In the previous post, I discussed ground rules for competing with AWS. In this, I will share specific thoughts for startups and larger companies. The good news is that it is possible to compete against AWS and win, but you need to be disciplined.
- Focus on specific use case.
Your product/solution has to solve specific use cases that have found a market fit. You do not have the luxury of building a dependent business. An example of a dependent business is” if Docker succeeds, we expect an increased demand for x, so I am going to build a product for x” . Solve a current pain that customers have. Your goal is to become the best known for that use case and thus make it harder for AWS to make a good enough solution.
- Deliver the best UX in the category
This is an area where you can beat AWS hands down. It is actually easier to build UX better than AWS. But, you want to shoot higher and build an UX that your users would love. I will give you an example. We currently use NewRelic. I signed up for a preview of AWS X-Ray and was turned off by the difference in initial setup experience of X-Ray. So , suffice it to say we will remain Newrelic customer for several more quarters. I bet there are many customers out there like that. NewRelic nailed the UX from initial setup to using the product. Can you do that for your chosen market?. ( To be fair, I did not give a fair shot to AppDynamics, so do not know if they are similar or better or worse than NewRelic ).
- Make pricing completely friction less
I have discussed about AWS outdated pricing model. They have lot of catch up to do. You can hit them hard on this. Make it an elastic pricing. Make it low risk to get started with your product and not need yearly subscriptions. You may still lose some large customers to AWS because of ELAs etc, but you can still have a major chunk of market by removing pricing fiction.
- Build a passionate community of users
As a startup, you will never have enough marketing dollars to create awareness and buzz around your products. Your users and community will help here. Investing in community and nurturing them will help you put speed bumps in AWS path. In a best case scenario they may come to you and want to OEM your product, like they did with Chef. Its not that they can not create a similar product, its that they know it takes much longer to create a similar community.
- Do this before AWS announces product that competes with you
You have to do all these before AWS announces a product in your market category. The intent here is to create doubt for AWS strategy folks that its going to be a hard and long slog, so they will find it hard to recruit necessary talent. If you are not already a leader in a market and AWS announces a product in this space, its game over for you. Pivot or find something else to do.
Its harder for larger companies to win, unless they are willing to make some tough decisions. ( I am not talking about Google or MSFT, both of which have head start in trying to compete with AWS ).
- IAAS battle is over
You need to invest in infrastructure and build capabilities throughout the world. But you are not going to win IAAS battle by lower prices or marketing. That battle is over. You need a strong IAAS to compete at higher levels.
- PAAS battle is over
Similarly, PAAS battle is over. Grudgingly ( because my gut didn’t say it back then), I acknowledge what Simon Wardley predicted 3 years ago that Cloud Foundry has won. I don’t see long term prospects of a PAAS game riding on coattails of a container orchestration platform.
- There is an opening in Serverless
We are still at year zero of serverless. AWS has slight advantage, but nothing insurmountable. In fact, I would argue that a large company can bridge this gap faster. The tech is not the hard part, its getting the business model right and aligning the company revenue officers to go along with it.
Serverless needs quite a bit of developer tooling. This is not a new problem, it is solved well by other companies. IBM and Oracle employs lot of the talent that solved these problems. IBM owns people that built Rational, that build a whole Java stack with WebSphere. These folks can be leveraged to build developer tooling for serverless. But they have to be unchained from existing revenue streams from WebSphere and Rational product lines. A separate BU that focuses on serverless with no constrants from existing sales and marketing teams is the way to go here. Similarly, Oracle owns Sun , BEA talent and even has people that build Application Development tools for Oracle, PSFT, Siebel and other apps. This talent can also be leveraged.
However, I do not see either IBM or Oracle having the courage to fork off a new BU and pursue serverless yet. Why? because revenue forecasts for that BU would look meager. Sales commissions would look so bad, there will be mutiny inside the company. So, its likely that the only competition to AWS will continue to be Microsoft and Google.
These are my quick notes before going offline again. Thoughts?