10 Best Practices for Integrating Your Customer Data

Scribe Software White Paper

10 BEST PRACTICES FOR INTEGRATING YOUR CUSTOMER DATA!

WRITTEN BY BETSY BILHORN VICE PRESIDENT, MARKETING MAY 2014 PIERRE HULSEBUS SENIOR SALES ENGINEER

The more information you have and the more effectively you use it, the greater degree of success you can achieve. Begin by establishing a common lexicon between all members of your team and all of your stakeholders so there is no misunderstanding or misconstrued intent or action along the way.

CUSTOMER DATA INTEGRATION WHERE DO I BEGIN? Information fuels success. The more information you have and the more effectively you use it, the greater degree of success you can achieve. The best of all possible worlds is achieved when all of the systems you use to manage information are tightly integrated so data from each can easily be combined to grow a deeper, more intimate understanding of each customer.

Depending on the size of your organization and the number of contacts you interact with, integrating these systems to accurately share customer data can be a formidable task.

Perhaps youve already undertaken this project, or perhaps youre in the process of persuading your executive management to approve or fund a project to do so and need help building your business case.

In either of these cases, this white paper focuses on bringing you the insight, the experience, and the information you require to leverage customer information most successfully.

In most enterprises today, information about customers and the business we conduct with them is managed by a variety of systems including:

Enterprise Resource Planning (ERP) applications to track transactions, inventory, accounts receivable and payable and more. The ERP tends to see each customer as a singular entity, a company, relating all relevant data to that company.

Customer Relationship Management (CRM) systems that are more complex than the ERP in that they see each customer company as containing many departments and individuals, and tracks sales, marketing and service activities related to those individuals.

Marketing Automation (MA) solutions automate and tracks performance of many of the marketing activities tracked by CRM.

Social Media, newest of all, creates new pathways between you and your prospects and customers.

Scribe Software Corporation 1750 Elm Street, Suite 200 Manchester, NH 03104 USA P: 1.603.622.5109 F: 1.603.622.3862 www.scribesoft.com

REPRODUCTION WITHOUT PRIOR WRITTEN PERMISSION IS STRICTLY PROHIBITED

1

Our 17 years of integration experience with thousands of different customers and their integration projects has produced the following set of 10 best practices to employ when approaching a customer data integration project of any meaningful magnitude.

So, where do you begin? Here are your first ten steps.

YOU SAY INTEGRATION, I SAY INTEGRATION

Shared understanding of what you mean by integration assures better integration!

Its never safe to assume that everyone is on the same page. Assure it instead!

Whether you mean combining, sharing, or referring to information, integration must result in people and the tools they use working better together to achieve common goals!

Any project of the magnitude involved in bringing together multiple databases to aggregate related data will be highly complex. Begin by establishing a common lexicon between all members of your team and all of your stakeholders so there is no misunderstanding or misconstrued intent or action along the way.

The word integration itself is often defined differently by different people. Some think of integration as a migration of one thing into another. Others use the word integration when describing the connection that allows access to certain resources. For some, integration is the process of providing functions from one system to another. An example of this is having a screen-pop triggered by an incoming call on a phone system. The data from the company ERP is integrated into the phone switch.

Still others think of business process integration, such as combining the activities of multiple departments, or multiple companies during a merger.

In the context of a customer data integration project, it is important that every stakeholder and every person involved share a common understanding of the end goal of the integration activity. It may be the enabling of a sales team to develop a deeper understanding of their customers to increase their closing ratio. It may simply be the automation of absorbing new leads from new sources into a prospecting list. Whatever the desired outcome, it is crucial to the success of the project that everyone share the same understanding of what that outcome is.

10 BEST PRACTICES FOR INTEGRATING YOUR CUSTOMER DATA >>> 2 PUBLISHED BY SCRIBE SOFTWARE CORPORATION 2014 ALL RIGHTS RESERVED

2

HAVE A PLAN WITH A MEASURABLE OUTCOME

With so many moving parts, its crucial to have a well-developed plan

Information aggregation is easier to achieve when you do it in stages, not all at once

Establishing, tracking, and achieving very specific metrics is your best strategy for assuring continued management support for your project

The saying A journey of a thousand miles must begin with a single step has been attributed to Lao-Tzu, the founder of Tao Te Ching.

While there is great wisdom in this observation it is also true that, for anyone starting the journey toward integrating customer data, that first step is preceded by extensive planning. That planning begins with establishing a clear set of metrics by which to gauge the success of the integration, the fulfillment of the goals of the project.

Accepting that the integration task itself is a large one, it is also important to break it down into smaller phases to bring a higher likelihood of success. Rather than attempting to boil the ocean all at once, it is recommended to seek milestones to measure progress, and to approach each of these more achievable steps in sequence building to the larger goal.

Beyond setting achievable and measurable objectives it is also important to be very specific. If the goal is to increase the number of new leads entered per day into a system through automation, by what percentage will we increase that number? If the goal is to reduce the equivalent of Full Time Employees (FTEs) assigned to a task, how many FTEs do we intend to reduce? What actions will cause this reduction to occur? In this way we can precisely define what success will look like, what the intended outcome of the integration will be.

If the goal is to realize a return on investment (ROI), how quickly will that investment be recovered through savings and/or increased revenue, and then by what percentage will it continue returning per year.


3

THE REAL BUDGET

Know your all-in budget, anticipating as many hidden costs and surprises as possible

Dont limit cost to funds. Include time, skills, labor and other resources in your budget

Consider the comparative cost of outsourcing certain functions

There is a difference in definition between the price of something, and its cost. For a retailer, that difference may be the cost they pay to obtain merchandise and the price they sell it to a customer. Subtracting the cost from the price yields the all-important profit.

For the project or operational manager, however, these two terms are defined differently. The price signifies the initial investment required to commence the activity, such as the purchase of capital equipment. The cost refers to the ongoing total cost of ownership (TCO), the ongoing cost of operating and owning the resources required to achieve the desired outcome.

Costs take the form not only of funds, but also of other consumables such as time, skills, labor, research, learning curve, licenses, real property, equipment, energy, and other resources. All of these must be budgeted as part of a quality planning process.

The budget process requires a careful examination of past activity to review how things were done in the past and at what cost. Then you can analyze why each either succeeded or failed.

Since the budgeting process includes ongoing cost, we must also consider the level of commitment required to sustain the investment and see it through to completion. How will various business processes be impacted by the project? If resources need to be re-assigned from daily operations, how will those be backfilled?

What functions may be outsourced? If an installation or modification is only going to be performed once, does it make sense to commit resources to learning how to do them, or does it make more sense to engage an external resource with experience performing these one-time tasks elsewhere?

Some refer to this as the all-in budget. Whatever you call it, what really must be allocated to assure the success of the project is a question that must be answered before taking that all-important first step?


4

Beyond the

fundamental data structures of your ERP,

CRM, and MA systems its key

to understand what the custom

workflows are and how their

dependencies and connections are

structured.

KNOW THY SYSTEMS

The deeper your understanding of each of the systems involved in your integration project, the better

Remember that reality often varies from the documentation. Anticipate inconsistencies

Many integration tasks are one-time events. It may be more cost-effective to bring in an experienced installer for these

By definition, the high-level goal of any integration project is to make things work well together. To even begin such an undertaking requires a thorough understanding how each of the things works individually.

Beyond the fundamental data structures of your ERP, CRM, and MA systems its key to understand what the custom workflows are and how their dependencies and connections are structured. What processes are triggered by each action, and which data entities are involved. An understanding of the metadata layers, how the workflows impact each other, and an understanding of what the cascading effects of integrations may be will also be important.

APIs

Armed with a thorough understanding of each involved system, youll need to develop thorough familiarity with the Application Programming Interfaces (API) provided by each. How do they permit you to map fields in one system to their corresponding counterpart in the others? How do they pass identifying information, authentication credentials, and specific data elements to each other?

Remember that the promise of a given API doesnt always match the reality. The fact is that many APIs were originally purpose-written for a specific use case and may have been modified or adapted to some extent for general use. Since your goals may not align with the original intent of the API, the quality of those adaptations may not always provide the facility required for your integration.

Some more mature products have APIs that have aged and not been updated or replaced. You may now want to use them in ways that werent originally conceptualized or envisioned. This legacy effect may negatively impact the efficiency with which you can use these APIs to accomplish your intended integration.


By definition, the high-level

goal of any integration project is to

make things work well

together.

Its also critical to develop a clear understanding of vendor limitations that may be imposed by trademark, licensing, or other rules for the data. There may be storage fees required. Using the API to access a given system may consume a user license that must be purchased. It is far better to uncover these parameters prior to creating an integration that depends upon resources you may not have rights to.

Perhaps the most valuable resource you could have would be the experience of people who are familiar with the realities of the APIs involved. What are the idiosyncrasies of the code? How does the real performance of the API vary from the documentation? How will performance differ from a testing environment to an actual production environment? If they are available, people who have utilized these systems and these APIs can be invaluable.

One last note on experience: There are many upgrades, migrations, and installation activities that are one-time events. That is, once you have performed them once you will likely not be called upon to do so ever again. Determine the relative value of having an external provider perform these rather than investing the extreme time and training funds that can be consumed when having your own people learn how to perform these one-time events. You may often find it far more cost-effective to have them done for you.


5

The most important system you will integrate

will be the community of

people who will actually use the

integrated systems and data.

As important as it is to know your technology systems

and data structures, it is

even more critical to know your user

community.

MAP TWICE, INTEGRATE ONCE

You cant get where youre going unless youve carefully mapped your way there

Involving stakeholders every step of the way always improves mapping processes

Your ability to choose the best tools for your integration project absolutely depends upon careful and accurate mapping of information, processes, and peoples roles

The most important system you will integrate will be the community of people who will actually use the integrated systems and data. As important as it is to know your technology systems and data structures, it is even more critical to know your user community.

Your planning must begin and end with them. Begin by understanding what you will be integrating and how those integrated systems will be used. Put the entire process down on paper. Map each process specifically and in full detail. To do this will require that you sit down with all of the stakeholders and users involved and walk through each process. Get screenshots from each involved system to aid your understanding, and really fill that use case out before you develop your plan of action.

When we refer to mapping, were not only addressing the mapping of like fields between databases, were also focusing on mapping out business processes step-by-step. You may map these processes using Visio, for example, and then detail the related data field mappings in an Excel spreadsheet, but everyone involved must agree on how things are actually done in real life to assure that the systems adapt and support those activities effectively.

This process mapping must be accomplished before any tools are purchased or approaches are considered as these decisions will often be deeply informed by the outcome of this all-important set of conversations with the users and stakeholders. Some tools will be more appropriate in situations where data must be moved from one system to another. Other tools are more event-driven, enabling related processes to be triggered by specific actions.

Our point here is your integration designs need to be helpful for picking approach and platforms. You must achieve agreement about what youre actually going to do.

Note: This step all too often gets skipped. Theres a degree of tension between traditional application design and more modern waterfall processing. The traditional way was to develop, design


6

and deploy. In the more modern, more agile approach, we discover changes as we go along. Get started and then hold workshops to discover more information. What you see today is a hybrid between planning and learning as you go.

GARBAGE IN, GARBAGE OUT

Bad data discourages users from adopting new systems, slowing your project to a stop

Good data is good in the context of the system currently using it

Cleaning data properly almost always requires multiple passes, multiple processes, and painstaking vigilance

It is obvious that bad data will always have a negative impact upon your results. What is not often considered is the impact bad data can have on user adoption.

When you evaluate system initiatives that have failed in the past in an attempt to determine where the project went off track, the likelihood is that the root cause was the failure of users to adopt the new system. This may be caused by a failure or lack of training, but more often than not it is caused by the users perception that there is no value for them in the new system. This perception is often created by a lack of data quality.

When data is clearly erroneous, the typical user will not assign blame to the data. Rather, they will see the system as failing. More than one occasion of bad data can easily cause users to simply stop using the system and find ways around it, usually using what they had previously used to accomplish the same functions.

Data Context

Each of the systems involved has a different focus, a different context that the data exists in, and may see specific data entities differently.

For example, an account in an ERP is usually a company that does business with you. Of course, if your finance department is looking at it, an account is really a categorization for a given income or expense flow that exists on a Chart of Accounts in the general ledger. By contrast, your CRM sees an account as a container for many contacts which may have a wide variety of activities and events related to them.

The ERP is focused on invoices and inventory. The CRMs focus is far deeper, a far more intimate understanding of each customer and each contact within each customer company. The CRM may be considered to have more focus on the hierarchy within the


Careful review during the integration process can

often uncover flawed or less

efficient processes that can readily be

improved during the integration

process.

customer organization; specific people and specific actions. These variations in context and focus from one system to the next must be considered when integrating data flows between them.

Also be sure to review any business processes which may impact the data flows. Careful review during the integration process can often uncover flawed or less efficient processes that can readily be improved during the integration process. You may also preclude finger-pointing later as you prevent one department from accusing the other of having less-than-accurate data.

Cleaning the Data

The ability of various systems to fully clean up data in any given database improves every year, but there are still going to be situations in which humans must review and cleanse the data, preferably prior to completing any kind of integration of those databases with others.

Casing the data is one good example. Many older users became accustomed years ago to clicking the shift-lock before entering data, so everything was entered in ALL CAPS. Of course, nothing looks more artificial than incorporating such data into a mailing. Nothing screams bulk-mail louder. One goal of modern marketing is to personalize the experience for each customer, and quality data is required to accomplish that.

Anticipate stopping your integration processes occasionally to assure that the data is clean and accurate. Include a staging process wherever possible so that data is staged somewhere so it is convenient to review and revise before completing each integration process. It may be more efficient to order the data by geography, or alphabetically by company or contact name, or some other sequence that facilitates rapid review. In this environment your data goes from source to staging where it is reviewed and cleaned, and then on to the new target application.


7

SELECTING AN APPROACH

Highest efficiencies are achieved when the process you create today can still serve you the next time and the next time

Break down the project into manageable tasks and prioritize each task prior to execution

Be careful to only make changes that will survive the next vendor upgrade. Stick to the APIs wherever and whenever possible

Scalability is always a key issue in any IT planning. Allowing for future growth avoids the need to discard investments and replace them with greater capacity solutions. This presupposes that one solution can and will do it all if it has sufficient capacity or potential capacity.

This is not the case when it comes to something as intricate as customer data integration. In most cases, integration will require a variety of tools and technologies to be available for use in various different ways at different times. To paraphrase the popular saying, customer data integration is not a nail, so a hammer alone will not suffice. One size tool simply does not fit all.

Decisions regarding what approach to take and what tools to use to accomplish customer data integration should not be made based primarily on available technology, but rather on the needs of the business at any given point in time. Some integration tools excel at the orchestration of transitions, and others are designed to be point-to-point migration tools.

Prioritize

Begin by breaking down the larger integration into smaller, more achievable pieces. Attempting to perform a complete integration of everything all at once increases the possibility of failure and frustration while approaching it in phases makes each phase easier to accomplish allowing for steady progress with meaningful gains at each step.

The first phase should be a comprehensive planning exercise. This will help to set and manage the expectations of all stakeholders while also making it easier to present and obtain approval for your business case.

When selecting tools to help accomplish each phase, stay focused on the importance of maintaining agility for the long run. Some of


Remember that users quickly

become impatient with anything they

perceive as not providing

reliability and value.

8

your platforms may include built-in integration tools beyond APIs. Remember that these were not necessarily designed with your other platforms in mind. They may or may not be fully compatible. They may also lock you into a singular approach selected by the platform vendor which ultimately may limit your agility.

The same is true of many free integration tools that may become obsolete or incompatible at the next release of any of your platforms, or of the tool itself. Select tools that will continue to serve you independent of the actions of any of the developers of any of the technologies involved in your environment. This was the driving force behind the development of APIs which has helped many to preserve and enhance their agility over time.

Remember that users quickly become impatient with anything they perceive as not providing reliability and value. In todays cloud-based IT market users can easily obtain external solutions by subscribing on a website using their credit card. It is easier than ever for them to become rogue users who operate outside of IT direction. This can cause confusion and later increase your ITs support requirements. Make certain that your integration planning benefits from their observations and input. Users are the most important component to integrate tightly into your plans and your approach.

DESIGN VS. PERFORMANCE

Resist achieving performance increases by going directly to the data. The APIs preserve your connection to the solution developer and their upgrades

Exercise caution when deciding to replicate data instead of integrating it

When you must replicate, be sure to include an intermediary quality assurance step so you dont transfer bad data

Design each phase in the context of a viable metric that is based on business value. How many records must be transitioned in a specified period of time? How accurate must the data be as a percentage of the total volume of data? This directly impacts how much you must budget for each phase, what the anticipated duration of each process must be, and what type of tools and technologies will be required to accomplish each.

The API Economy

We mention in the previous section how developers may modify their applications over time. This includes changes to the fundamental structure of databases, to the overall schema of the


9

data employed in each platform. Having to then modify your own customizations and modifications created during integration of systems can be a very expensive proposition in terms of lost productivity, lost time, and repeated expenditures.

This is why we speak of the API economy. While the database might change, and the schema might be modified the API is a programmable way for a developer to connect directly into the application. They can then pass a user name and password from one application to another and programmatically trigger a desired activity. The API exposes the application in a secure way allowing you to interface programmatically with their system. Instead of direct access to the database you are getting access to an application layer.

While it is seems it would be faster to go directly to the data, performance can become an issue as each access requires authentication and careful checking to maintain database integrity.

Integration vs. Replication

Remember also that there is a substantial difference between integration and simple replication. In many cases you may simply want to access and display specific fields of information to facilitate user action instead of replicating that data across to another system. The replication could include millions of records which would introduce unacceptable latency.

When replication is definitely required, consider using a three step approach that takes the data from the original source, replicates it to a staging area where it can be quality assured through various verification processes before it is finally passed along to the target environment.

SETTING EXPECTATIONS

Manage your users response to your system first and foremost. Its valueless unless they use it

Find or create champions within the user base to help promote your cause

Commitment to specific results is critical, but remember that preserving some agility always helps

One of the most important things technology support personnel are taught is to fix the customer first. That is, be sure to carefully set and manage the users expectations. Often users will not understand what will happen when, and how it will happen. In many cases, not knowing what is possible, they will want and expect the impossible. This must be managed from the beginning and throughout the integration process.

The best way to assure effective user expectation management is


10

to make users a constant part of the process, from consulting them during initial integration planning, to building in frequent User Acceptance Testing (UAT) rounds in your process. Not only will this assure that your approach is addressing their needs, it will also make them your ambassadors to their own teams as they go back and report on their progress with the integration of the systems.

Much as technology providers issue a Statement of Work to inform their clients as to what will happen and what will be accomplished during any given project, it always helps to give your user community tangible evidence and clear communication of what can and will be done during the integration of systems and what functionality they can expect as a result.

Keeping agility in mind, be sure to constantly remind users that your statement of work can and will change over time as a result of their input and other factors. This assures them that they have input into the process, and sets their expectation that nothing is set in stone.

CUSTOMER INVOLVEMENT

The real measures of success are the results users get from using your integrated solution

The best way to prevent users fears is to keep them involved hands-on as much as possible

Your project is only successful when your users are successful

Customer data integration is not an abstract exercise. It does not end when the applications are sharing process and data. In fact, it does not end. People use the integrated platforms on an ongoing basis. Properly executed, your integration activities will enhance the ways in which users work, increasing their productivity and therefore their job satisfaction. While your intended metrics may focus on number of records transferred, leads input, or hours saved, the real metric will be the adoption and acceptance by the users over time.

Any change is met with some fear, some trepidation. Part of your goal must be to mitigate these concerns by offering users many opportunities to get hands-on with your integration project as often as possible. Eliminate surprises. When users are disappointed by the results of your integration that may delay adoption resulting in broken workflows and the inability to actually achieve your goals.

Remember that data only gains value when it is in motion, and it only goes into motion when users are actually using the data to accomplish business objectives. Involve the users early in your planning and keep them involved throughout the project. This is


The more you learn about

each contact the more finely you can tune your

targeting to reach each one

with just the right message

through just the right channels.

the best way to eliminate surprises and design for enthusiastic acceptance, adoption, and use.

GETTING STARTED - LEARNING MORE

Integration of CRM, Marketing Automation, ERP, and social applications is best approached gradually and organically. Start by identifying the highest value understandings you want to obtain, and also by identifying the understandings you can gain most quickly. In this way youll enjoy some early success at integrating relevant data while also establishing pertinent goals for the process. Trying to architect and build an overarching integration strategy will most likely result in more delay, more frustration, and fewer immediate results.

Dont expect the knowledge to come solely from within your organization. Consult with experts in the field of data integration. They can guide you to the right tools to use to define business rules and snap the data between systems. Upon identifying a new anonymous visitor the right tools can locate an identifiable piece of info that takes the data back to a prospect master record in CRM and captures that change.

The more you learn about each contact the more finely you can tune your targeting to reach each one with just the right message through just the right channels. By taking a stepwise approach and committing to it long term, youll immediately begin to enjoy more value from the data youre already collecting, allowing you to engage prospects and customers on a far more personal interactive level.


About Scribe Software Corporation

Scribe is an established global provider of solutions that easily bring customer data anywhere it is needed regardless of IT infrastructure. Scribes award-winning products help 12,000 customers and 1,200 partners use customer data cloud-based, on-premise or a mix to increase revenue, provide superior service, and create business value faster. Its easy-to-use, enterprise-ready solutions are backed by extensive support options and training, and service customers across a wide array of industries including financial services, life sciences, manufacturing, and media and entertainment companies. For more information about Scribe please visit www.scribesoft.com.

About the authors

As Vice President, Product Management Betsy Bilhorn is responsible for leading Scribes new product initiatives and guiding the companys corporate and product marketing efforts, including the execution of Scribes go-to-market strategies.

Prior to joining Scribe, Bilhorn held several leadership positions at SaaS pioneer WebTrends, including participation in the evolution of WebTrends OnDemand. She has also held leadership positions at system integrator, Lease Dimensions, where she managed large deployments of global CRM, ERP, and financial systems, including acquisition and integration. Her clients have included American Honda, Baxter Healthcare, Cisco Systems, Ford Motor Credit, HP, Transamerica, and Volkswagen.

Senior Sales Engineer Pierre Hulsebus has over 24 years of experience in information technology, sales, and marketing, his duties take him all over the world meeting with organizations to help them understand how to integrate complex customer data. After 10 years in the Hardware and telecommunications side of the business, he started CRM consulting in 1999 and has since been involved in hundreds of CRM projects. He has worked with Microsoft Dynamics CRM since its release in 2003, and built a successful Microsoft CRM Gold Certified consulting practice. Pierre is a nationally-recognized expert in Microsoft Dynamics CRM, and complex data migration and integration projects.

Scribe Software Corporation 1750 Elm Street Suite 200 Manchester, NH 03104 USA Tel: 1.603.622.5109 Fax: 1.603.622.3862 Email: [email protected] www.scribesoft.com


Customer Data Integration Where Do I Begin?You Say Integration, I Say IntegrationHave a Plan With a Measurable OutcomeThe Real BudgetKnow Thy SystemsMap Twice, Integrate OnceGarbage In, Garbage OutSelecting an ApproachDesign vs. PerformanceSetting ExpectationsCustomer InvolvementGetting Started - Learning More

Documents

10 Best Practices for Integrating Your Customer Data