Integrations; It’s Broken, Who Fixes It?

System integrations come in many different forms, from the manual keying of data between systems, through flat file transfers, screen scraping, APIs, middleware and many variations of each of these. Just a few of the many potential integrations in a typical talent acquisition or contingent workforce technology deployment might be:

• Master Data – Feeds from core systems to provide org structures, cost objects and users.
• Requisition – Perhaps a requisition is created and approved in an ERP or workforce planning solution and transitioned to an ATS or VMS for resourcing
• Sourcing – With an ever-increasing array of channels and tools available to access talent. Connecting an ATS or VMS to relevant channels can deliver significant efficiency.
• Approvals – Perhaps more relevant for contingent hiring than direct hires, but we often see integrations between VMS and cost control tools
• On/Offboarding – Compliance, access management, IT user management and more can all be streamlined by the use of intelligent integrations.
• Time & Expense – For the contingent workforce a common cause of invoicing delays can be traced to having multiple different timesheets, which integrations can address.
• Invoicing – For enterprise clients, electronic invoicing is increasingly becoming a pre-requisite.

However, whether data is pushed, pulled or transferred (pushed and pulled), almost universally every integration involves at least two systems and at least two providers.

Having multiple parties involved isn’t in itself an issue, until something goes wrong. Hopefully, when an issue occurs, the root cause is quickly and efficiently identified and rectified. In reality, this is not always the case.

It is not unheard of for integration problems to quickly get to a situation where the providers of the two systems involved are both convinced that the problem lies with the other system and therefore it is the responsibility of the other provider to rectify the issue. Meanwhile, as the client you find yourself in the situation of having an imperfect or non-functioning process and trying to translate highly technical responses from providers as to why the other party needs to own the issue – stalemate.

So what steps can be taken to minimise the risks associated with this type of problem?

• Error Handling– In my experience, many technical integrations are built without any thoughts towards error handling. Although these capabilities are not a silver bullet, having a level of logging and reporting so that where any points of failure are occurring certainly speeds up the diagnosis of issues. When specifying the design requirements for an integration, be sure to incorporate error handling as a must have. Any extra investment in the build costs will repay itself many times over.
• Primary Vendor– As previously stated, most integrations involve more than one provider, so for each system touch point, agree with your vendors which one is going to assume responsibility for which integration in the event of an issue.
• Governance– Your program governance model is a vital component in issue resolution of all types. Ensure that the escalation paths and channels, in the event of technical challenges, are clearly defined and published to all parties.
• SLAs– SLAs are not just for your staffing providers, but for your technology providers as well. In relation to technical support having SLAs that have teeth e.g. service credits, are a key tool in progressing issue resolution. You may already have support SLAs with your technology provider, but you may also consider drawing up some definitions as to what urgency and severity would apply to the inoperability of different integrations.
• Conference Calls– What sounds like the most obvious solution is often the most effective. If something is not working, get the technical teams from both providers in the same meeting or conference call and replicate the issue in that session so they can collaborate and diagnose the problem. Talking on the phone rather than trading e-mails and tickets will dramatically reduce the resolution time.
• Arbitration– Finally, don’t be afraid to ask for help. Your own IT team may well have many professionals with the requisite skills to liaise with the providers technical teams and act as arbiters to help diagnose and fix the issue.