Friday Rant - moving to a hyperscaler is NOT digital transformation

Profile picture for user brianssommer By Brian Sommer August 21, 2020
Summary:
The spinning is intense as tech firms try to position old tech as a necessary antidote for the pandemic. But can old tech enable digital transformation? Is lift and shift a remedy for the times? The truth is in the motives of the players (not the technology).

ranting
(Pixabay )

It seems everyone in the technology industry has a quick-fix pitch to get your #COVID19 affected firm on the road to digital transformation. If only this wasn’t a way to separate your firm from some of its hard-earned cash.

Here are three sales strategies being deployed today.

The lift and shift journey mirage

My diginomica colleague Jon Reed and I were speaking the other day about the digital-transformation-washing going on right now. Never to let a good crisis to go to waste, integrators and others have been packaging the movement of clients’ computing loads to their own private cloud environments as some kind of miracle solution during the pandemic. 

The marketing hype, or logic, if you will, for these journeys seems to follow this tortured plot:

  1. Take your on-premises application portfolio. 
  2. Prepare to move it to another firm’s data centers. 
  3. Utilize virtual machine technology to facilitate this. 
  4. Re-work all of your integrations to cloud applications. 
  5. Pay this third-party tons of service fees to monitor service issues, track service tickets and/or maintain applications, systems software and databases.
  6. Eventually, you might feel that you’ve gotten comfortable enough with this starter set of cloud computing and will want to move this environment again to a major public cloud hyperscaler solution provider.
  7. At this point, you can now declare victory as you’re “digitally transformed!

The reality is that this is an expensive death by a thousand cuts. You’re being helped by a self-serving firm that wants to collect a lot of service fees by lifting and shifting your application portfolio a couple of times. That same firm also wants to keep its army of service professionals chargeable while they maintain your applications on their cloud. Instead of getting your firm to a public cloud immediately, they want to collect years of fees. And, of course, none of this gets you digitally transformed. You just moved your problem applications around – nothing more.

There are variants to this scenario, too. One frequent variant asks you to move some of your on-premises apps to a third-party/private cloud service. At that point, you can upgrade some of those applications to a single-tenant cloud version of the applications. The problem with this approach is it’s a two-step dance AND the customer is still on the hook to keep their applications current. To do so, they’ll either need their IT group to do this or use a third-party firm (e.g., an integrator, application maintenance outsourcer, etc.). 

In none of these scenarios is the customer digitally transformed. These are teeny, incremental systems changes that do not materially alter the business, change its business model(s), enhance its competitiveness, etc. They also don’t utilize NEW sources of data, especially Big or Dark data. This is like putting a bow and ribbon on a worn-out pair of shoes – you can call it “New” but it’s not. 

Pandemic and moving to Public Cloud

Some integrators are pitching a direct move to a public cloud or hyperscaler for a specific pandemic rationale. The intent is to provide a way for Work From Home (WFH) employees to get easier access to the company’s applications. Apparently, your closed-off, on-premises solutions were difficult for WFH employees to access. The theory is that it is easier for remote employees to access applications and data from a hyperscaler’s public cloud than it is to so from a business’ unique IT architecture that was not optimized for large numbers of remote access users. 

This approach is not without challenges though. If the old solutions are simply ported to the hyperscaler’s cloud, it doesn’t mean that the applications themselves were changed. If you needed a remote access, VM or VPN bit of software to get to the application before, you will still need these. In a WFH world, you’ll like need more of these licenses. Moving to a hyperscaler doesn’t make your single-tenant applications multi-tenant. It also won’t introduce new/better application and data security unless your firm is willing to do that work. In fact, you probably could have upgraded that on your old system. 

What this move could do is enable your IT people to support applications and their computing environment remotely. They wouldn’t need to come into your on-premises data center to maintain much of anything. 

The other big advantage to such a move might be scale. If the existing technology in your data center wasn’t designed to handle so many remote users, then the hyperscaler option helps you quickly avoid a costly capital expenditure. 

Yet, nothing in this scenario is getting your firm closer to a digital transformation existence. Jon Reed opines: 

It’s just more lift and shift.  Remote access via the hosted "cloud" is a stop gap. Don’t confuse it with a step on the path to digital transformation. From a transformation view, moving your data silos to the cloud accomplishes nothing. 

What's even more galling is some vendors who provide this service will tell you all about their ML and AI capabilities. If there's one thing ML doesn't like, it's data silos with limited access. Taking full advantage of AI is about standardizing core systems, opening them up via secure APIs, and feeding AI systems information without having to put a team of twenty legacy programmers to work building customer interfaces - which an "AI partner" will happily charge you for.

To Jon’s point, I’d add that great ML tools can benefit greatly from utilizing data aggregated across thousands of software customers not just one firm. For example, you’d get better preventative maintenance suggestions from a database with hundreds of millions of data points than whatever data you firm has (or might ever collect). This is why so many cloud HR products have rich ML-based capabilities around engagement, discrimination and other issues as they evaluate data from thousands of customers. A cloud-hosted, single-tenant, clone of an on-premises solution will deliver sub-par ML results because it is starved for digital inputs – the opposite of what a digital transformation requires. 

The path to a digital future?

And then there are the software vendors (and their implementer partner ilk) who want to take you on a “digital transformation journey”.  It’s a journey alright! This path can be expensive, time consuming and full of steps. It also won’t create much value unless you and these implementation partners are willing to do a lot of work. 

This journey involves companies slowly moving off of their old, on-premises application software. The first step is to accept a newer version of the applications running on a new platform. Amazingly, some of these platforms are just single-tenant cloud solutions often running on private clouds. As such, these products may not get the automated upgrades to new releases/upgrades that a multi-tenant public cloud product would get.  

Worse, some of the new versions of these applications have less functionality that earlier products. Yes, the vendors have an answer for that: you need to adopt their standard process designs and practices. In that one statement, they are advising you to forgo competitive advantage and be just like your competitors. 

Vendors and implementers will then encourage you to consolidate the number of versions of their software that you are using and they’ll want you to cut out a lot of the customizations, too. While you’re at it, you can also eliminate a lot of old technical debt. 

But, in the end, all you have is a slightly newer version of your applications. It’s like when I put a new set of tires on my Civic. I restored the road worthiness a bit but it was still essentially the same old car. 

But, that’s not the promise the vendors and implementers are selling. They say there are more chapters to this strategy and some of those haven’t been written yet. Specifically, they maintain that you have to have new software with a more modern backbone/platform to take advantage of advanced technologies (e.g., machine learning), big/dark data and more. And you also need patience and a big checkbook to get those advanced capabilities as the best features may still be on the product roadmap.  

Smart buyers might want to investigate offerings from non-ERP vendors to see what’s actually possible today. Vendors like Aera, Uptake, Celonis, C3.AI and more might recommend you simply ring-fence your old ERP and let it do its traditional transaction processing while their Big Data/IoT/etc. tuned systems can handle the advanced technologies and business problems. 

This kind of “journey” only makes sense if you really do your homework, check out ALL the options out there, do a great job of building a business case, etc. 

My take

I’ve seen tons of digital misfires and few transformations. I even wrote a book on this subject. I was stunned by the amount of work being expended on incremental improvements – improvements that were, at best, minor improvements to the firm and its users. What companies weren’t doing was looking at the universe of potential changes and who could help them. Seeking advice from the implementer and ERP firm you used 20+ years ago was rarely a gold mine of actionable and competitively advantageous capabilities. 

The best firms see what is the art of the possible BEFORE committing to any new projects. They develop insightful, forward-looking plans that describe a new to-be state and how they’ll get there. Then, they rationalize these plans against the capital, human and other resources/constraints the company has. Only then do they proceed with the digital transformation. 

Caveat Emptor folks!