This syndicated post originally appeared at No Jitter - Recent posts by Zeus Kerravala.
Don’t look for a hockey stick-shaped adoption curve just quite yet.
As I mentioned in my previous No Jitter post, “Thinking About the Cloud 2020,” I had been tasked with being the prognosticator for cloud 2020 as part of the one-day Enterprise Communications & Collaboration 2020 conference-within-a-conference that took place on the opening day of Enterprise Connect Orlando. During my presentation, I shared my vision of a world where the cloud enables predictive, contextual communications, driven by machine learning, and where information is pushed to us, when we need it, making our personal and professional lives much simpler and more efficient.
My panelists — from 8×8, BroadSoft, Cisco, Genband, and Microsoft — mostly agreed with my prediction of what cloud 2020 would look like. Knowing that technology shifts are never easy, we spent some time discussing the barriers to the cloud 2020 vision. Here are the main points:
- Security and privacy — Without a doubt, this is the top concern regarding the vision of cloud 2020. Getting even part way to the utopian state requires users to be willing to freely share information like location, preferences, social network, address, and other data. There is a school of thought, and it’s certainly valid, that people will not willingly provide the necessary contextual information because of privacy and security implications. I understand why people have this opinion, but I fundamentally disagree with it. Once the cloud enables us to do things we could not do before and removes the manual effort required to do many day-to-day tasks, I believe people will opt-in to services that require them to provide data. The key is to make the services compelling enough. In the business world this isn’t as big of an issue as it is in consumer markets since the organization owns documents, emails, blogs, PowerPoints, cellphone records, and other sources of information. The business may choose to keep the data private, but again if the business value is captivating enough, I believe most will opt-in — and that will put pressure on other companies to keep pace. It’s up the cloud providers to convince customers that their data is indeed being kept secure and the risk of a breach is minimal.
- Federation of information — Machine learning and artificial intelligence can only be effective with data. The more data there is, the more accurate the results. The problem is that no one company owns all the data and the cloud providers must be willing to share their information. This is being done well today in small pockets, such as with cloud contact center and CRM providers, but in general we see very little data sharing among cloud providers today. Sharing data will create value for everyone, bringing us to a contextual cloud. Open platforms, APIs, and co-development must be the norm if the cloud 2020 vision is to be a reality.
- Total cost of ownership(TCO)/return on investment (ROI) uncertainty — One of the audience questions was how one could actually calculate the TCO of a cloud solution, compare it to an on-premises version, and figure out the ROI. The only way to do this is to make the new stuff look like the old. If a customer leverages UCaaS simply to replace what the organization was doing with on-premises solutions then, sure, calculate the cost of service, MACDs, etc. Productivity costs are likely to be harder to calculate, but it’s possible. With all due respect to bean counters, however, asking about the cost of one versus the other shows a lack of understanding of what you can do with the cloud but not with an on-premises solution. This is analogous to trying to calculate the ROI of email or Internet back in the mid ’90s. No calculations were needed, as the business could no longer be competitive without either.
- Picking a platform — Literally dozens of companies are now trying to do machine learning in the cloud. Cortana, Google, Einstein, Siri, Monica, and Watson are all names we have become familiar with as platforms that can perform analytics and AI. However, should a company choose and standardize on one, or should it use them all for different uses case? Try and get two or three that interoperate? Unfortunately, there is no right answer to this question as we are at the very start of this evolutionary cycle. In the short term, I recommend that businesses should start small with a vendor they can trust and then expand from there.
- Long migration cycles — With many technologies, particularly with communications, the evolutionary arc is much longer than experts anticipate. Many organizations still have legacy PBXs that are in the range of 15 years old that should have been replaced half a decade ago. However, they still work, and the industry hasn’t created a compelling enough case to get some customers to ditch working equipment.
Personally, I think 2020 is an aggressive date to get us to the next phase of the cloud. I think we’ll need more time to resolve the issues I’ve outlined above and to develop compelling use cases before we start to see a hockey stick-shaped adoption curve for the cloud.
Latest posts by Zeus Kerravala (see all)
- Forget Heliocentrism-Embrace the cloud and Zeus-centrism - April 17, 2017
- Avaya’s post-bankruptcy plan should not impact customers, partners - April 14, 2017
- There’s never been a better time for Cisco Services - April 12, 2017