Welcome to this week’s Think Tank with Frank Kenney
With decades of analyst and integration industry experience, Frank Kenney is a fountain of knowledge on all things tech. Now, he aims to share that awareness with you. Come back every other Thursday for your biweekly dose of thought leadership in this blog from one of technology’s most insightful thinkers and gain perspective on a variety of topics ranging from what’s happening in integration today to what’s on the horizon, poised to disrupt the integration space going forward.
Let’s try an interesting experiment. I want you to dive into your integration logs. Perhaps, you’ll need to scroll through them like a long, and needlessly complex novel. Perhaps, you’ll tap an elastic search API to take a slightly less manually intensive method to pull out all of the purchase orders and settlement notifications that you have received electronically from your business partners over the last 24 hours.
Tally up the numbers. You’re probably looking at one sizable number representative of the total “business” that you’ve completed in a full day. Now, multiply that number by 10. Where does the number 10 come from? You’ll see in a moment.
Imagine there’s an issue
Let’s pretend that business didn’t happen. Let’s pretend for the sake of argument that those purchase orders got hung up on an expired digital certificate attached to your AS2 connector. Some firms want you to renew that certificate every year, some firms want it done every other year, and still, many companies don’t care!
But let’s say these orders didn’t get through. Look at your current business processes and systems and tell me how quickly you would be aware of an exception or an outage. Would you know when your customers called you? Knowing it is only a portion of the battle; however, how quickly can you resolve the error? And finally, how quickly can you reprocess the transactions without impacting the current workflows?
Amplify the impact by 10
Now let’s get back to that 10 number. The reason I asked you to multiply by 10 is so that you acutely feel the pain. Because there are very few instances, usually in legal proceedings, that you can crisply articulate the monetary damage to your brand. The monetary damage to the trust of your customers and your partners have in your ability to serve them. Yes, errors happen. It’s the reason why some of the largest infrastructure providers only go to “five nines.” It’s not the error, it’s the resolution. Specifically, it’s that critical time to resolution…
But you’re okay. You’re working with a managed service provider that is proactively monitoring your service level agreements (SLA). You’ve negotiated agreements up front, and you know that there is mutual and contractual understanding of accountability. At the very least, you can finger point and negotiate the continued level of “trust”.
But you’re working with an integration provider whose specialty is internal application integration or cloud application integration. The impact of errors is slightly diminished because it happens to the “back office” applications. Salesforce connection is down? That stinks! But I can pull out (and potentially dust off) the old black and red notebook. Workday connection is down? Man, does that gum up the network! But I can still queue up items to process. That simple AS2 connection go down? Darn, I’ve missed my window for bidding load to tender! Darn, I’ve missed my window to acknowledge a purchase order from my biggest retailer! Darn, I’ve just violated an SLA!
Doesn’t feel so simple anymore.
What’s the point?
Here’s the point.
Well, really, here are two points.
1. Not all integration service providers are created equal.
The real trick to choosing the right provider is how much skin are they willing to put in the game? How much risk are they willing to take? That answer comes in the form of guaranteed SLA’s.
2. Some iPaaS vendors only provide guarantees for the environment.
This arrangement naturally discounts all the aforementioned revenue-generating transactions, and also the integrations executed within said environment. When an exception happens in this scenario, you are most likely responsible for recognizing the cause and managing the issue through to resolution.
But, maybe that’s okay. Because most of the time when you’re buying an iPaaS, you’re buying a toolset. But there are times, exception-al times (see what I did there?) where you shouldn’t be asking for a can of paint and a brush… especially if you aren’t an assuredly skilled painter.
Perhaps, rediscovering the art of effectively managing integration exceptions doesn’t necessitate you to be the artist. Rather, you get to walk the halls of a grand museum and appreciate the mastery of exceptional artistry.
Today, your business should know when to spring for the managed service versus doing-it-all-yourself approach to integration exception management. Self-service was a powerful capability when supplied in the right places and for the right things – you oil the chain, not the seat. Today’s self-service is yesterday’s toolbox approach, very useful in some cases but not applicable for everything.
Maybe, the approach we should be thinking about is a blend of self and managed services? (Shameless plug number two.)
For a bit more on the topic of services models, this is a great post from my colleague and member of Cleo’s product management team, Dave Butcher.
About Frank Kenney
A former Gartner analyst and current market evangelist and strategy director, Frank Kenney is widely credited as the creator of the term managed file transfer (MFT) and was the first to write about and discuss its modern architecture, platform, and use cases. Previously, Frank served more than 10 years as a research director at Gartner, where he defined the MFT, B2B gateway, SOA governance, and cloud service brokerage (CSB) markets.