Geographic expansion: why, when & how

Sharing Thoughts With Gustav From Spotify

Over the past few years I have been exposed to very different philosophies and approaches related with the issue of geographic expansion.

I learned that there is no one rule fits it all but rather a plethora of factors that must be considered each single time.

To help me to debate this issue, I asked my friend Gustav Söderström, Chief Product Officer at Spotify for a hand. He is an entrepreneur himself and investor/advisor in multiple startups.

Roberto: When I look at all the best startups in Europe, they all ask themselves when is the right time to open an office in another country. It’s amazing to see how each company tends to give a different answer to this question. Wooga for example is located in Berlin. The company employs several hundreds people but they are all based in a single location. With users all over the world, the choice of a running the operation from a single location might sound bizarre. On the contrary, I believe that this is one of the company’s key strengths. Gustav, what’s your take on this? Do you think opening up offices in different geographies is a must for any startups?

Gustav: The short answer is no, it is not a must for every startup, on the contrary, I think it adds a lot of complexity and risk to a startup. The normal leadership style of most startup CEOs of “just being everywhere all the time” doesn’t scale to multiple offices. I think it depends on the business idea.

A few examples:

- An iPhone game: The core of the company is the game itself and it’s mechanics. It varies little by geography and distribution is fairly easy. In this case it makes tons of sense to leverage a single location for as long as possible.

- Uber: The core product is fairly straight forward to copy once a company nailed it. The whole business is really about scaling to new markets faster than anyone else (another example would be Groupon). Price, competition and regulation is often very local. In this case a big piece of the success of the business depends on an organizational model able to easily scale to multiple markets.

- Spotify: Because of the local dimension of the music catalogue and the need of local advertising sales organization, geographic expansion has been a must relatively early in the company building process.

Roberto: Ok, so Gustav you think that it makes sense to keep product development in a single geography? I guess on one side this should make things much easier to manage. But, on the other side, what about talent acquisition at scale? Is this an area where we (Europeans) have a bit of a competitive disadvantage because of geo-fragmentation? Could you have found enough people in Stockholm to build all aspects of the product in the same location?

Gustav: Yes, I think it makes sense to keep product, design and engineering in one single location for as long as you can. But to your point, at a certain level and for certain types of talent you may need to go to other markets which eventually became the case for Spotify. As an example, the kind of experience in scaling system that exists in the US (due to the FBs’ and Googles’ of the world) is very hard to find in Sweden/Europe. The same is/has been true for machine learning skills where Spotify eventually acquired a company in Boston (i.e. close to MIT).

Roberto: So, how do you decide the right timing to expand geographically? I have seen some cases in which entrepreneurs waited too long and slowed growth down. On the other hand, I have also seen a lot of cases in which new offices were open too early and the culture of the company was practically destroyed by a total lack of identity. How do you deal with culture as soon as you have multiple offices but you still are an immature startup?

Gustav: I think a good way to look at it is to your point, to look at the skills you need and if they exist in your market. If you really depend on skill you can’t find, you’ll be at a competitive disadvantage vs your competition.

Regarding culture, I think this is THE vital point of scaling geographically: this is where experienced CEOs do make a difference.

Roberto: Therefore, the obvious question: what’s the secret to spreading culture across multiple geographies?

Gustav: Spreading culture is very hard work. At Spotify we’ve experimented a lot in this field. We have worked with formal tools and wrote down what we think our culture is, but we have also experimented with rotating people. Obviously, culture in an office is a mix of three dimensions: 1) the company, 2) the country and 3) the local people.

I also believe is very important that culture is captured bottom up through the organization.

We’ve therefore chosen to make our culture (specifically engineering culture) fairly public, which really helps as people get the right expectations even before they apply for a job.

Roberto: Ok let me get this straight. I think what we are saying here is that there are two main drivers that lead a startup to geographically expand:

1) customer requirements and need of proximity;

2) scale up the organization and access multiple sources of talent.

It seems to me that you guys decided to expand because you had both requirements at the same time. Are you suggesting that this should be a standard approach for European entrepreneurs?

Gustav: Yes, and again, it’s important to distinguish there are different parts in the organization and is important to keep them distinct. While we’ve had small local sales offices in almost all markets very early, we were always very cautious with splitting product/design/developments and that happened only much later.

Roberto: I often give two suggestions to the entrepreneurs with whom I work:

1) let’s avoid spaghetti businesses (nano businesses in many geographies);

2) let’s move key leadership members if we are serious about adding a geography.

Gustav: Agreed, I think that opening several offices should be seen as “a necessary evil” that you only do if you have to and have strong rationales for doing so for your particular business. It will be a huge time sink for the CEO and the whole management team.

However, when you do have to do it, I think physically moving the culture is the best way and we’ve done it by rotating people for up to 6–12 months (which also happens to be a great way to let people in another geography like Sweden try something new without having to find another company). Let’s keep in mind, however, that it is also important to find local strong leaders (as you want their networks in that market). To facilitate this we’ve focused a lot on onboarding and all the leader of a new office will spend 3–6 months in Stockholm first, as part of their contract. We actually fly every single employee to Stockholm for intro days (once per quarter) where they spend a bunch of time, meet me and everyone else in management and learn a ton about the company.

Roberto: Gustav thank you. Just to get a sense of the scope of your experience on the topic of geographic expansion, what has been the size of your actual experience?

Gustav: At Spotify We have product and engineering in 5 places now actually. Stockholm and NY are basically our hubs with many multi purpose teams. Then we have teams with specific skills and missions in other geographies, eg SF is close to many of our partners like FB etc so are focused and partnerships, Boston Echonest is obviously focused and machine learning and music intelligence. In general, we’ve either put people from another office to start the new one, or we’ve hired very strong local talent and onboarded them carefully in Stockholm.

Roberto: It sounds like you have some real life experience on the topic! I love that ;)