The demand for software engineers exceeds supply. And this gap only will keep on growing for the next decade. Trends predict that the coming years the software development capacity decrease to less than 25% of the need. The call for more productive engineering methods is inevitable.

Model driven development (MDD) is a solution. Not to be confused with Model Based Development, or “Rapid Application Development” (short: “RAD”), that has been around for years - methods based on code generation but never able to satisfy the needs to offer all required capabilities and a full roundtrip model.

Low-Code / No-Code (LC/NC) is the thing today. As the name implies it is about modelling, configuring and assembling. It is up to 10 times faster -according to pure-play LC/NC platform Mendix. Their competitor OutSystems talks about 75% development time reduction. A similar figure as claimed by Pegasystems, the vendor of a full BPM-ACM/enterprise platform, also based on LC/NC.

LC/NC is very suitable for “systems-of-innovation”, the top most layer in the widely embraced pace-layer model ( http://www.gartner.com/binaries/content/assets/events/keywords/applications/apn30/pace-layered-applications-research-report.pdf ). Basically this concerns small temporary applications with limited scope and short live cycle. LC/NC allows business to quickly deliver IT functionality. Often through citizen development. Business value is generated almost instantly and business users are more flexible and in control than ever before!

Truly a great development... As long as pure-play LC/NC applications are indeed used for systems-of-innovation: short lived and focused.

The spreadsheet analogy

Now an analogy with an old acquaintance comes up: The spreadsheet.

Also a tool intended to quickly solve small temporary business problems. Business users were empowered to gather, analyze and present data. Low thresholds prevented a long application development cycle.

But as we all know: Many spreadsheets did evolve into something bigger. They kept on existing - even after their purpose was fulfilled. Not seldom business considered them “quiet handy for other matters” and thought it was a shame discarding them, wasting the effort building it. So: many spreadsheets lived on and grew into something that eventually was used for operational decision making. And this decision making became part of vital business processes.

And see: The “temporary” excel had evolutionarily become a core business IT asset. Evolved into an uncontrollable stack of data and visual basic code, gluing it all together. Only too late IT departments would find out about these monstrosities - acting like enterprise applications, but lacking the necessary quality. Unmaintainable. Unsecure. Impossible to track. Hard to integrate or reuse.

Parallel

It is not hard to see the parallel here. LC/NC too enables the business to craft nice applications, meeting short term business needs. But what if business users may not be able to suppress the urge of extending their applications, either in scope or in life cycle? What if they -possibly without realizing it- are creating new enterprise applications?

The core aimof pure-play LC/NC platforms is to satisfy business users. They focus mainly on how to deliver functionality fast and pay less attention to non-visible IT/quality aspects as they would make the platform more expensive. Sure: these platforms offer some nice features for workflow, security and integration. But just enough to satisfy business needs to quickly deliver functionality at low license rates.

LC/NC: Which platform: pure-play or enterprise application/BPM?

So: What if the intended Systems-of-Innovation (have to) become Enterprise Systems or part of business processes? Systems that have far more strict quality demands from risk and total-cost-of-ownership perspective. Having advanced requirements relating security and governance; compliance and auditability; process analysis and reporting; maintenance and extendibility; availability and reliability; deployment and scaling; integration and process management.

Would you still address this with a pure-play LC/NC platform? And if yes: How? How tight would the IT governance processes need to be, and what professionality-levels would it take from the IT organization? What additional IT and platform capabilities would you need to get in place to meet the required quality? How would you transform or rebuild the systems-of-innovation realized by business into real enterprise applications? How would you enable reuse of large parts of standard functionality over multiple axis like product, customer, location and channel?

And would the total-cost-of-ownership still be less than an BPM-ACM/enterprise platform with LC/NC capabilities like Pegasystems - Platforms designed to support all those requirements out-of-the-box, allowing systems-of-innovation seamlessly be scaled up to enterprise applications. With all due quality aspects when following the included governance model.

LC/NC are surely the way to move on to cope with the expected IT resource scarceness. The bottom-line question is: Will a pure-play LC/NC tools be the best solution for your demands on the longer terms, or would a complete BPM-ACM/enterprise platform be more beneficiary ?

A legit question to ask when considering a platform for the future.

Over de auteur

  • Harald van der Weel's picture
    Harald van der Weel

    Deze ervaren all-round architect (Enterprise, Informatie, Business, Integratie) heeft zijn sporen in verscheidene sectoren en uiteenlopende projecten verdiend. IT, Energie, Overheid, Finance, Bouw… Verschillende uitdagingen, maar altijd met hetzelfde kritische oog voor kwaliteit en een vleugje Brabantse betrokkenheid. In zijn vrije tijd verkent Harald graag de wereld, te voet of op de fiets, en drinkt hij graag een goed glas wijn.

    Harald van der Weel over SynTouch

    Bij SynTouch denk ik aan… kwaliteit, respect en pragmatiek.

Deel deze pagina