How to Optimize Your Application Ecosystem

Like it or not, you probably participate in an application ecosystem where you interact with multiple service providers, partners and possibly suppliers to support your business services. Most of these ecosystems are Web-based and many exploit Web 2.0 as well as classic Web technologies. If you’re using any kind of external Cloud computing resources, even if it’s just to extend your infrastructure, you’ve already become dependent on a highly dynamic and quite possibly disruptive community of services.

So, how do you manage your most critical applications when they may be impacted by 10 or more service providers, and the multiple business partners who now share in extending your business model? And, how do you optimize your infrastructure when large chunks of your infrastructure are becoming invisible?

Some research done in conjunction with Internet infrastructure services provider Neustar, revealed there are tons of commonalities across many different types of ecosystems. Just for starters, if you happen to be in the Internet game, B2B looks a lot like B2C in the complex interdependencies you have to capture and manage. But the same is true for healthcare, financial services, manufacturing, retail, and high technology, etc. In other words, as long as you have some critical dependency on Web-based applications for internal productivity, or how you come to market, or how you partner to extend your business model, or how you deliver your critical business services externally—you should be asking yourself the following questions:

  • How do I create effective service level agreements (SLA) to support my business commitments and objectives that include all of my key partner and service provider interdependencies?

  • What kinds of guarantees can these service providers give me that answer my real business requirements versus just what’s convenient for them to monitor?

  • Is there any possible way to get these service providers and partners to be responsive to me as a community versus just an independent set of resources; a problem made all the more challenging since some of them may compete with each other?

  • How should I share information with them?

  • How should they share information with me?

  • What process requirements do I need to put in place that now extend beyond my own IT organization to deal with, for instance, collective problem resolution issues (incident and problem management), or security, risk and compliance issues?

  • When it comes to optimizing my infrastructure, do I just give up, or do I try to rein in these many parts with something like a federated CMS?

  • What monitoring tools work best across ecosystems, and how do they combine with my existing monitoring capabilities?

Before you panic, here are a few hopefully helpful hints on how to grapple with potential ecosystem chaos:

Take the time to understand how individual outsourced services impact your top-tier critical business needs. For instance, you may be dependent on a credit card processing applet on your website for financial transaction processing of various kinds. If so, that may not fall into the category of something you need to prioritize as top tier for monitoring―unless of course it’s been a problem for you in the past, or if you are dependent on it for your core business transactions.

Guarantees are negotiable. You’re not alone in facing the bigger issues of ecosystems. Some service providers are already grappling with a parallel set of challenges and therefore should be more “ecosystem-friendly” in their negotiations with you.

Sharing information across ecosystems is paramount and research shows that it’s on the rise. There are, by the way, real parallels here between a single IT organization’s growth in sharing information across domains and information sharing across partners, clients and service providers in an ecosystem. Similar processes and policies must ultimately apply to both.

Closely related to the prior point is the very real demand for new or expanded types of best practices and process definitions to extend beyond the enterprise, and even linear enterprise-service-provider relationships. These new process definitions will be designed for a community of interdependencies and embrace new types of technical, and more importantly, political and cultural barriers.

Processes for change and configuration management across multiple organizations are also beginning to get attention in more progressive environments. So don’t give up on a CMS requirement if you face competitive hosted data centers. I’ve already seen CMS initiatives succeed in enterprise situations targeting, interestingly enough, a cohesive and non-disruptive way to virtualize infrastructures that span multiple, competitive hosted environments.

On monitoring tools: you will need to monitor your services from both outside the firewall and inside the firewall. Those outside the firewall are typically software-as-a-service capabilities such as those from Neustar, Keynote and Gomez/Compuware .

Probably the best reassurance I can give is that just by being aware of your “ecosystem situation” and taking the time to understand it, you’re already ahead of many of your competitors. Beyond that, applying common sense and prioritizing meaningful and pragmatic phase-one objectives can go a long way in helping you to make the right choices as you progress from “reactive” to “proactive” in your approach to ecosystem management.

Dennis Drogseth is VP of Boulder, Colo.-based Enterprise Management Associates, an industry research firm focused on IT management. Dennis can reached at [email protected].