Thursday, March 18, 2021

Enterprise Buyers Not Looking for a One-Stop Shop

There's been an interesting discussion on Twitter over the past few days, which I started with this deliberately ambiguous tweet. 

IMO, very few enterprise buyers are really looking for a "one-stop shop." 

As intended, that brought out replies from several friends and associates, such as Vijay Vijayasankar, Oliver Marks, Holger Mueller, Jody Lemoine, Shane Bryan, John Appleby, and others. 

So, what did I learn from the dialog? 

First, I was thinking back to client meetings I've sat through over the decades, where business leaders positioned "one-stop shop" as a key element of their desired strategy. 

In other words, in the market we serve, customers are typically looking for 10 things.  But today, we only offer seven. If we can offer all 10 things, we can become a one-stop shop! Customers will not have to go anywhere else but will have the convenience of having us satisfy all their needs. 

In enterprise software, this might translate to an ERP system vendor attempting to offer a CRM system or supply chain management suite, or product data management, or a host of other complementary products. Invariable, because these systems take years to develop from scratch, in practice this means acquiring those complementary products. It may also mean offering other elements of a complete solution, such as a development platform, tooling, system integration services, even databases or hardware. 

I don't know if Oracle ever used the term "one-stop shop," but it certainly behaved as if it had. It has been on a multi-decade acquisition spree, not only in business applications, but also in databases (its roots), infrastructure software (BEA), even hardware (Sun). To be fair, it also plowed profits from those products into new development, such as for its Fusion cloud applications. And it is now competing with Amazon for cloud infrastructure services. It is a poster child for the one-stop shop. 

SAP has had its own version of the one-stop shop, acquiring a variety of systems (Holger calls some of them the seven sisters). It also built its own proprietary database, and it also has its own development tooling. 

What About One Throat to Choke? 

One can imagine why such a strategy might be attractive to technology sellers.  But is it attractive to technology buyers? 

I say, no.  In decades of consulting, I don't think I've ever heard a client say, I just wish I could buy everything I need from a single vendor. What I need is a one-stop shop. 

But isn't a one-stop shop the same as "one throat to choke?" I say no. One throat to choke means that in a system implementation, for example, there is a prime contractor or service provider ultimately responsible for delivery. If another partner in the deal is not meeting its commitments, the prime contractor or service provider serving as overall program manager is responsible.  It doesn't mean that there is only one service provider or vendor in the deal. 

What About Integrated Suites? 

Holger asked, "Are you saying that [integrated] suites are done?" Not at all. But I have two responses to this. First, many integrated suites are anything but.  Especially if, as noted above, the vendor built its suite from piece parts that it acquired over time. It takes years to integrate software acquired from various sources. So, buying from a vendor attempting to be a one-stop shop does not ensure you are really getting an integrated suite. 

Second, I have seen very few large deals where there was only a single software provider in the deal. There are almost always complementary products whether they be for sales tax reporting, factory data collection, data analytics, or countless other niche requirements. 

Third, no IT organization's application portfolio only has software from a single vendor, not even a handful of vendors. Even small companies buy software from dozens of vendors. There is no one-stop shop in enterprise software. 

What About Application Rationalization? 

But what about vendor consolidation? Maybe one vendor isn't reasonable, but isn't it a good idea to limit the number of software providers and rationalize the applications portfolio?  Certainly, many companies need to consolidate applications, especially if they grew through mergers and acquisitions and now have two, three, or more ERP systems, for example. 

But that does not mean they need to only buy from one vendor. 

Vendors love to talk about vendor consolidation, as long as the surviving vendor is them. They call this gaining in their "share of wallet," as in the buyer's wallet. 

In my view, when it comes to vendor consolidation you can have too many vendors and you can also have too few. You don't want to have so many vendors that you have redundant types of systems. On the other hand, you don't want to have too few vendors to the point that they gain leverage over you.  

To this point, I've heard of customers engaging in multi-year programs specifically to reduce dependence on certain Tier I vendors, as they become too powerful and attempt to engage in wallet fracking, as my friend Brian Sommer calls it. 

Is there a way to have the benefits of integration and applications rationalization without becoming overly reliant on a single vendor?  I think there is.  Modern cloud systems have become API-oriented. And to be fair, the major vendors, even those aspiring to a greater share of wallet, are building with this model. They have to, if they want market acceptance. Cloud leaders, such as Salesforce, do it by providing a platform that partners can write to, even leveraging Salesforce objects, to provide that integration. Oracle's NetSuite offers a similar capability. Cloud ERP vendors, like Acumatica, Plex, and Sage Intacct are very integration-friendly. Oracle's cloud applications and SAP's offer open APIs, as does Workday. Microsoft has similar capabilities. 

If this is the future, then maybe vendors will give up the strategy of the one-stop shop. 

Wednesday, March 10, 2021

Deploying Low-Latency Applications in the Cloud

Cloud has become the preferred deployment option for most categories of enterprise systems. But conventional wisdom is that some systems that require low latency and a high degree of system availability, such as warehouse management systems (WMS) or manufacturing execution systems (MES), are best deployed on-premises.

The argument is that response time over the Internet is never as fast as over a local area network and that such systems cannot tolerate any level of unscheduled downtime inherent in using a cloud application. 

Nevertheless, cloud systems are invading even this category of software. Although still in the minority, there are some vendors providing such low-latency applications as a cloud service. 

One example is Plex Systems.  And, interestingly, it is not a new example. 

Read the rest of this post on the Avasant website: Deploying Low-Latency Applications in the Cloud.