Data
Get information on our data sources, forecasting methods, accuracy, revenue definitions, B2B data availability, and our in-scope and out-of-scope definitions.
ECDB uses different data sources as a basis for revenue, market and KPI calculations. Transaction data from credit cards, debit cards, bank accounts & e-wallets are the most important source. All of the transaction data comes directly from card issuers or processors based on the transactions going over their networks.
ECDB processes more than 1 billion online & offline transactions per month. This represents approximately 2-3 % of all transactions for the analyzed companies and markets.
Almost real-time: ECDB receives the transaction data with a delay of 7 days after the purchase event and processes it immediately.
ECDB complements the transaction data with other information: company data from public companies, traffic & clickstream data, web scraping and crawling.
Transaction data
- Credit cards, debit cards, bank accounts and e-wallets.
- Transaction information: description, merchant name, industry sector, purchase amount, date of purchase.
- Daily data processing.
- Fully compliant with global privacy regulations.
Other data sources
- Company data from public companies: approx. 5% of relevant retailers are stock listed.
- Traffic & clickstream data.
- Web scraping & crawling.
ECDB buys its data from different international suppliers. Some of them generate the data themselves, some of them act as wholesalers. All the data is GDPR conform.
Data modeling by our analysts is the heart of our data generation. Based on researched, collected, harmonized and cleaned data, the different datasets are compiled into a consistent data model.
Our experienced data analysts dive deep into individual markets, countries and players and merge all data strings. With the help of various in-house created tools, we achieve data modeling that produces unique data as its output. As the data we publish are an estimation, we constantly check our estimation with published data from stock-listed companies. This helps to continually improve our estimation model.
In our forecasts, we apply diverse forecasting techniques. The selection of forecasting techniques is based on the behavior of the relevant market. For example, the S-curve function is well suited for forecasting digital products and services due to the non-linear growth of technology adoption.
For markets with projected steady growth, we use exponential trend smoothing to illustrate the continuous market development. The parameters are adjusted individually depending on the market-country/territory combination.
As these are just some examples, more details and information on the main sources can be found in the methodology box on the content page of the respective market.
Approaches, assumptions, input data, and scope are improved from update to update. Therefore, data from previous updates might not necessarily be comparable with current data.
“Global revenue” refers to net sales eCommerce of a specific store. “Filtered revenue” refers to a share of the revenue which can be allocated to a specific product category and/or country. If the product category selection is “all categories” and the country selection is “worldwide” then “filtered revenue” equals “global revenue”.
For the moment ECDB does not include B2B data. The modelling approach to B2C and B2B are very different from each other. It is planned to include B2B data in the next years.
Our product categories cover 250 different markets. The categories include all relevant B2C physical eCommerce categories. A detailed overview can be found here. Not included are digital ecommerce, traveling or B2B eCommerce.