Sarah Lacy argues that enterprise application software is due for a change. Her reasons seemingly boil down to:
- Users are increasingly eager for friendlier, consumer-like technology.
- The current generation of apps was installed long enough ago — often before the Year 2000 deadline — that enterprises are willing to contemplate rip-and-replace.
I’m inclined to agree, although I’d add some further, more technological-oriented drivers to the mix.
Changes I envision to enterprise applications include (and these overlap):
- Better integration with communication technology.
- Social software.
- Better stakeholder-facing interfaces.
- Voice control.
- Better integration with analytic technology.
- Dashboard-first UIs.
- Search-first UIs.
- Alert-first UIs.
- Analytic assessment aids (job performance, supplier desirability, expense approval, etc.).
- Automated decisioning.
- Some true analytic apps, interesting or otherwise.
- Better use of different kinds of data.
- Analytically-derived data.
It’s natural to ask how this all relates to two increasingly important internet use patterns — mobile and SaaS (Software as a Service) technologies. Actually, the connection isn’t very tight; I see these changes mainly as trends that would and will occur even in the traditional world of desktop clients and on-premises server software.
That said, I’d think hooks in all this to mobile technology can be found among:
- Social software.
- All of the UIs.
- Some of the machine-generated data.
SaaS delivery seems relevant mainly insofar as:
- In general, SaaS could make it easier to implement next-generation applications.
- With its rapid upgrade cycles, SaaS can make it easier to adopt newer technology, especially in the area of user interfaces.
- salesforce.com is a large SaaS company that happens to be leading the push toward integrating social technology into operational applications.
A little more explanation
Back in September — influenced by what I heard at salesforce.com’s conference — I suggested that the opportunity to use social software to communicate and collaborate around almost any kind of business process could yield major improvements to operational applications. If nothing else, it could be used by companies to help users — not necessarily their own employees — fight through the awful interfaces many apps exhibit.
There are numerous ways in which your systems are or can be opened up to users beyond your own employees. Examples include (and these overlap):
- Customer-facing websites.
- Anything to do with supply chain … even including social-software-based collaboration.
- Stakeholder-facing analytics.
This is all a fertile area for application innovation, often revolving around user interfaces. After all, if you want somebody to use your software — and you can’t force them to do so by the power of the almighty dollar (Euro, Lire, shekel, clamshell …) — then probably you should at least ensure that doing so is not a hideously unpleasant experience.
Speech recognition interfaces have been slow to evolve. But much of the reason has been a lack of utility due to hardware shortcomings. If you’re at your desk with a headset on, speech recognition has long worked — but that’s also where it’s least needed.
I don’t know enough electrical engineering to understand exactly what the issues are. But I do believe that Apple Siri is a sign they’re finally being addressed. That increases the incentive to reengineer certain enterprise applications so that they can be commanded in more simple ways than is currently the case.
Text data management is a confusing (and confused) subject. In particular, it’s hard to find one good technology stack on which to build applications managing both text and relational data. Still, it’s long been possible to put text fields in relational databases and then treat them like text documents for most purposes. And increasingly, it’s going to look quaint to do CRM (Customer Relationship Management) without heavy reliance on relational and text data alike. Ditto supplier relationship management. Ditto medical records. Ditto a few other application areas as well.
Machine-generated data has been getting huge amounts of attention, and rightly so. People want to monitor it moment-by-moment; then they want to analyze it as well. But I’m hearing more about operational apps around machine-generated data too. Certainly, that could be said to characterize a lot of internet personalization and digital media use cases. Electricity “smart grid” applications are coming to the fore now too (for example, SAP is proud of a “Smart Meter” application). Hospitals have had their own little machine-specific applications for a long time; more in that area lies ahead. It feels like we’re due for another round of security technology upgrades. In general, as an ever-larger fraction of data is generated by machines, an ever-greater fraction of operational application technology will address the data those machines produce.
OK. This has gotten long enough. I’d love to keep going and say more about analytic/operational application integration, but that’s going to have to wait until future posts.