Hot Topics
- Home
- /
- Campaign Management Solution Overview
- /
- Predictable data consolidation vital for on-time campaign execution
Figure 1 – illustrates the lack of predictability in the start time of campaign operations. On day 1 in the above diagram, campaign operations begin on schedule at 6AM. However, on day 2, the start of campaign operations is delayed by 4 hours. Late availability of data required for campaigns was the reason for the delayed start on Day 2. Such delays although not caused by campaign operations lead to loss of business opportunities and lower productivity.
Predictable data consolidation vital for on-time campaign execution
Introduction
Effective campaign operations require an infrastructure for consolidating data from different sources. Typically, sources have different timings at which they provide data for campaign infrastructure.
This data consolidation process plays an important part in ensuring timely execution of campaigns. Most organizations face difficulties in ensuring that consolidated data is available for campaign execution in a predictable manner. In order to devise a solution for predictable execution of the consolidation process, we need to understand the challenges involved in data extraction and consolidation for campaign operations.
What is the difficulty in consolidating data?
Effective campaign operation requires consistent customer and campaign data from discrete source systems. Customers interact with different source systems for services. Since customer interactions with these systems may vary in time and sequence, any single system is likely to hold only fragmented view of the customer. This fragmented data needs to be synchronized from different systems prior to running campaigns to get a consolidated and consistent view of customer data and avoid loss of business opportunities.
Building agile and efficient marketing teams
Ad-hoc approaches in implementing online campaigns are impacting business. It is taking longer than ever to implement new campaigns and operation costs are on an increase.
Resolving these synchronization issues is a challenge because each source system provides data with differences in terms of start time, volume, availability and mechanism to handle dirty data at source. Jobs implemented to extract data from source systems are designed using static scheduling rules and do not account for constantly changing patterns of source system availability, data volumes or the need to properly sequence data extraction from different systems.
Consider the example of source systems handling IVR transactions as shown in Fig 1 above. Source systems A, B and C consolidate data on customer transactions at end of business day and make them available for extraction by the campaign data infrastructure. On Day 1, Data consolidation at source system A and B ends at 6PM, 7PM for source system C. The job for extraction of source data from A and B is scheduled to start at 6 PM and for C at 8 PM. The process proceeds smoothly because all three source systems were able to consolidate their data before the scheduled start of their respective extraction jobs.
Measuring effectiveness of campaigns
Campaign teams adopt ad-hoc approach to monitor status and measure campaign effectiveness. Standardizing the process helps to eliminate manual collation of status information.
However since the volume of customer transactions varies from day to day the time required by IVR system to complete consolidation and make the transactions available for extraction keeps changing. On Day 2, data consolidation at source system A, B and C overshoot the static scheduled extraction time of the extraction jobs. The extract jobs started at their scheduled time but are able to extract only the subset of data that was consolidated till time of their start. This situation requires execution of data backlog jobs, to extract the backlogged data and consolidate with already extracted data, which are time consuming and resource intensive. The end result is delays in data consolidation and finally campaign operations.
The above challenges are particularly acute in enterprises with large transaction volumes and an aggressive campaign marketing operation.
Visualize campaign data to gain benefit
Critical aspect is to provide data to campaign managers in a way that makes sense to them. However fast changing business needs are exposing inefficiencies in the way most teams implement this aspect.
Dynamic scheduling of extraction jobs
The challenge of synchronization of fragmented data requires implementation a well-designed ETL [Extract, Transform & Load] layer. Extraction logic implemented at ETL layer extracts data from source systems and loads the data into a staging database.
Such extraction jobs should be designed in a manner that takes into account dynamic nature of availability of data at the source systems. The scheduling of an extract job changes based on information received from source systems such as expected data volume, its time of availability and business rules that determine how the data is extracted. Extraction jobs should include mechanisms to handle their restart, rollback of committed data either by “truncate and load” or “delta load, and activation or deactivation of any job”.
Using above approach ensures data extraction starts dynamically and reduces the necessity for executing data backlog processes.
Lets understand how this dynamic scheduling helps to reduce the possibility of execution of data backlog jobs explained in the example above. Extraction jobs for source A, B and C is to be scheduled based on the information shared by the source systems. This information includes the data volume and time of data availability, which is 9:00 PM for A & C. This information is shared after data consolidation starts at source A, B & C. The extraction jobs are then re-scheduled based on the expected timings of data availability. This dynamic scheduling avoids the execution of data backlog jobs. Even if sources could not finish consolidation at 9:00 PM data backlog job will have to process lesser delta and thus increasing the possibility of starting campaign operations on time.
Data from staging area is consolidated in the pre-presentation layer using ETL jobs thereby completing the data in terms of time, customer transactions and other customer information. These jobs should also be scheduled, synchronized, monitored, activated and deactivated according to business understanding of the data.
Managing marketing campaigns
Marketers face challenges in operating their campaign smoothly despite having multiple systems. They need "ready to execute" infrastructure and analytical approach to take charge of the campaigns.
Improving marketing campaigns
EXILANT helped one of the largest consumer electronics product company to establish and enhance their campaign system. The solution enabled the client to manage and improve sales and services of their products.
Summary
By implementing well-designed ETL jobs, the extraction of fragmented data from unsynchronized source systems for consolidation can be improved. This ensures reduction in delays in executing campaigns that target rapidly evolving business opportunities. With our expertise in implementing the campaign management systems, we can assist clients in their initiatives to improve performance of their campaign data infrastructure.
Recent Links :
Building agile and efficient marketing teams
Ad-hoc approaches in implementing online campaigns are impacting business. It is taking longer than ever to implement new campaigns and operation costs are on an increase.
Measuring effectiveness of campaigns
Campaign teams adopt ad-hoc approach to monitor status and measure campaign effectiveness. Standardizing the process helps to eliminate manual collation of status information.
Visualize campaign data to gain benefit
Critical aspect is to provide data to campaign managers in a way that makes sense to them. However fast changing business needs are exposing inefficiencies in the way most teams implement this aspect.
