Going the cross ERP way

Ashish Gupta says that over the next three to four years we will see cross ERP products as customers demand more from enterprise applications

All businesses are wired and most software companies accept that their products must talk to each other. The Net is both omnipresent and a still highly attractive environment for applying technology to business processes and problems. Finally, even the most hardened competitors in the software world now accept that it is in their self interest to share previously secret application programming interfaces (APIs) to ensure different companies products work easily and effectively.

Easy sharing of data and applications will happen, but the business world, already questioning the RoI of previous technology investments, will not do business with technology companies that make unrealistic claims and then fail to deliver. Finally, software companies with mature management and a realistic understanding of their market can and will prosper—Neither Rome nor the Taj Mahal were built in one day, but built they were.

The Net has precipitated two of the most fundamental breakthroughs of the last five years:

* A mass acceptance of the state of connectivity: More important, for our current limited discussion, is the acceptance of the state of connectivity at the CXO-level—where a wire running out of the enterprise is not anathema but rather potential for efficiency of some sort.

* Democratisation of the concept of APIs: This is the single biggest contribution of Web services. While the average user of the word ‘Web services’ does not understand APIs, they now understand the top-level concept of two programs talking to each other. Several people use Web services to refer to programs, solutions, data, and computers—as nouns, verbs, adjectives and so on. The point is that few people really know what Web services really means. However, they now intuitively understand the concept and get the fact that Web services enable programs to communicate in ways that were previously not explored.

For instance, pretend that the French have long been aware of secrets about their language that make translation to and from English easier. Nevertheless, they have been hesitant to share those secrets for fear that if the secrets were widely known English would increase its stranglehold as the world language of commerce. Only now do they finally realise that if they share secrets, French will be more easily used and therefore strengthened, not weakened.

Bright future

Opportunities abound, but only for those who take the time to explore and learn. One common factor in both the previously discussed breakthroughs is that no one really understands where most returns will be obtained. While connectivity is good, it is not good enough to understand where it will be most useful from a business perspective—ditto with Web services. In fact, several companies are beginning to think about and implement applications that rely on connectivity and Web services. There are two primary flavours of these applications, and we will refer to these companies as XERP:

* Applications that span established software silos of CRM, manufacturing, finance, and supply chain solutions. Companies like Metric Stream want to measure and report cross-application metrics. Other companies validate cross application logic, some aspects of PRM and customer information sharing.

* Applications that span previously established corporate silos in more ways than before. Examples include newer supply chain companies, demand chain companies and PRM companies, etc.

Opportunity and challenge for cross ERP applications

XERP companies are dealing with the precedent set by Ariba and Commerce One. The implication of the precedent is that newcomers have to not only justify their own RoI but also have to justify why the company spent so much money on its previous solutions. In other words, there is great hesitation in enterprises to buy more software given that their investments over the past few years have not yet returned the desired results in many cases.

Further, given the recent embracing of the state of being connected, customers do not understand where they will realise the highest returns. For example, it is not clear to customers whether maximum return will be realised by connecting the CRM system to finance or by connecting their CRM system to their OEM partner’s CRM systems. There are near infinite choices for the ways and points at which different systems can be connected. Enterprises are yet exploring which connections and/or integrations are the most meaningful. This exploration is expensive and time-consuming because connecting any two systems is difficult. Further, whenever two systems are connected, new functionality emerges, which needs to be built as an application. Thus it is not just a matter of integrating systems but also a matter of integrating systems in the context of new functionality.

It is clear that data and application connectivity are necessary, and that explains the multiplicity of companies addressing the problem.

Inventing the future

While most customers of XERP applications do not yet know where they will realise the most RoI by connecting applications or with partners, they do have ideas about what they want to try first. Unfortunately, for the vendors of XERP solutions, different enterprises have different ideas about the highest return connectivity points. That is:

* Several customers of XERP applications are willing to try focused solutions for specific pain points that they have observed.

* The pain points are not yet accepted by enough customers for XERP companies to sell the same product to several different customers.

* Addressing one pain point is itself a large enough task that the same company cannot try to address different pain points for different companies. Typical integration-based applications require a lot of domain expertise and a lot of dirty detailed work resolving integration points.

* There is a proliferation of XERP companies, each going after a particular integration-based solution based on feedback from a handful of customers.

Given the above factors, we claim that the model for building an enterprise software company for approximately the next five years will be centred on providing solutions. By that we mean that the XERP company will need to start with a very thin layer of customisable functionality and add the rest of the functionality based on a particular customer’s need. This exercise will lead to a better understanding of the customer’s pain and may lead to a slightly fatter layer of customisable functionality. This process will have to be iterated over several different customers and over several years.

The vendor will then understand the common pain points and customers will get an opportunity to learn from each other’s experiences so that they can start sharing a view of the value they can realise by connecting their ERP systems internally or externally. Only when this consensus emerges will there be a market for most XERP products. Only with the education that a vendor gets by implementing several part-custom solutions will the vendor understand what parts of the system really are generalisable and what are the knobs that need to be introduced to capture that generality. Thus, over the next three to four years, vendors will be able to generalise their offering and customers will arrive at a more common definition of what is needed. Then we will see XERP products.

The author is vice president of Tavant Technologies