Strategic-Approach

Business Intelligence, often referred to as 'decision support' or 'CRM analytics' when aligned with CRM software systems and customer strategies, empowers decision makers to better understand, analyze, forecast and impact business performance. Business intelligence software tools help transform raw data from multiple sources into useful information and distribute this insight to all who can use it, when they need it, in order to improve decision making timeliness and accuracy.

As with other enterprise software applications, implementing business intelligence (BI), decision support or CRM analytics software tools without accompanying strategy, business process support and IT alignment will risk implementation, challenge adoption, and likely not achieve objectives or return on investment (ROI). This advisory proposes a 10 step strategic approach to successfully plan, justify and deploy a BI or CRM analytics solution.

The Business Problem

Management's ability to consistently make timely and accurate business decisions—at both strategy and operational levels—is extremely influential in determining whether the company surpasses, or gets surpassed by, competitors. Yet for too many business executives, decision making is an art that comes with experience. And experience is what you get when you don't get what you expected. These decision makers find themselves trying to make decisions based on incomplete, inaccurate, irrelevant or stale information.

Decision making from intuition or just gut feel lacks replication, predictability and scale of successful outcomes—and decision makers seek methods and processes to leverage data-driven, fact-based decision making.

Excel normally emerges as a stop gap measure on the road to business intelligence. However, the inherent flexibility of spreadsheets is a double edged sword—enabling data massaging at will—but also leading to lack of data integrity with a simple transposition error or when a sum formula misses a row, or eventually multiple versions of the truth when team members working on a common project argue over who's data is 'more accurate'. Questionable reliability ultimately leads users to distrust the data and seek alternate solutions. Spreadsheets are powerful tools which should compliment decision making solutions, but put decision making at risk when pretending to be the system of record.

Decision makers recognize their need to graduate from spreadsheets and leverage a more strategic technology tool in order to access more data, more timely and with greater integrity. Executives, managers and other decision makers also seek automation solutions which allow them to spend less time retrieving and compiling historical information and more time analyzing information that supports their most pressing business initiatives, allows them to better plan for the future, quickly identifies areas that need attention and systemically delivers the insight to contribute to improved decisions.

The Business Solution

Organizations normally possess plenty of valuable data, albeit in different repositories that resemble silos. Prospect data in the sales force automation (SFA) system, lead data in the lead management system, service history in the call center application, customer data in the customer relationship management (CRM) system, product and sales data in the enterprise resource planning (ERP) system, sales performance data in the incentive compensation system, business plan and budget data in spreadsheets and so on. Only when the data is consolidated can relationships, patterns and otherwise difficult to identify insight be discovered.

To better execute business strategies and outperform competitors, business leaders are pursuing a combination of systemic processes and decision support software tools which synergize to better source, aggregate, contextualize and deliver business insight to knowledge workers, operational managers and decision makers throughout the enterprise.

A Recommended Course of Action

Consider the following strategic approach to achieve a predictable outcome:

1. Verify Pressure For Change

Successful BI deployments are driven by a pressure to change. Only when the opportunity associated with better decisions or the pain of poor decisions is clearly recognized and exceeds the collective effort of cash outlay, dedicating resources and business interruption will enough pressure exist to go the distance with a BI solution. Attempting to implement a BI solution, however legitimate the intention, when decision makers don't recognize a decision making dilemma is an uphill fight that will be met with change management resistance such as waning sponsorship, IT reluctance and user adoption challenges. If you find yourself in this position, begin with an education agenda and see if you gain momentum.

2. Quantify Stakeholder Objectives

In determining stakeholder objectives it's necessary to first define your stakeholders. As with all enterprise software deployments, visible and vocal executive sponsorship is a must, so soliciting executive team expectations early is paramount. Beyond that, if you're beginning with a departmental or line of business project, your stakeholders will likely include those business unit directors or managerial staff as well as line managers and support staff who are held to performance standards which materially affect the departments measurable objectives. Each of these roles holds relevant first hand information, is key to a successful deployment and must be afforded the opportunity to identify their objectives. Not all objectives may make project scope, but all should be surfaced, heard and considered.

3. Dedicate Resources

While executives and decision makers seek business intelligence solutions which can be configured, tailored and exercised without much or any IT involvement, it's a mistake to select or attempt to implement a BI product without IT participation. Even with Software as a Service (SaaS) BI solutions, which are easily provisioned on-demand and more quickly deployed, BI requires data cleansing, data staging and transfer, and integrating and consolidating data from a number of usually disparate information systems and technologies. These tasks will certainly benefit from technical talent on the IT team.

Similarly, BI solutions must align with business objectives, focus on business requirements and deliver business and operational insight. Therefore, BI deployments should not be IT driven. Business intelligence is best accomplished by leveraging the highest and best skills from both IT and business staff, and ultimately result in a symbiotic relationship as each side is dependent upon the other for success.

BI tools are not just for top executives or a privileged few. The goal is to connect as many operational decision makers with operational data as possible. Wider participation leads to better decisions at more levels in the organization, increased operational alignment with the company's top strategic goals and a culture of learning. Consider naming a cross-functional team or committee to define the information strategy and decision support roadmap.

4. Determine The Most Salient Metrics

Identifying the key performance indicators (KPIs) which are most influential in advancing business objectives is a pre-requisite to a successful BI program. KPIs must be aligned with the business imperatives that they are benchmarking. It's critical to make sure you measure KPIs that really drive performance, and not measures which are easily retrievable or traditionally reported.

Identifying the right performance metrics is not a onetime event. As business plans, budgets and management directives change so too will the performance metrics which align to those objectives. Too stay abreast with changing business conditions, new opportunities and competitive threats, business leaders must implement a process of continuous review to reaffirm, adjust or replace metrics and see to it that the metrics measured are at all times optimal for achieving the missions for which they are aligned. It's during this review process that many managers identify new business drivers for the first time.

Performance metrics are unique and highly dependent upon individual business goals, however, several performance indicators are common across industries. Marketing metrics may include campaign performance indicators such as response rates, conversion rates and ROI while sales may align performance with pipeline quality, forecast accuracy and win rates. Customer service is sure to measure strategic metrics such as customer satisfaction and retention as well as several operational metrics such as first call resolution (FCR) rates and up-sell conversions. It's key is begin with fewer, more relevant metrics as opposed to initially measuring everything that can be measured.

5. Identify Data Sources

After identifying what data is needed, create an inventory of data sources. For CRM analytics, the bulk of data is likely to reside in the Customer Relationship Management system. However, additional data stores are likely, and may include a marketing automation system, email and groupware systems, a Content Management System (CMS), an accounting software system, an Enterprise Resource Planning (ERP) application, shadow systems or even Excel spreadsheets.

While internal transaction-based applications hold a wealth of structured data, it's the unstructured data residing in external social media and other repositories that presents both opportunity and challenge. Unstructured social content residing on social networks, blogs, rating and review sites, and online communities, to name a few, possess valuable contributory data that can be appended to existing information to add further perspective and insight for decision makers.

Fortunately, best of breed CRM analytics and traditional BI platform suppliers have stepped up to address this opportunity—although the methods and objectives vary significantly by vendor.

6. Select Your Tools

Software technology enablers are necessary to transform raw data into business insight. A CRM analytics or business intelligence platform generally includes an underlying data warehouse or data marts which store, aggregate and correlate data; integration or middleware tools which extract, transform and load (ETL) data across disparate sources; and data visualization technologies such as dashboards, scorecards and OLAP (online analytical processing) which permit query, reporting and interactive analysis.

Consolidating data into a central system is an analytics best practice in order to achieve a system of record and avoid inherent problems associated with disparate and redundant data, including potentially conflicting data or multiple versions of the truth. A central data warehouse, or at least a small number of domain specific data marts, will ease data inquiry, extraction, reporting and deliverability as well as system administration.

Information doesn't need to be delivered in real-time as much as it does in right-time. Real-time may be necessary in order to capitalize on short term opportunities, however, less frequent periodic data refreshes may be fine for monitoring more strategic efforts.

Business intelligence solutions are available in a variety of software technologies and deployment options.

Consider Software as a Service (SaaS) BI when … IT resources are at a minimum, budgets are tight, capital expenditures are unrealistic or you seek a proof of concept to validate the solution. SaaS analytics solutions excel where business requirements are straight forward, data sources are relatively contained and management wants to go deep within particular domain areas. SaaS analytics are extremely popular for functions such as website analytics, text mining, speech analytics and pipeline analysis. Sample SaaS CRM analytics vendor solutions include Angoss, Birst, Cloud9, Convergys, Coremetrics, GoodData, Oco, Pegasystems/Chordiant and PivotLink.

Lead with traditional, on-premise BI software when … You have excess capacity with both IT computing and staffing, there is a cultural preference for on-site computing or you seek to an enterprise-wide solution to integrate with many heterogeneous information systems. Market consolidation has resulted in the largest vendors accounting for the bulk of the market. However, resilient pure play vendors continue to emerge and gain traction by illustrating unique value propositions, accelerated and simplified deployments, superior time to value and reduced costs. A few popular enterprise CRM analytics systems, available in multiple deployment models, include three solutions from IBM – Cognos, SPSS and Unica; two solutions from SAP – Business Objects and Crystal Reports; as well as solutions from arcplan, Board International, Infor Epiphany, Information Builders, Microsoft, MicroStrategy, Oracle, Panorama Software, Portrait Software, QlikTech, SAS, Tableau, Targit and Tibco.

Evaluate open source BI software when … Your BI needs are likely to require customization and you seek control in modifying the tools. This option can also be attractive to ISVs (independent software vendors) seeking to embed a BI solution in their software products. Open source BI is experiencing its highest adoption in certain industries such as government and certain regions such as Eastern Europe, Southeast Asia and Latin America. Open source CRM analytics solutions are available from vendors such as Actuate, BEE, Cignex, Jaspersoft, Openi, Pentaho, SpagoBI and Talend.

BI software solutions vary greatly in terms of scope, target market, delivery model and value proposition. A proper software selection project is your best assurance to make sure you acquire the analytics system that best matches to your business requirements and objectives.

7. Clean Your Data

We've all heard it – GIGO – Garbage In Garbage Out. Nowhere is this more absolute than with BI projects. Dirty data is the top cause of BI project delays. Few organizations anticipate the data quality issues, and subsequent data cleaning requirements, before they initiate the data population phase. To avoid this repeated mistake, sample each of your data sources early in order to determine data quality and allow for the needed time to scrub the data before its imported to the data warehouse.

Once you begin seeding the data warehouse with clean data, don't stop there. Look for tools which can automate clean data acquisition and ongoing maintenance. For example, CRM systems can use tools such as account merging, address verification, spell-check and de-duplication. Getting data right at the source speeds data consolidation, reduces data maintenance and lowers costs.

Once into production, you will need a formal process supplemented by enabling software tools to both clean and enrich the data on a go forward basis.

Never lose sight that quality data is a pre-requisite for quality decisions; or said another way recognize that BI solutions with faulty data aid users in making poor decisions more quickly and confidently.

8. Pursue a Phased Approach

Business intelligence and CRM analytics programs are akin to other enterprise software applications in that well-defined, multiple-phase projects lower risk as compared to big bang or waterfall deployments. This approach also allows early lessons learned to shape future roll-outs, permits project managers to publicize (even small) victories and facilitates user adoption.

Start with a relatively small department or business unit, possibly one with a never ending backlog of report requests. After achieving success, methodically expand your roll-outs to include more business units and integration with more information systems.

Also recognize that seldom do performance variables and factors for even departmental objectives reside entirely within a department. Instead, process flows traverse departments and more often than not data resides in often redundant operational silos. Even with a goal to start small and advance in phases, the project should commence with the end in mind. Decision makers in any particular area must understand the enterprise-wide impact of their decisions.

9. Measure and Refine

Measurement should begin by actively tracking staff utilization of the BI tools. User adoption will grow over time so tracking who's accessing the tools and the volume of users over time will provide an early indicator toward ROI.

Information analysis reveals learning and insight, however, it also raises more questions than answers, thereby requiring extending the data models for new interrogation of the data, inserting new measures or dimensions to discover new relationships, displaying the data differently for various roles and contextualizing the data to make it more actionable.

Business strategies, business unit operational goals and functional targets are fluid so the metrics and BI solutions which support them must continually advance to remain relevant.

Top business achievers maintain a deep understanding and direct link between their business objectives and the operational performance required to achieve those goals. Businesses advance their growth objectives when the metrics that drive those objectives forward are measured and improved as needed. Many companies recognize this connection, however, the process of aligning, religiously measuring and quickly implementing corrective actions remains elusive. For sustained success companies should implement a formal process whereby performance metrics are faithfully measured and learned from—and modified or adapted based on that learning or as the business shifts.

Measurement also includes calculating the BI project return on investment (ROI). No enterprise software project is complete without periodic evaluation, determining whether slated goals have been accomplished and calculating payback. To this end, information management costs and benefits need to have been calculated ahead of the BI project in order to have a baseline comparison point. BI ROI should be determined periodically as additional phases are completed, information grows and adoption rates increase.

10. Raise The Bar

Making better business decisions throughout the enterprise is a perpetual journey. Some additional follow-on initiatives may include:

  • Grant information dissemination to more people, or increase cross-functional information exchange among more lines of business. The goal is to democratize BI analysis throughout the business.
  • Improve the speed or timeliness of information delivery. However, not all information needs to be real-time, so align information types with the need for speed.
  • Implement alert notifications. These notifications are normally delivered in real-time upon a performance metric threshold value being exceeded—and can provide management the opportunity to remedy a performance deviation before it gets out of hand.
  • Tailor data visualization by roles, and experiment with new information delivery tools, be it new forms of dashboards or tools which permit new methods of slicing and dicing data, possibly by dragging-and-dropping data results with new measures.
  • Step up to predictive modeling and analysis. This type of analysis uses technology to discover hidden patterns and support 'what-if' scenarios or pro forma modeling. Being able to accurately forecast the effects of new or proposed efforts is a powerful tool in allocating budget and scarce resources among competing alternatives.
  • Append more unstructured and social media data with existing repositories. Social CRM tools are increasingly acquiring more unstructured customer data. However, for most companies that data remains isolated from their CRM software records.
  • Moving in the other direction, new social CRM tools and mashups are permitting users to develop analytical scorecards, dashboards, charts and graphs and distribute them to social sites such as blogs, wikis or even Facebook.
  • Compliment your CRM data, which is largely historical and a lagging indicator of customer behavior, with more current and dynamic data generated from customer surveys, customer loyalty programs and voice of the customer programs. Data from these measures can be used as leading indicators and is generally more telling of the true customer relationship.
  • Leverage mobility to make insight portable and device agnostic.
  • Implement training and coaching programs which aid uses in reviewing, interrogating and acting upon the data.
  • Consider implementing a knowledge management solution in order to centralize and better share policies, training and education materials, instructional procedures, advanced capabilities and best practices.
  • Check out self service BI solutions, which put powerful but easy to use, browser-based query and reporting tools, interactive graphics and wizard driven creation into the hands of staff. This can be an ideal method to keep up with users growing information demands or relieve under-staffed IT resources.

Conclusion

Information is the fuel that power's intelligent organizations. However, management's traditional use of reviewing static, generic, historical reports which usually only pull data from a single source is no longer sufficient in competitive markets. These reports are designed for passive viewing and do little to proactively advance business initiatives in the shortest cycles possible. What's needed are the strategies and software tools to help decision makers advance from looking in the rear view mirror to instead looking forward through the windshield to see what's ahead.

BI solutions are increasingly popular, and beginning to be embedded with traditional and on-demand Customer Relationship Management software solutions. These tools offer increasing value when extended throughout the business. Strategic, data driven projects such as annual business plans, budgets, forecasts, period-end reporting packages, incentive compensation plans, product price elasticity exercises, workforce analysis, regulatory compliance and more can strongly benefit from automation and analysis with BI solutions.

Business intelligence tools have historically correlated and displayed data as a means to view and analyze what happened by interrogating the data with various measures and dimensions. However, with more powerful BI tools which deliver increased data mining and predictive analytics, more types and volumes of data to increase confidence levels and enhance increased learning, and with new technologies such as cloud and SaaS to simplify and accelerate BI deployments—analytics are no longer just for enterprise companies with deep IT resources.

Executives, managers and staff throughout the organization are responsible to the make the best decisions they can based on the information they have. If the timing, relevancy and insight from that information improves, so will their decisions.

When pursued strategically, CRM analytics and Business Intelligence solutions apply more relevant information to decisions, involve more contributors in decision making processes and reach better decisions in less time. If you improve the quality of your decision making processes, you will improve your execution of business objectives.