Two Views of the Software Market

A Bridge Too Far?

In a blogger briefing this week at SAP’s TechEd, SAP CTO, Vishal Sikka, drew a chart of the application software market and what new areas of technology are of interest to the firm. I’ve tried to reproduce his freehand drawing into the following PowerPoint rendering. I believe I have captured the essence of his points but apologize in advance if I omitted some of the subtleties of his sketch.

Vishal discussed how new technologies, like social networks, are forcing ERP vendors to process new kinds of information (i.e., unstructured data that may exist in great volumes with little organization), understand the explicit and tacit insights within this data and connect it to the decision making processes of modern companies.

I agree that new kinds of information are presenting themselves to businesses. Sadly, most ERP vendors have often ignored new kinds of information and, instead, focused single-mindedly on those transactions that eventually end up in a general ledger.

Now, let’s look at the broader picture. ERP vendors are behind the eight-ball on several data fronts. Here’s my abbreviated list on this subject:

- event data – Businesses make all kinds of decisions based on external (and many internal) events that are never found in an ERP. For example, sourcing personnel make a number of decisions as to how much to buy, what to pay, etc. based on current commodity price movements. If a key commodity your firm needs suddenly jumped up in price, would your ERP system notify a buyer? I doubt it. Other events are triggered by legislators, competitors, regulators, the press, bloggers, employees, etc. Does your ERP monitor and assess the changing market share and business fortunes of your competitors or is it still trying to perfect last year’s financial statements?

- the context behind data – Can your ERP system scan non-structured data, like blog posts, and determine whether your firm (or a competitor) has suddenly got a product reliability or customer service issue? I doubt it. Most ERP’s are still stymied with problems like how to reconcile the company’s headcount in the HR module with the headcount used in the Budgeting module or the headcount figures in the company’s EEO reports. If an ERP can’t understand structured, accounting data, how can it deal (credibly) with data from less structured sources or data from external sources?

- data for non-traditional users – Your ERP probably makes a bunch of accountants, clerks, data entry people, outsourcers and systems integrators happy. But what does it really do for the significant number of other data-hungry people who are constituents of your firm? Does your ERP have dedicated applications for your board of directors? Your suppliers? Your customers? Different regulatory agencies? Your external auditors? The activist shareholders who are watching your management team? ERP products were designed to satisfy an internal, mostly accounting/Finance user group. While lots of vendors tout ‘analytics’, ‘business intelligence’ or ‘data warehouses’ to address some of these new constituencies, these are mostly bolt-on reporting tools that simply re-work existing, internal transaction data. I really doubt any ERP was designed as ‘business operating system’. Recently, I saw a picture of a horse-drawn cart where the owner had so overloaded the cart that it tipped backward and lifted the horse off the ground. That’s ERP today – an overloaded horse cart. We need vendors who will take a fresh perspective at the totality of a business and its information needs. The incremental approaches of late may work for a while but if left unchecked, create a poor long-term solution.

I got the impression that SAP’s leaders are looking at many of the right kinds of data and business questions but their self-imposed need to do so within the context of keeping everything in the original ERP wrapper may be limiting their vision and potential.

Their desire to push everything into the same technology, same business construct, etc. may not be correct. It feels like they are pushing, a la Montgomery’s Operation Market Garden, for a bridge too far. I’m unconvinced that a solution designed for manufacturers and designed to report, process and store internal transaction data is necessarily the best platform for the businesses of today and tomorrow.

Don’t believe me? I give you a couple of things to ponder. When most ERP systems were designed, technology was heavily constrained. Memory, throughput, processor speed, disk storage, etc. were all in short supply and expensive. ERP designers created systems that worked within these constraints. They intentionally constrained their products to have less than a whole, world view of business. Their products cherry-picked a few key computationally intensive or labor intensive internal tasks and automated them. Over time, as computing became less constrained, more function points and process components were automated. But, at its core, these solutions are still constrained as they were designed as internal systems, using internal transaction data that creates backward looking reporting data.

ERP vendors could create better, more relevant solutions now if they’d just envision a technology world without constraints. When you create systems assuming you have unlimited storage, terabytes of in-core memory, etc., you realize that business information doesn’t have to be internally generated data only. You realize that work is not just comprised of internal work processes (just watch how a sourcing professional does their job – they spend much of their time checking prices of suppliers, analyzing the financial statements of suppliers, etc. – they spend just minutes a week actually keying in a purchase order) but a mix of internal and externally facing tasks. When you realize that the old systems’ views of processes were artificially constrained and limited to an internal view of the world, then you understand that the old data model for ERP is just obsolete, irrelevant and desperately in need of a new perspective.

Data isn't the only sticking point. ERP systems were designed for internal users. They weren't created to serve other constituents like board members, activist shareholders, customers, regulators, suppliers, external auditors, etc. Many corporate constituents were never part of the original data model of these products and bolt-on efforts to mollify/pacify these non-traditional users are 'limited' by design.

Bolting data from social networks, group-think collaborations, web crawling activities, etc. into ERP solutions may tax the ERP data model to the breaking point. Even if these solutions have the technical elasticity to support these new data types, the basic limitations with the inwardly focused old ideas of business, business information, etc. will remain within old ERP solutions. The old ERP design is ready to be replaced with one more relevant for today’s firms.

We need visionary ERP vendors who will take the fresh piece of paper and envision what a new generation of software product should look like. The technology maturity curve for ERP solutions has run its course. The S-curve has hit its apex and has flattened. It’s time for a new kind of product. It’s time for some real innovation and not more of this innovation at the margins.

Vishal's vision looks good and SAP has the R&D resources to make big things happen. But will we see a vendor brave enough to re-imagine what ERP should really be? Or will we see more stuff bolted to the exterior of the old ERP and business thinking of yesteryear? I'll keep hoping for the former.