Salesforce Mission Bay Rendering
A few days before was poised to receive the City’s approval, Salesfoce pulled the plug on their plans for a 14-acre campus in San Francisco’s Mission Bay.
From Reuters on Friday:

The company’s stated reason for changing course was that the new campus would not be big enough for its growing workforce. But people closely involved with the ill-fated development paint a picture of an out-of-control project that lurched forward even in the face of stratospheric costs and tepid support among employees. Only a construction expert hired late in the planning process convinced Benioff that moving forward would be folly.

The canceled development has already cost Salesforce tens of millions of dollars, and the price tag could rise further if the Web-based software maker fails to find a ready buyer for the land it bought, according to recent regulatory filings.

The cancellation has also singed the company’s relationships with city officials, including San Francisco Mayor Ed Lee, who had touted the development as an economic boon and was stunned when it was suddenly abandoned.

From 2010 to 2011, the projected cost of constructing the campus ballooned from $750 per square foot to roughly $1,000 per square and a project cost of roughly $2 billion.
First Four Campus Buildings Poised For Approval [SocketSite] Campus Key Elements And Design Preview [SocketSite] Kills Mission Bay Campus, Open To Offers For Land [SocketSite]
Insight: Salesforce’s plan for opulent campus a costly debacle [] Acquires 14 Mission Bay Acres To Build 2 Million Feet [SocketSite]

22 thoughts on “Behind Salesforce’s San Francisco Campus Debacle”
  1. “The company’s stated reason for changing course was that the new campus would not be big enough for its growing workforce.”
    Even an idiot could see the absurdity of that comment.

  2. I thought Mission Bay was supposed to be a biotech/pharma magnet? I think SF should keep its eye on that long-term vision: new hospital, new biotech companies. Dotcoms can go anywhere…

  3. Good thing they canceled it. In more times than I care to recount, a company building a monument to itself in terms of a building is a “sell” signal for the company’s stock.

  4. Seeing this colorful rendering, reminds me of Yahoo circa 1999.
    Still seems odd when unprofitable companies make huge forays into real estate projects…

  5. I don’t see how this cancellation is positive unless you’re either a commercial landlord who now stands to gain from renting space to Salesforce or a penny pinching shareholder in CRM. For San Franciscans in general, the failure of this project is a net negative.
    From an extended excerpt from Daniel Gross’ book Pop! Why Bubbles Are Great for the Economy:

    In that magical world known only to a few economists, where resources are allocated efficiently and investors and consumers behave rationally, bubbles would always be unambiguously negative. They cause all sorts of distortions and silly, self-defeating behavior…but that’s not the world in which we live, especially in the United States. We are not blessed with a population composed entirely of hyper-rational individuals…

    Looking back through the last 150 years, a familiar pattern emerges. A wonderful new technology or economic idea arrives. A few good years of solid growth help engender a sense that things are different and that new rules apply…Hype and rosy projections…justify investing at stratospheric levels. The trend, previously confined to the business community, crosses over into popular culture…And then, pop! The bubble bursts, heroes become goats, and bankruptcies spread. As corruption and venality are exposed, self-loathing and recriminations rule the day. (See: subprime lending, spring 2007.) And that’s when all the moralizing narratives about the tragedy of bubbles get written.

    But this is only half the story! After all, the process of growth and innovation doesn’t end when a bubble bursts. The Internet wasn’t unplugged and shut down in 2002. In fact, once you gain a little historical distance from bubbles, it is clear that some bubbles–some, not all–leave behind something that is a little bit boring but extremely useful: infrastructure. The bubbles that have left behind commercial infrastructure have been incredibly important contributors to America’s remarkable long-term economic performance.

    Simply put, bubbles are how new commercial infrastructure gets built in this country…In the 1880s, vast competing, and often redundant, rail networks were built way ahead of demand. By 1894 about a quarter of the rails were in bankruptcy. The 1990s saw an orgy of commercial infrastructure built for the Internet. We all know how that ended.

    Unlike Wai Yip Tung I was not impressed by the Reuters article, because the theme was that salesforce was somehow overspending irrationally; the reporter threw in lots of anecdotes about lavish launch parties and the architect designing a home for a Salesforce board member. But here’s what really stuck in my craw:

    By last autumn, word spread among local real estate insiders that new projections put the overall cost at over $2 billion, or roughly $1,000 per square foot (about $11,000 per square meter) – a 33 percent rise from previous estimates and far more than the market rate of $600 per square foot (about $6,500 per square meter) for new corporate offices.

    That would have made it the most expensive office development in San Francisco history, said Colin Yasukochi, vice president of research at commercial real estate firm Jones Lang LaSalle.

    Benioff maintained that the cost projection did not rise dramatically.

    Here’s the thing. If San Francisco wants to continue to grow, it seems to me that somebody, somewhere at some time is going to have to pay for new development that is “the most expensive office development in San Francisco history”.
    The current market rate for corporate offices, even new(er) ones, includes lots of drab, boring and retread offices that are just trying to be cheap. With this project, Salesforce was trying to do something cutting edge. Of course it would be more expensive. Campuses that are open to the neighborhood are more expensive than simple towers, no? But which one is better for residents in general?
    And if you can’t finance cutting edge projects (and this project was daring and cutting edge and innovative, even if I don’t personally agree with a lot of their aesthetic choices) with new, bubble-obtained equity financing, then when can you finance it? When can projects like this get built, if not with funds from a firm like Salesforce?
    Rational allocators of capital in a mature business aren’t going to attempt a campus like this. Like Gross says, bubbles are how new commercial infrastructure gets built. And my answer to Drew’s comment above is that once the campus was built, there would be very little stopping the buildings from being revamped to house biotech if Salesforce went belly up a few years after project completion. I wasn’t aware that there’s some large biotech company that’s waiting in the wings to build here and because of Salesforce, they can’t get the space they want. If Salesforce turns around and sells the property quickly at a profit, then we’ll know I’m wrong.

  6. So the “penny-pinching shareholders” of CRM should be forced to pay for San Francisco’s development?
    Hmm, with thinking like that no wonder so many people are hosed for their retirements.
    Capital should be invested where it will generate the greatest rate of return for its owners. And communities can decide how they want to encourage/discourage economic opportunity.

  7. To clarify, I like the article because it explains what happened behind the scene. I guess most of us would have suspected there were more going on than the Saleforce’s official rationale.
    I don’t know if the collapse of the project is good or bad for San Francisco and for Salesforce. I don’t know enough to make a judgment. I do appreciate the effort to make it open and public serving. Mission Bay development is boring so far and the Legoretta design does stand out from others. In this respect it is regrettable that it cannot carry forward.

  8. re: the “sell” the stock of the company that is building the ego driven HQ–anybody know how that Apple flying saucer structure is moving forward?

  9. I don’t think anyone’s arguing that it would necessarily have been good for Salesforce to build the thing– just that it would have been good for the city. These companies are going to blow through billions anyway– why not leave a something that lasts?

  10. I always thought it was a shame that none of the billions in the boom left any buildings in SF. SF blocked every single one of them.

  11. Salesforce should have taken a page from Amazon regarding their new corporate headquarters in South Lake Union in Seattle. Get a developer to build the headquarters for you and you lease them. It’s a much better alignment of interests and keeps costs in check. If Salesforce ultimately wants to be the owner they can structure that option into the lease and be the owner once the project is complete. Bottom line, a company like Salesforce has absolutely no business leading the development of a project this size.

  12. djt.
    “Seeing this colorful rendering, reminds me of Yahoo circa 1999.
    Still seems odd when unprofitable companies make huge forays into real estate projects…”
    There are two other examples:
    1) Barney’s in NYC
    2) SGI which I am personally familiar with.

  13. Apple is actually a great example for when a company should built their own project, especially if you buy into Joshua’s rational capital allocation assumption.
    Apple has a significant number of employees in Cupertino in particular and in the Santa Clara Valley in general that are not housed in the buildings centered at Infinite Loop and is paying rent to house them. As everyone knows, the company has literally billions of dollars in cash just sitting around not doing anything, and the company is paying lots of little commercial landlords rent.
    Apple doesn’t need financing to build their project (although they might still borrow for tax efficiency purposes) and when they rid themselves of dealing with all the two bit commercial landlords, they’ll save money over the long haul and therefore be even more profitable once the reduction in overhead costs is taken into account.

  14. But then Apple will be stuck with a building that can not be reasonably built in phases, nor can it be expanded since it is assumed to be perfect upon its creation. And if Saleforce’s employees overwhelmingly resist moving even a short Muni LRT ride from downtown, why wouldt Apple’s employees (and more importantly potential future employees) be different? (OK I know they “think different” – they really really do). But when every indicator says that young creative tech professionals crave urban settings and alternatives to driving Apple is hobbling itself by tieing itself down in a suburban space donut. All of the strategies that Brahma suggests can be accomplished in an urban setting with a scalable phased project. Maybe they could take over the old Salesforce project.

  15. JC, I think you have good points regarding the proposed second Apple campus, but Apple appears to be moving forward with it, so those two factors must not have been deemed all that important by Apple’s executives. And even though the building is circular, they have an option built into the plan to develop an additional 300,000 ft² of office space on/near the site.
    Apple could have moved to downtown San Jose anytime in the last fifteen years and the same goes for San Francisco, so “young creative tech professionals crav[ing] urban settings and alternatives to driving” must not have held all that much sway either.
    I can do a straight discounted cash flow analysis on the rents that a company like Apple is paying commercial landlords all over the Santa Clara Valley, so I assume that the highly-educated and compensated managers at Apple can and did do that analysis, and at a level of sophistication much higher than I could dream of, before they decided to initiate this project.
    I can’t at all do anything similar with vague general assertions like “better alignment of interests and keep[ing] costs in check” as an argument in favor of paying a commercial real estate developer to build the project and then leasing the resulting space. I would love to see the cost accounting on situations like the one outsider describes above, but unlike Joshua, I don’t believe that managers of enterprises always “invest capital where it will generate the greatest rate of return for its owners” in the real world we live in.
    I don’t have any insight into the contracts that Apple may have signed in the past with commercial landlords, but given what the result was, the terms couldn’t have been too favorable to Apple.
    If you’re a commercial tenant, assuming that the actual developer is a rational revenue maximizer, aren’t you by definition covering the cost of the building AND the cost of the developers overhead AND the developer’s profit if you took that option? Maybe it is Amazon that isn’t a wise steward of their investor’s capital, or they just had other priorities than optimizing allocation of real estate-devoted operating costs.
    Back to the cash flow analysis…can’t do anything similar with the living arrangement preferences of “young creative tech professionals”. How’s that preference revealed? Apple isn’t having a lot of trouble with recruiting personnel, are they? And there are lots of not-so-young tech professionals that want detached SFHs for their children to live in, and Cupertino provides that, no?
    btw, downtown Cupertino is not an oppressive auto-oriented suburb like Dublin (the one in the East Bay) or Livermore. As I pointed out in the last thread on the proposed Apple project, the Santa Clara Valley Transportation Authority serves both the current Infinite Loop campus as well as the one proposed at the former HP site, so a potential employee wouldn’t necessarily have to drive. And Apple has a corporate fitness center complete with showers at the existing and proposed sites, so one could always ride one’s bike.

  16. The most unfortunate part of this entire debacle to SalesForce’s neighbors is how poorly the property is being managed. The site is a constant location for tagging and graffiti. The site is poorly secured and has become an attractive nuisance for partyers going who scale the old cement factory structures in an effort to make their mark.

  17. As a SFDC employee, I am very happy this was cancelled. Public transportation to 1 embarcadero is much better than to SOMA. The company doesn’t need to be buying expensive real estate, and the city was making it very hard to develop. For example, they wanted to put in a daycare center as an employee benefit. What next — if an employer provides a cafeteria or gym, these must be open to the public, too? A campus requires security and access controls for who can be present. It is not a publuc space, but a private place of work. There are numerous regulations as well as customer requirements that control who can be CRM property — as well as common sense. E.g. Non city of SF employees cannot access daycare or other facilities provided to city employees. Try using the restroom at the church and market muni station. Its for muni employees only. But private firms are held to a stricter standard. It’s the thousand little regulations like this that cause the costs to the employer to go up. For me, this was a great example if SF greed checking the CRM ego, resulting in a good decision being made.

  18. Sorry, the above was incoherent (shouldn’t post from iPhone) 🙂 I wanted to say that the city required that the proposed daycare center be open to the public, but this conflicted with CRM security policies that all non-employees be accompanied by an escort while on CRM property and that all those not escorted be cleared with a background check. As a hosting platform for third party data, CRM is obligated to follow security procedures that control who can and cannot be on the campus.

Leave a Reply

Your email address will not be published. Required fields are marked *