A Startup at Microsoft

Big companies and startups have different strengths. Most big companies have one or more strong brands their customers trust, an established customer base and products on the market that allow them to make new bets while still bringing profits. Startups, on the other hand, are free of the past and can operate in a lean way, quickly reacting to changes in consumer preferences and able to continuously adapt while looking for customer value.

At Microsoft we certainly sit on the “big company” side of things. However, when we look at successful startups we see many learning opportunities. For example, successful startups tend to be particularly effective at searching for customer value, at building a product their customers will really want and, critically, at adjusting (pivoting) their direction as needed when evidence shows that what they built is not what their customers want.

We were exploring this topic a couple of years ago here in the Cloud + Enterprise part of Microsoft. We knew there was no better way of learning than actually doing. We were lucky enough that at the time Microsoft was partnering with TechStars to run startup accelerators for new companies wanting to explore business opportunities around Microsoft technologies, Kinect first and then Azure in a subsequent round. During the organization of the Azure accelerator round, an opportunity emerged for a Microsoft team to participate as if it were another startup, together with the 10 other non-Microsoft startups that TechStars selected out of many hundreds.

Long story short, Liam (@liamca), Eugene (@chaosrealm4) and myself (@pmc) convinced our leadership that we had the right project for this (“a search-as-a-service thing”) and jumped in. We stopped going to work at Microsoft –although we were still Microsoft employees- and instead started joining folks from the other 10 startups every day to learn from the TechStars mentors and work on our idea, all companies together in a big open space in a South Lake Union building in Seattle.

Fast forward to March of this year, we were announcing the general availability of Azure Search, a new search-as-a-service offering that is part of the Azure platform. Today, although well-funded as part of C+E at Microsoft, the team is still a startup within the big business, operating in a lean manner, making scarcity a virtue and not something to endure, and ready to pivot in different directions as we obsessively focus on learning from our current and potential customers. Luckily for us, Microsoft has also changed during this time and as it turns out the team values and the new company tone align nicely.

What did we learn?

If I have to pick a single takeaway, it would be this: a startup is a machine that searches for customer value above everything else. In the context of a big company, this means a startup group needs full business autonomy. If you create a startup to build a product that has already been defined, or in some other way try to inject a bunch of agenda to it, you’re doing it wrong. If you give it full autonomy, the result can be a great product idea that customers will love, just probably not the one you thought you were going to get.

Another thing we learned was not to boil the ocean. To be effective you need focus. Our initial reaction was to build a product that would work for any scenario that could be possibly related to the problem space we picked. Customer evidence pointed in a different direction, where a generic platform didn’t address anybody’s needs. By the end of the process we had a strong focus on three areas (it’s still the case: catalogs/ecommerce, user-generated content/social networks, line-of-business apps) and had let go of all the other cases.

Most importantly in the context of this discussion, we learned that it can work.

What’s the same?

We learned that there are indeed a number of things in common between stand-alone startups and startups in big companies.

Evidence. There’s only one way to search for what customers want — talk to them. As Steve Blank would say, you need to get out of the building. You need hard evidence. Ask non-leading questions. Listen carefully. Don’t let your existing ideas filter the message. No action without evidence.

Be ready to fail fast. Failing sometimes has a bad connotation in the corporate world. Startups don’t even use the word, they just “pivot” into a different angle, sometimes a small tweak and sometimes a whole new company. Startups in big companies can only work –and can only be fair to those involved- if all stakeholders are ready to see it fail serially before it succeeds.

People. Building something from nothing is a challenge in many ways. There is no frame of reference, only vision and a steep path ahead. People joining a stand-alone startup and people joining Microsoft are likely different in many ways, but passion is a common trait in those that want to create something new in either context. Startups of either kind need people that want to build something that will be used broadly, not just building something because it’s hard, different or simply because it comes with a paycheck. Every team member needs to believe in the product, the approach and in everyone else on the team.

Scarcity is a virtue. In the context of business, scarcity gives you no choice but to focus on what really matters. Not having time or people to spare is a feature. In most startups this is not a choice. This is an easy one for big, resourceful companies to get wrong.

Customer obsession. Ultimately, startups are about customer obsession. You need to create an environment where everyone legitimately cares about the product and everyone is proud every time a customer success story comes back.

What’s different?

There are certainly some important differences between stand-alone startups and startups within big companies.

Brand. While I love the product we’ve built, lots of customers listened to us during the customer development phase only because they were already customers of Azure or some other Microsoft product. It’s up to us to show value for them to stick around, but sometimes it can be easier to get that first contact going when you’re part of a big company your customers already know or are doing business with.

Risk equation. For individuals involved in a startup, the risk equation between a stand-alone one and one that’s within a big company is fundamentally different. When you’re part of the big company you take significantly less risk if you set expectations right (see “be ready to fail” above), and you expect less reward as well (if you come up with the next Facebook, you won’t be the owner). Which point in the spectrum of risk is right is a personal decision; There’s no right answer for that.

Resources. While I’m a big believer in scarcity, we still have access to many resources we tend to take for granted and that help remove roadblocks and focus on valuable work. We don’t need to worry about office space, learning how accounting works, finding a lawyer when we need advice, etc.

Should startup-inside-the-big-company be the only way?

If the startup approach to customer development is so effective, why do anything else? There are cases where alternative approaches could work better.

Whenever there’s potential for a new idea, something disruptive or at least untested, something where customer value is not obvious, a startup-style approach focused on lean customer development is often the thing to do. This is now done at Microsoft more and more often and the approach has great support from our leadership.

On the flip side, I’m not sure that it’s the right approach for really big “top down” business cases where you feel you have a solid grasp on the situation. I imagine for example the conversation between the folks that wanted to start the Xbox business and the CEO/board: “Pitch: We observe that the console space is great and Microsoft needs to be part of it; Here’s a bunch of hard market data. We need infinite dollars to enter this market. Executives: ok, infinite dollars granted” *. At that point, scarcity and small iterations might not be the best thing. You may want to bet big and go for the big splash. You’ll still want strong customer evidence for your decisions but the rest will have a very different tone.

*I have no idea of how this conversation actually went, but I’m really curious now.

Back to Azure Search

If you sit down to chat with any of the engineers in the Azure Search team you’ll hear that in addition to curiosity and passion for technology, a common theme among them is that they joined this team because they loved the proposal. The idea that we’d run with business autonomy, make choices solely based on customer evidence and work hard to make sure customers love what we build and in turn use it to build great things themselves.

We’ve come a long way, but the startup tone is still with us. Customer obsession permeates everything we do all the way down to process and technical choices such as how we do DevOps so everyone understands our customers’ workloads and why we do continuous delivery so we can get enhancements to customers every few days.

If you want to learn more about Azure Search, go here. You can find me on Twitter as @pmc, and if you tweet anything with #AzureSearch you’ll get a quick follow up from one or more folks on the team eager to hear about your experience with our product.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.