If your career included a technical phase it’s likely that the first project you were involved with included integration as a deliverable. The last IT project you found yourself involved in probably included integration as a deliverable as well.
It would be unsurprising if they were the same project.
Regardless of the amazing coverage of your ERP or CRM, or the depth of a new point solution … wait. I need to start that sentence over: Not “regardless,” but “because of” the amazing coverage of the systems you have or are in the process of implementing, they will have data and functional overlaps. To take an easy-to-understand example, your ERP and CRM systems both manage data about your customers somewhere in the depths of their databases.
Failure to integrate them means that any time you want information about a customer, or knowledge about customers in the aggregate, the two systems will disagree.
These disagreements are gaps. Your business sponsor will either find them, or hear about them from someone else who did. And, they’ll have sufficient sophistication to know the gaps could “easily” be closed by integrating the systems. And they’ll want to know why you didn’t take this obvious step.
Let’s role-play the conversation you would have to have with the business sponsor to get yourself off the hook. You’ll need to ask the business sponsor a few questions, to help them understand the tradeoffs the team will need to make in order for the project to move forward.
Let’s rehearse a few of the points that go into this conversation, starting with:
- Where does one system start, and where does the other one pick up the mission?
Where systems overlap, that is, which is the source of truth? Breaking this down further, what we are really seeing are three overlaps—Overlapping Data (both systems might need a street address), Overlapping Functional Logic (both systems need to make sure that a delivery is going to a valid location) and Overlapping Business Logic (both systems are involved in order fulfillment).
To say this gets messy is an understatement. Ideally, if you ask your CRM and ERP systems the same question about an order, you should get the same answer, in terms of payments, fulfillment stage, delivery location, billing location and customer. But depending on how your solution will synchronize them, the answer might be to let them disagree. Is this okay? Which gets to the next question:
- If you must choose, which system needs to be “right”?
In our conversations with our colleagues, we do need to ask which system should be considered the System of Record, which systems depend on the information from this system, and when they all must agree.
- What’s the flip side of the coin?
How often, that is, do the two systems need to re-synchronize? Near-real time? Overnight through a batch process? At month’s end as part of closing the books? This is when you give your business sponsor the bad news about synchronization: The closer we get to real time, the more complex the engineering and the higher the cost. Not to mention the higher cost. If the business sponsor wants real time or nearly so, are they willing to pay for it?
- Every system has data that is in some way, shape, or form, “dirty.”
CRM systems, for example, are really, really good at helping you stay connected to customers. They’ll track every interaction imaginable. They are also notorious for creating an almost schizophrenic portfolio of contacts that are, in fact, the same person, but with one letter in their name different, slightly different addresses, birthdays, and so on. It is not uncommon to have 10+ entries associated with the same human being. Which of the ten should your ERP system synchronize to?
It’s a good question with no right answer. The dirty-data problem mucks up expensive marketing campaigns, recalls or RMAs … even interpersonal interactions. CRMs are likely the worst offender, but not the only one for introducing bad data to other systems.
- Is data cleansing in your company’s future?
Without it you’ll never finish implementing the new system. With it comes expense, implementation delays, and the certainty that three years from now you’ll have to cleanse all that data all over again.
Will this conversation with your business sponsor be easy? Sure it will. Conversations about trade-offs are always fun and games, aren’t they? But especially with your business sponsor, and then recapping the results to your team, you are going to gain trust and build alignment—which might at least make later conversations easier for everyone.
Great article. Keep up the good work.
re: Is data cleansing in your company’s future?
The second most expensive and difficult aspect of implementing a new system (matching/mapping processes is first).
Unfortunately, it is often an afterthought, rather than a deliberate part of the project.
That is a good point!
Yeah, as the old proverb goes: A man with one watch knows what time it is. A man with two is never sure.
Data “cleansing” is another way to create dirty data… It is extraordinarily difficult to do it well and we almost always settle for less than that!