Data Integration Requires Trust

By Len Silverston
Jul 2007

When there is a lack of trust, there may be unnecessary boundaries and political obstacles that are the underlying causes for lack of success in data integration efforts.

Why Trust is a Key to Integration
What is trust? Webster defines it as “Assured reliance on the character, ability, strength, or truth of someone or something.” Trust represents a bond between people that creates a connection. Trust is a fundamental basis for relationships, and without relationships how can we become connected and integrated? If people and organizations are not connected and integrated, data does not get shared appropriately; thus, data does not become integrated.

If we approach data integration, data management or data governance only from a “police” perspective of being an enforcer, then we are jamming down rules and are not creating a healthy environment in the long term. There may be situations where force is needed; however, generally, we want to establish trust so that things go smoother and easier.

For example, on one highly successful integration effort where we integrated data across a state-wide community college system, trust was a key. My client was the data warehouse leader, and he had developed so much trust from his history of delivering on time, within budget, caring about others, and being open. People believed what he said, and they believed in his ability to deliver. In this highly political environment, people were willing to share, work out data issues, and we delivered a very successful integrated solution.

Dr. M. Scott Peck explains in his book, The Different Drum: Community Making and Peace1, “the word ‘integrity’ comes from the verb ‘to integrate.’ Genuine community is always characterized by integrity.” At another client, a large financial institution, I witnessed a genuine community where people trusted and respected each other and each other’s department. This integrity or truth created an environment where people were willing to share data and where a data governance was welcomed.

Dr. Peck further declares in his book, “A lack of integrity always reflects a lack of integration and the reality that something is being left out.” Integrity occurs when you can count on someone, count on their integrity, and in their competence. Psychologists refer to a term called compartmentalization where people without integrity have different parts of their lives that are inconsistent or un-integrated. For example, consider a Sunday churchgoer who vows to have high morals and then on Monday goes to work and dumps toxic wastes into the ocean because his boss demanded this of him. When people and departments become compartmentalized, data becomes compartmentalized or siloed.

On one project for an international association, the head of the organization had all their people carry an embossed card in their wallet that said their company mission included respect for the individual. I witnessed a conversation with a manager where the manager said, “How dare you not carry the company mission card in your wallet! You are a disgrace!” This type of behavior creates an environment where people are protective and do not want to share; therefore, data integration becomes extremely difficult.

In a well functioning machine or system, all the parts work together, like gears moving together, each one doing its own thing yet interlinked and functioning together for the end goal. One of the most successful data governance and integration projects that I have seen was at a large financial institution where they worked together like a well functioning machine. For example, I have talked and written extensively about using the idea of a PARTY model to maintain a single instance of a person and/or organization. Often there is tremendous resistance on this concept with some saying that it is too theoretical or impractical. At this organization, I was amazed that there was no resistance to this idea. When I asked why, the response was that “We trust the data management area to come up with solid common constructs that will serve us all.” I also noticed that they selected a Sybase platform for their database environment, and there was literally no discussion about whether this was the best platform. When I asked about it, they commented similarly, “We trust the database group to make good choices for all of us.”

This enterprise successfully developed an integrated intranet platform for all their investor information over 16 Sybase servers with one common data schema that offered tremendous, recognized business value. When I asked people the reason for their success, people replied “trust in their leadership.” People trusted each other and trusted in the overall organization.

At a data integration effort within a large insurance organization, I noticed the opposite. Teams did not trust each in each other’s motives or character as demonstrated by the indirect comments made behind backs. Teams did not trust in the competency of the other members and they were competitive, vying for individual success. The company did not see results coming out of the data governance and integration effort, so they cut it – even though the needs to integrate data were growing. So they didn’t change the teams, processes, ways of accomplishing data integration, and they did not address the key cultural issues – they simply cut funding for any type of data governance or data integration effort. Even though this program had plenty of resources, methods, tools, technologies, they did not have a trusting, cooperative environment. Thus, the funding was eventually cut.

How to Develop Trust
Three ways to develop trust are:

  • Be vulnerable and open – When we are vulnerable and open, there are less barriers and usually more trust. For example, when we are willing to admit mistakes rather than hide them, this creates trust.
  • Caring about others – When we demonstrate selflessness and genuine concern for others, people are more apt to trust.
  • Earning it – When we demonstrate that we walk the talk, for example, when we deliver what we say we will, when we promised it, this creates trust.

Vulnerability is an important aspect in creating trust. I was involved in a data governance and integration effort for a large distributor. One day while I was there, there was a major issue in their operations. All of the shipments were scheduled out of the New Jersey location, and there were no shipments scheduled in any of the other locations throughout the United States. The whole company was wondering what had happened. One of the people in the data management group realized that he had made an error. He thought he made a change to the development environment and was going to test the change. Instead, he mistakenly made the change to the production environment. Oops! What is the first thing that you would do, and what course of action would you take if you were in this situation? Would you:

  1. Fix it first and then tell everyone what happened?
  2. Fix it and then let it blow over?
  3. Hide it or blame someone else?
  4. Alert everyone right away that you messed up and then address it?
  5. Something else?

The person that made the mistake chose option “d.” He e-mailed all people to let them know immediately that he made a mistake. He left himself highly vulnerable and just communicated the truth. Then he got many people’s help and worked diligently to fix the issue. What did management do? They praised him for admitting the mistake right away. Management believed in the philosophy that people should have permission to fail. This created a trusting environment. This is a company that created a very successful data integration program!

When people know that they are cared about and there is a sense of selflessness, more trust is earned. Why is this? When we are able to get outside of ourselves, we create more connections and we live with a bigger picture. If one is out to advance one’s own career as the first motive, people usually can sense this. In a data management program, if the leader truly thinks about the needs of the entire organization as well as the needs of others as much as his or her own needs, then, in my experience, the data integration program has a much better chance of being successful. In the community college integration program, there was a sincere concern about how they could help other efforts, and this created synergies and ultimately resulted in a very highly successful program.

Another key way to develop trust is to earn it! This simply means do what you say. Keep your commitments. Much more easily said than done. However, it is essential to developing trust and to creating success in data integration efforts.

There is a distinction between having someone (such as your boss) tell you that you are obligated to do something and making a commitment to do something. It is important to say “no” at appropriate times to develop trust. So to establish effective governance, how and when should we involve people when establishing policies and standards that they must follow? A key in the success of both the statewide community college effort and the financial institution effort was that my clients had both developed very strong reputations for doing what they said they would do and saying “no” in difficult circumstances. This allowed them to make sure they delivered what they said they were going to deliver, and they were highly trusted.

Should One Always Trust More?
Does one always trust? What if there are continued mistakes? Or what if someone is known to not be trusted? What if there are politics such that it is not safe to trust?

I belong to a group that shares very openly, and many of the people have commented that if they shared in work what they have shared in that group, they would probably be fired. Who is in charge of our environment – the leaders or the people working in the environment? Can we choose our own path independent of politics?

I am not saying that one should always trust and be open. There are times where it may not be safe. However, in general, I believe that there is a need to develop more trust and openness in data integration programs.

Conclusion
Data silos usually come from people silos. When people do not trust each other, they are usually reluctant to share information. When a department does not trust another department, there may be unnecessary boundaries and political obstacles that are the underlying causes for lack of success in data integration efforts.

I believe that many of the issues in data integration programs have to do with lack of trust. Organizations spend huge amounts of money on data integration efforts; and much of this, at the core, is spent on issues relating to trust. I think if we invest more in learning how to develop trust in our teams and throughout the enterprise, then we can be more successful in data integration efforts.

Endnotes:
1.    M. Scott Peck, The Different Drum: Community Making and Peace, second edition, Simon and Schuster, 1998.

Published in The Data Administration Newsletter