kosmin - Fotolia


B2B processes get smarter in the cloud

Major cloud platforms are offering more and more analytics tools. It's a development that should be beneficial for the continuing growth and alignment of B2B partnerships.

The evolution of B2B processes and technology has been slow, but inexorable. Increased connectivity between partner companies, viewed with suspicion early on, gave way not only to greater efficiency all around but to the need for sharing internal data, and eventually integrated systems and processes.

Cloud technology has accelerated that integration considerably, and is opening doors to even greater efficiency and performance: B2B partners can now share not only channels and data, but also analytics.

And the same cloud technology that facilitates the first two can now be leveraged for the analytics, as the major cloud platforms continue rolling out baked-in artificial intelligence (AI) and analytics staging tools.

AI wars

In the B2B processes arena, last year was very much the "year of the smart cloud." From Microsoft Azure Cortana Intelligence Suite to Salesforce Einstein to IBM's Watson Analytics, baked-in AI and cloud analytics are now more convenient for businesses with B2B processes based in the cloud. Each of the major vendor tech conferences unveiled something new and exciting, and businesses seeking to add analytics to their shared B2B resources now have a number of platforms to choose from.

It's important to realize that there's very little "me too" among the cloud offerings; each vendor has chosen a different direction in how this new technology is deployed and used.

Microsoft, for instance, created its Cortana suite with build-it-yourself in mind. As with its SQL Server BI tools a decade ago, its strategy is to entice customer uptake by offering easy-to-use components for building analytics processes, rather than offering actual canned processes.

With service-oriented approaches like IBM Watson, there is unprecedented convenience and simplicity -- no expertise required -- but this is a problem in itself.

As an example, B2B partners could consolidate shared schemas (purchase orders and invoices, customer profiles, schedules, and so on) within a shared data lake in Azure, to simplify integration of local systems. Participating partners could reference common data in an Azure data catalog, to simplify discovery; and these resources could easily be made single-point accessible to in-house systems and other external processes via a data factory, making it possible to integrate gradually, rather than all at once.

On the other hand, many applications of analytics are more straightforward, rendering a building-block approach as overkill. The best use of shared analytics in some B2B partnerships might simply be a common platform for uploading data sets to be scrutinized for process-improving trend analysis. For this, the IBM Watson Analytics tool set might be a more prudent choice.

And some cloud intelligence is already deeply wedded to processes crucial to B2B processes: CRM, for instance. Salesforce, the industry leader in CRM, offers its Einstein intelligence suite, which comes already integrated with its CRM platform and tools. Businesses can begin using it immediately, without building anything.

Lower cost of entry

Apart from shared resources and the impetus to standardize within B2B partnerships, these platforms offer a lower cost of entry for those alliances interested in leveraging these new capabilities. It is simpler to buy cloud subscriptions and share them than to purchase new software and servers and come to agreement on their maintenance and administration.

It is also less costly when participants across several organizations are learning the same technology, rather than how to integrate different ones. Knowledge transfer between companies becomes not only cheaper but faster.

Words of warning

All of this said, there are still some drawbacks to smart cloud technology, especially when it is shared across partner companies.

On the build-it-yourself side, there is administrative overhead that does not exist with other approaches. Commitment to the Microsoft Azure platform requires at least moderate expertise in that platform within each participating organization (though this need not go as deep as developer-level). In addition, an extra layer of administration is required to create and implement shared standards in the management of schemas, data and connectivity.

With service-oriented approaches like IBM Watson, there is unprecedented convenience and simplicity -- no expertise required -- but this is a problem in itself: When submitting a data set for analysis, the user has no real idea how the resulting analytics are generated. This leaves a question mark hovering over such operations, as "analytics" isn't a black box (or shouldn't be, in any case). Different analytical operations are suited for different kinds of problems. It's too easy to trust processes that may be yielding non-optimal results.

Finally, with intelligence that is completely bound to existing business software, there's the problem of having to invest in a proprietary platform to access the value at all.

It's the beginning of a new era in B2B processes and technology. But, as with all such giant leaps forward, a great deal of collective forethought and planning is called for.

Next Steps

IBM urging partners to adopt Watson

API strategy important for B2B processes

What you need for B2B integration in the cloud

Dig Deeper on ERP and supply chain management