If you’ve been reading our blog, then you’ve likely realized by now that we don’t believe the cloud is the only option when it comes to building and scaling a successful SaaS company. That’s because, over time, cloud costs can really add up. What may once have been a fast and cheap tool for early stage startups to build and launch a SaaS offering, can quickly become a huge bottom line expense for companies as they grow larger and more established.
It’s hard to know when exactly the pendulum will swing, turning the public cloud into more of a burden than an asset. Luckily there are certain things you can do throughout the lifecycle of your company to gut check your use of the public cloud and set yourself up for success in the future. This means you don’t have to rule out the public cloud as a resource altogether, but rather do the work to understand how to make it work best for you.
Below, some of the top things to consider at each stage of your SaaS company’s growth to help you best leverage the public cloud.
In the early days, SaaS startups must decide quickly whether they want to invest in building their own cloud infrastructure or utilize the public cloud. If the latter, they must also decide which provider to go with. To make the best choice for your business, get your priorities straight and evaluate the benefits each provider offers.
When building in the cloud, startups are forced to make decisions all the time that weigh the value of portability against the value of speed. In the development process, cloud providers like AWS provide fantastic native apps and services that allow you to spin up new features quickly. But, by using these services instead of building features on your own (which could take much longer!), you’re tethering your product to that specific cloud environment. This may not seem like a big deal early on, but it’s important to consider the future impact of this decision and make sure your executive team is aligned early on about the value of portability against the value of speed.
Blindly opting for speed over portability can cause problems down the line. Say you build your SaaS offering using AWS native apps, but a year later a large portion of your business is coming from federal government contracts and you need to maintain certain FedRAMP certification standards. You may want to move to Azure, because Microsoft has a leg up with FedRAMP compliance offerings — but your ties to AWS could be too complicated to unravel at that point, without significant time and investment. The same issue may come up for companies who grow large enough that they want to move their offering to their own cloud infrastructure instead of the public cloud for cost reasons. If you didn’t plan ahead to prioritize portability, this can be a major problem. An early decision to opt for portability may include more upfront time to build a product sans native apps, but you’ll thank yourself for investing that time later.
Another early lure of public cloud providers comes in the form of credits. Cloud companies will offer financial incentives for startups — often to the tune of up to $100,000 in cloud credits — to win their business and emphasize the ease and speed of their service. The catch? Most times, these credits need to be used within 12 months of entry in the cloud environment. In the early stages of building a business, it can be hard to use that many credits quickly. Our advice here is to negotiate, negotiate, negotiate. Try to broker a deal for an 18-month window, or ask for the cloud credits to kick in 6 months after you’re up and running. By then, you’re much more likely to need and use the credits, instead of missing out by leaving some unused credits on the table.
Once your company reaches a certain level of maturity — with multiple customers and customer data to analyze — you reach a position of power in your public cloud relationship. Leverage that power by advocating for yourself within your cloud partnerships and using data to guide product development.
As your company matures, you may find yourself wanting to change cloud providers for one reason or another. But many companies are hesitant to make a switch because of time and monetary investments involved. In actuality — fearing change is the worst thing you can do. If you’re unhappy with the quality of service or are facing constant price increases, make a move to a new provider or consider a multi-cloud strategy (working with multiple providers at once). Many times an early decision to make a change can cause less headaches down the line. It can also put your public cloud partners on notice. If a partner sees that you’re willing to change providers or add other partners to the mix, they may offer you better incentives to stay, which could really benefit your business.
Operating in the public cloud at this stage of your company’s growth and development can also be hugely beneficial from a product development standpoint. There are tons of analytics tools out there — like Pando Analytics and Heap — that help you analyze data in the cloud and turn that data into actionable insights that can guide the future of your product. If your product runs on a customer’s servers, or is part of a homegrown on-premise environment, it can be harder to build tools and put processes in place to mine data efficiently. Take advantage of your public cloud infrastructure at this time by learning as much as you can about user behavior and product performance and letting that guide development priorities.
When you’re in growth mode, the cloud is the best way to build quickly and efficiently. But at a certain point, you become large enough that you need to evaluate your growth through the lens of earnings and expenses. At that point, cloud costs can start to become a problem. That’s when you need to be extra savvy about your use of the cloud.
Our first tip during this stage of growth is to get smart about reserved instances. Purchasing reserved instances ahead of time can help you avoid massive fees for burst cloud capacity. Naturally, you need to be skilled at forecasting to accurately predict the reserves you’ll need. But fear not — miscalculations don’t have to come with wasted dollars. You can actually sell your reserved instances in a marketplace — a smart way to recoup some of that money spent, if you no longer need all of the capacity you planned for. A little bit of effort spent gaming this system can help your company save a lot of money in the long run.
Beyond reserved instances, later stage startups should emphasize smart spending as an essential skill for all DevOps leaders. Ideally, all leaders should have a deep grasp on cloud spending and trends and be able to act like a CEO and financial advisor when it comes to decision making about cloud spending. Automated tools can help with this, too. Pinterest has tackled this issue by appointing “spend captains” to watch cloud costs and sound the alarm about out of control spending. I’d argue that every DevOps leader should have the skills and the know-how to act as a spend captain as part of their day-to-day work. Instead of assigning this role to certain people within your organization, make it a tenet of all operations from day one so you always spend smartly.
In real estate, you reach a tipping point where it’s cheaper to buy a house than to keep renting. The same goes for use of the cloud. At some point, your organization may be so big that your public cloud costs become exorbitant. At that point, consider if a hybrid model or investing in building your own cloud infrastructure will save you money in the long run. It might have been a pipe dream when you were in the early stages of your startup, but at some point the switch is flipped and it can actually become a cost saver. It’s an inevitability for many companies, and the worst thing you can do is wait too long to make a big change (albeit a big investment, too) that will benefit the future of your business.
There are plenty of ways the public cloud can benefit you, and also plenty of ways that it can become a burden for your growing business. By understanding the pros and cons of the cloud at each stage of your business you can set yourself up for success and learn to play nice with the public cloud.
The most important thing to remember is that the cloud is not the end-all, be-all. At any point, it may make sense to switch gears and migrate away from the cloud — or even avoid building in it to begin with. That’s okay. We’ve seen plenty of successful companies forego the cloud, despite the pressure from traditional venture capitalists to go all-in. Our philosophy as investors is that there are many ways to build a successful business — and if you’re on to the next great idea, we’d love to hear from you.