Rethink: A New Generation of Business

For thousands of years, Zen masters have taught us the value of living in the moment. Don't ruminate about the past, or fret about the future. Just live in the now, they say. Easier said than done, perhaps, but most people understand the intrinsic wisdom of this mantra.

In the business world, such an approach has been traditionally eschewed. There's too much at stake, too many salaries on the line, so many suppliers, and all those customers in need of their products and services. Besides, there hasn't been a sufficient technological solution to enable a business to run in such a real-time capacity, to run "live" as some would say.

No longer! A confluence of innovations over the past half-decade has opened the door for an entirely new paradigm of business: real-time. As is often the case with information technology, the innovation wave started at the hardware level. And although the hardware in question isn't new, the cost and level of refinement have combined to forge a new, remarkably different future.

Towards a new paradigm

The tech in question is memory. Long a critical component of any computer, memory has dropped in price so much that many organizations are now embracing information architectures that are stored entirely in-memory. Merely a dream 10 years ago, this new reality poses radical changes for how data is captured, stored, processed and analyzed. A new paradigm is here.

One of the leaders in this space is veteran German software supplier SAP, which began architecting its HANA solution roughly a decade ago. Anticipating the price drop, and recognizing the sea change that an in-memory architecture would yield, SAP mapped out an ambitious future for its many well-heeled customers. 

Central to SAP's ongoing success is a robust partner ecosystem. Company executives realized years ago that working in tandem with consulting firms would provide ample opportunities for client engagement, while freeing the organization to focus on its core strength: software engineering. One such partner is VISEO, which employs 1,200 in France and around the world.

In a recent InsideAnalysis interview, Emannuel Deverre of VISEO, explained the significance of the in-memory revolution: "The big change came from the technology side with the in-memory," he said. "This technical change made many things possible. We could start looking at the data on direct live access. Everyone was doing operational reporting using a data warehouse where they worked to replicate the data. They consumed the data with BI tools. But now with the in-memory data, you can directly go and consume this data live. There is no more need for replication."

Eliminating the need for replication represents a massive change in data management, one that results in many benefits. As Deverre explained in the interview, the process of replication began in the days of data warehousing, after many organizations realized they could not efficiently do operational reporting from within their enterprise resource planning (ERP) systems. The processing power just wasn't there, and as a result, the practice of data warehousing evolved.

Over time, using a data warehouse became commonplace for large organizations who sought to better understand cost structures, pricing, customer engagement and other aspects of the business. In order to feed these warehouses, a complementary industry grew to prominence, primarily with the bread-and-butter functionality of ETL (extract-transform-load). Companies like Ascential (acquired by IBM), Informatica, Ab Initio and others dominated this market.

But even at its height of popularity, a warehouse was arguably no more than a necessary evil. After all, it introduced many additional costs, and also required a data pipeline that was fragile and error-prone. Even the most sophisticated information architectures that centered around a warehouse were commonly plagued with significant levels of redundancy in data movement; and almost invariably led to multiple versions of the truth. 

A complete in memory solution

There were also political issues in many organizations that led to user frustration, and ultimately spawned another cottage industry of data warehouse appliances, such as Netezza, Greenplum, Dataupia and numerous others. In many cases, frustrated departments circumvented the warehouse and created their own prodigious data marts. Although these solutions generated targeted results for their users, a side result was the proliferation of additional SILOs. 

By architecting S/4 HANA to be a completely in-memory solution, SAP sought to solve this significant conundrum. The raw speed of in-memory computing enables an architecture that feeds both transactional and analytical processing, a combination that for years has been considered nothing short of the Holy Grail in enterprise computing. This tectonic shift has massive impact on those fragile data pipelines, which are now largely obsolete.

Deverre noted: "Now with HANA and in-memory technology we can bring back the operational reporting straight within the ERP. For all the people within the company, that's the only way you can drive your business processes at the end. How can you monitor your business processes if you don't have the correct and the realtime operational reporting? Now we have the answer for that... At the end, we're back to the promise we made a few years ago."

Delivering such a real-time view of the business has other major benefits, especially in the realm of forecasting and planning. In the warehouse age, planners would need to request specific subsets of data in order to do their forecasting, or hash out new ideas. This was problematic because of the latency, which could range anywhere from days to weeks. By the time they had enough data to make their projections, the situation on the front lines of their company could well have changed, and dramatically.

Rethink the business processes

In this new, real-time environment, planners can make much more accurate determinations about which supplies they'll need in order to satisfy customer demands. For example, in the manufacturing realm, tremendous amounts of time, effort and money can be saved by streamlining processes, and only ordering the components necessary for current or near-term orders. 

Deverre noted: "You have full control in real time of how your business is going. You don't need to wait for a day plus one and then see what's going on. You can directly see what's going wrong. You can take actions directly on that. You can also simulate, which was not possible in the past. In the past when you were running your MRP (Material Requirements Planning) processes for replenishment, or whatever, or your saved forecasts."

"There was no way you could simulate the sync because the system could not technically handle any simulation. You had to wait for one day migrating all the data to another system and then time simulating. Three days later you could potentially have an input of what could be a simulation. Now you can do that directly in real time. That makes a big difference."

Perhaps the broadest area of benefit comes from the ability that companies will now have to completely rethink their business processes, and even their business models. In an age when Uber can upend the taxi business, or Airbnb can throw a serious curve ball at the hotel industry, having the ability to quickly redesign business processes, and invent whole new business models, is a critical success factor.