ICD User Guide
ICD User Guide |
|
CHAPTER 5 RESEARCH METHODOLOGY
Creating an initial data model
Revising the initial data model
Creating demographic value splits
Datamonitor has worked with independent auditors to develop a research methodology that provides clients with the most accurate data possible, and ensures that all data points we publish are fully traceable. A leading global audit firm audits our research processes, to ensure that all our researchers and analysts fully follow the methodology that we have agreed and published.
For more detailed methodology information, please follow this link (LINK TO CLIENT METHODOLOGY PDF).
Datamonitor's consumer goods market data is gathered in three key stages over the course of each year:
The most important research process is the annual data review, which takes place at the beginning of each year. Researchers review every single data point within Datamonitor's consumer markets databases; they carry out over 5,000 primary interviews and review more than 20,000 secondary sources to ensure that that all the information Datamonitor provides is accurate and timely.
The annual review is a complex and detailed process, involving research, analysis and market modelling. It is broken up into four key stages:
During the initial stage of the annual review, the secondary research team carries out detailed research based on published external sources. The sources checked include:
The information collected not only highlights the key developments in the market over the previous year, but also helps confirm that the previous year's figures are accurate.
In addition to the above Datamonitor also purchases primary data on a number of core markets concerning category segmentation by demographic group. These data are subject to Datamonitor’s usual standards of due diligence, ensuring that it meets our own research criteria.
In preparation for the final interview stage, analysts prepare a detailed estimate of all data points covering a particular market. Once finalized, this estimate is circulated to primary research contacts. The market modelling process is based on a wide range of sources:
The modelling process initially involves preparing an estimate of historic market size; this will be generally based on the previous year’s data. If Datamonitor has carried additional research on the relevant market since the last data review process, this research is used as a base. If the data has been revised in line with client concerns since the last data review process, the revised figures are used as a base.
The analyst preparing the market model for any country/market combination then produces a trended forecast for the current year onwards, based on:
Once the basic forecasting is complete, the researcher applies the results of the secondary research process to their initial model. The key aim in this process is to ensure that the effects of company-level market drivers are incorporated in our data findings; additionally, it allows Datamonitor to take into account other organizations' estimates of market size in determining the accuracy of our own estimates.
Once the analyst has taken into account the impact of the secondary research, they produce a provisional data estimate for the relevant country and market. When this process is complete, a full sanity checking process is then carried out.
In order to estimate value splits by demographic group for our ConsumerGraphics data, we carry out a second, parallel modeling process. This is also broken up into two distinct phases:
The end result is a set of data segmenting each category modeled by demographic group of consumer; this is comprehensively sanity-checked before being combined with the modeled market value data and converted into a questionnaire.
The primary research process is a major component in Datamonitor’s data research program, ensuring that all data contains input from people active within the relevant market in the relevant country. Datamonitor’s researchers follow a tightly defined primary research process
The primary research interview process is based around Datamonitor's database of more than 15,000 food, drink and personal care industry contacts. The research team contacts all the contacts in Datamonitor's database that are relevant to the program.
Once all the interviews for a particular market have been carried out, analysts reconcile these with each other, the original data, and all other sources originally used.
Where there are disagreements between sources, the data finalization process is necessarily reliant on analysts' judgement at evaluating and weighting conflicting values. The methodology involves:
If the analysis team is unable to resolve final market figures, then the research team is contacted to carry out further primary interviews.
Once the data consolidation process is complete for a particular market, finalized data sets covering that market across all countries covered are compiled from the individual datasheets.
The analyst compiling these market data sets runs a quantitative checking program designed to draw out any possible discrepancies and to allow for ‘sanity-checking’.
Throughout the course of the year we carry out additional research, including extra interviews in countries with relatively low confidence scores, customer surveys and responding to customer feedback, Internet surveys, consultancy projects and work for other reports. All these sources are fed back into the following year’s Annual Data review.
© Datamonitor 2004