height extension was a good trade off between environmental impacts and minimizing the eyesore of a large commercial building. We got our height extension and there was much rejoicing.
The building after height extension and various IgCC changes
With many of the governmental hurdles cleared March was spent working with all the engineering groups (HVAC, plumbing, civil, electrical) to hammer out construction documents. This is the period where all the needs of SparkFun and its 140 employees get boiled down and communicated. How many data drops do we need and where? We’ll need compressed air in production but what about shipping? Will you ever, in the next 20 years, need a 2nd elevator? IgCC requires outlets that automatically shut off–where do we locate the overrides so that people can turn on their soldering iron after hours? We generate an amazing amount of cardboard and recycling–where do we locate the huge dumpsters to make it as easy as possible to recycle? There are countless discussions and debates that ensue for every aspect of the building.
The level of detail increases as we get to the final construction documents
Want to see the full spec? It’s massive.
While the building design was being hammered out the construction budget for the building was put together by Wyatt Construction, the general contractor.
How much will the doors for the building cost? About $68,000. How much sheetrock and metal studs? About $475,000. How much does it cost to build a 38’ building with 80,000 sq ft? About $7,037,354. This number doesn’t include a few other large expenses namely:
The total bill looks to be about $10,500,000 with monthly payments to be about $70,000 a month. $10M is a terrifying number but to put it in perspective, this is part of the calculated plan for the next 20 years of SparkFun. I’m ready to invest now to be sure we have a good work environment with space to grow.
Once a final plan set is created you can apply to the county to get a building permit. Only after the permit is granted can construction begin. We were hoping to break ground in March but because of the special use review hearing, the schedule got pushed to May.
Checkout all the groundbreaking photos!
So after much work, on May 1st, during a late spring blizzard, the SparkFun crew stood out in a frozen field and dug some holes. It was a crazy start to an extraordinary adventure.
It should go without saying that it takes a tremendous amount of expertise and time to pull off a project like this. We have a great team of people and it’s important to thank each and every one of them.
If you’re interested in hearing more about the building, please let us know! There will be plenty of updates to share over the coming year.