ESOMAR 37

DatAction has its own proprietary panel community. These carefully developed and maintained panels are the cornerstone of our data collection capabilities. We recruit panelists through various methods, including online media campaigns and social media.

We also have contracts with other local and global suppliers. If our own panel resources are insufficient due to specific requirements or project design, we can utilize our partners' panels. Any such additions are agreed upon with clients in advance.

DatAction's proprietary panels serve as our primary source of high-quality data, ensuring consistency and reliability for our clients. Approximately 93% of our cases are conducted using these panels, while the remaining 7% utilize our partners' panels. We apply a rigorous selection process to ensure that the quality of our partners' data meets our standards and provides accurate insights.

We strive to utilize a wide range of recruitment sources to ensure the best audience representation. These sources include social networks (both local and country-specific), contextual and search advertising, postings in local communities and media, email campaigns, advertising in messaging apps, digital and mobile advertising, and all other available channels. While our recruitment methods are generally consistent across all geographic markets, the proportions may vary based on specific market characteristics, aiming to cover both local and global media.

Our panels are non-probabilistic and welcome participants from all socio-demographic backgrounds in line with ESOMAR guidelines and national laws. We do not reject registrations based on specific demographic criteria during the registration process.

Additionally, we run a referral program that allows our existing panelists to invite their friends to join our online panels. Users acquired through referrals constitute about 15% of our panelists, varying by country.

We have three major groups of sample quality checks and balances:

1. Recruiting/registration check When someone is making a new registration, the panel management system checks if the registration email is already assigned to any existing account, and notify if the email is already used. Then the panel management system checks if the new email’s domain is included in ‘disposable email’ domains or deceitful domains (services that provide an email address for several minutes or hours just for registration purposes) blacklist. We do not allow registration with temporary emails. To proceed with the registration process a confirmation email is sent for double opt-in. When accepting the confirmation email, for the next step, the system is using 3rd party software to create a digital fingerprint. It contains a set of data that identifies browser setup and device as unique to the new panelist comparing it with all existing ones. In case similar snapshots are found the system puts a ‘do not contact’ flag for a new registration account and it will be excluded for all statistics and invitation sendouts in the future.
   
   
2. Pre-survey check DatAction has a panelists’ ‘trust rating’ for all online panels. All new panelists have 100 points ‘trust rating’ as we do believe that all of them are honest and considerate respondents. In case if the panelist is being disqualified from the survey, his rating is being decreased for 10-30 points depending on the disqualification reason (fraudulent answers in open ended questions, speeding, straight-lining, etc). An out-of-range rating results in the participant being blocked from our sample.
   
   
3. During-survey check We use cookies and captcha verification systems to avoid fraudulent attempts.
   
   
4. Post-survey check DatAction always asks clients to provide the list of accepted and fraudulent IDs with reasons for deletion. The system sends out information emails for the removed respondents to clarify the reason for incentive removal and survey completion guidelines.
   
   

About 90% of our panelists access surveys through our web portal and receive invitations by email. These channels play a vital role in engaging panelists. The remaining 10% of our panelists chose to receive survey invitations through wide used messengers. The proportion varies at a panelist demographic, country, and project level, but overall, we see 40% of traffic coming from email notifications, 50% from our website, and 10% from the messengers. We use fully localized domains for online panels in each country, so if you are interested in receiving a list of our panel domains, please, send us an inquiry to rfq@dataction.eu

Ninety-three percent of all the research we conduct is based on our proprietary panels. Depending on the study design and project specifics, we may supplement with data from our partners. All additional sources are carefully preselected to ensure data quality, and their usage is pre-agreed with the client.

We have a system in place to allocate quotas per partner and control sample distribution. To avoid respondent duplication, we use cookies and unique links.

We supply sources for online questionnaires only. Samples can be provided for shorter or longer questionnaires, for mobile or desktop only. We can perform longitudinal studies and provide access to the same respondents or exclude them for consecutive waves of the research. We do not conduct online groups or communities.

Once a client commissions a sampling project with us, our first step is to identify the specific target population they wish to survey. We carefully match panelists to the client's target criteria and distribute them according to pre-set quotas to ensure balanced representation of all segments within the target group.

After selecting the appropriate panelists, we send them survey invitations via email and social media notifications. We strategically time and size these invitations, taking into account factors such as response and drop-out rates, to ensure we meet the required number of completed survey responses within the allotted fieldwork period.

We collect and store detailed information about our panelists, covering more than 200 parameters to achieve precise targeting. These parameters include:

• Demographics: Region, gender, age, marital status, household details, income and family finances, education, profession, and occupation.
  
  
• Consumption: Household items owned, shopping behavior, and categories of goods and services consumed.
  
  
• Travel and Leisure: Traveling habits and leisure activities.
  
  
• Health: Health-related details.

Detailed profiling allows us to effectively match panelists with research projects, enhancing the quality and relevance of the data collected. Some parameters are localized based on the country of residence and its specifics, such as broadband and mobile providers, banks, and income strata. This localization helps achieve better precision in targeting and accommodates regional differences in geographically extensive studies.

We can supply some of the points as addends in the data set on request.

To provide a feasibility estimate, we require the following minimum set of parameters: number of completes, quota structure, if applicable, and length of interview (LOI). For more precise estimations, especially when targeting more specific audiences, additional information such as qualifying screening criteria and estimated incidence rates provided by the client is helpful.

In cases of complex study designs or tracking studies, we need to be informed about additional parameters, such as exclusivity periods and the number of waves

We aim to inform the client in advance if the sample is insufficient and there is a risk of not completing the fields. This is monitored by the project manager and a built-in warning system in the program. In such cases, after aligning the approach with the client, we can:

• Add our pre-authorized and verified suppliers to obtain a sufficient sample
  
  
• Use the client’s preferred supplier
  
  
• Consider changing some of the qualifying criteria
  
  

When working with our partners, we ensure the quality of the sample and mitigate duplications. Based on ISO standards, we exercise strict guidelines to select partners. Judging criteria include, but are not limited to, data quality, sample feasibility, pricing, timeliness, reliability and safety.

We do not use the router. However, for each quota, we create a sample and forecast potential completions. We then activate an automated distribution system to achieve the required number of interviews. Randomization ensures that everyone has an equal chance of receiving the survey, with priority given to those with high ratings. Invitations to do a survey are sent out via email and social media messanger.

We also adhere to tracking limitations by collecting subsequent waves according to the specified requirements. Additionally, we observe moratoriums on excluding participants from surveys within the same category for the recent period.

Data quality is our highest priority, so we avoid overloading respondents and only assign them to surveys for which they are perfectly suited.

Instead of using a pre-screener, we simply notify respondents that they qualify for a survey based on their profile. They are then granted direct access to the client's survey link.

We use the same approach whether panelists access the survey through their personal account or a unique link. To ensure panelists are prepared for the survey, we provide them with the following information: length of interview (LOI), reward, type of device required (mobile or desktop), and any applicable technical requirements (e.g., longitudinal study over several days, access to a camera). To maintain the integrity of the responses, we do not provide any information that could be used to pass the screener or give hints about the category or brand.

Survey participation can occur through either a general format direct invitation or via the respondent's personal account where only the LOI, reward details, and technical requirements are stated.

While we can adjust incentives dynamically during fieldwork and this adjustment is reflected at the individual level, we refrain from using higher incentives to encourage respondent participation in surveys to avoid potential long-term behavioral bias.

As part of our regular procedure, participants who successfully complete a survey are invited to provide feedback on their overall experience. They rate their experience on a scale of 1 to 5, which helps us assess their satisfaction with the survey process. Moreover, we offer an open-ended question for participants to share additional comments or feedback if they choose. This information can be shared with the client upon request.

We provide a comprehensive debrief report including incidence rates, respondent completion statuses, quotas, and metrics influencing the report's cost. Additional metrics can be included in the debrief report upon client agreement.

By default, each panel member can receive a maximum of three survey invitations per week. Additionally, upon client request, we can implement quarantine or elimination rules that may pertain to specific time frames, survey topics, or individual projects. For instance, in tracking surveys, it is often necessary to eliminate completed interviews from the past three or six months, a capability we fully support as needed.

We can share certain data with clients, but each case is discussed individually. We ensure that any shared information complies with GDPR guidelines and is not protected by non-disclosure agreements (NDAs).

At Dataction, we implement three key controls to verify the authenticity and identity of panelists, ensuring high-quality data collection:

1. Duplicate Control: We use a fingerprinting system and unique cookies to identify and reject duplicate responses, guaranteeing that each panelist provides unique data.
   
   
2. Captcha: Panelists must complete a captcha verification to distinguish human participants from bots, ensuring genuine responses.
   
   
3. Sex/Age Verification: We include questions about sex and age in all surveys and compare responses with recruitment data to validate demographic information.
   
   

These measures work together to authenticate panelists, prevent duplicate entries, and verify essential demographics, ensuring reliable and credible data for our clients.

We can monitor each source within our system, managing partner contributions and allocating and controlling sample structure.

We can share this data with clients on request upon project completion.

We have implemented a panelist ranking system based on their behavior within the panel. Each panelist starts with 100 points upon joining. The system monitors performance, deducting points for non-compliant behavior such as low-quality answers, neglecting open-ended questions, or inconsistencies with the speed checker. Panelists with the highest rankings are given priority when assembling survey samples.

Additionally, we promote secure methods like S2S (Server-to-Server) and Hashing to ensure end-to-end security.

To maintain high-quality data, we implement several standard quality checks in our survey templates to ensure participant engagement and accuracy:

1. Automated Checks: Detect illogical, inconsistent, and non-responses.
   
   
2. Trap Questions: Verify attentiveness by including questions with only one correct answer, screening out incorrect responses.
   
   
3. Minimum Survey Length: Set a time threshold for survey completion; responses below this threshold are excluded.
   
   
4. Age and Gender Validation: Cross-check participant-provided data with our records for consistency.
   
   
5. Mandatory Questions: Require responses to all questions for comprehensive data.
   
   

Additionally, we offer customizable checks based on client-designed questionnaires:

1. Real-Time Open-Ended Checks: Ensure quality open-ended responses.
   
   
2. Logic Checks: Identify inconsistencies between related questions.
   
   
3. Grid Quality Checks: Detect suspicious patterns in grid-style questions.
   
   

These checks ensure data integrity and provide high-quality insights for our clients' research projects. As the result clients get clean, ready-to-use data.

Can be provided upon the request. Is inline with the GDRP policies and local legislation.

1. We adhere to the EU General Data Protection Regulation (GDPR), which represents one of the most stringent legal standards for data protection and to Estonian internal regulations.
   
   
2. We are using consent as the legal basis, and we while obtaining the consent we ensure that the process in compliant with GDPR. We store the proof of received consent.
   
   
3. For the cross-border transfer, data retention and addressing data-breach we use internal policies developed and maintained in compliance to GDPR legislation.
   
   
4. We have appointed DPO.

All policies are available to participants on the panel's website. Upon joining, participants confirm that they have read the policies and consent to the collection, storage, and processing of their data. Panelists can withdraw their consent at any time through their personal account on the website or by contacting the online helpdesk for assistance.

We adhere to the laws of the country where our services are provided and where payments are made to data subjects. Any changes to relevant laws and regulations are promptly followed and implemented by our legal and accounting departments.

Currently we are not formally certified by the external agency, we can confirm that all our internal policies are inline with the GDRP and the legislations of the counties, where samples are gathered.

Yes, we implement “data protection by design” as a ground principle and therefore have implemented and running data-protection policy that is fully adhere to the GDPR and Esomar guidelines and is compliant with the legislation of the panelists country of residence.

We adhere to ISO 27001 standards, conducting regular audits and risk-assessments as well as stress tests on our internal information system.

We are not ISO 20252 certified by we hereby confirm that we adhere to framework and quality standards.

We can provide all the following information to our clients upon the request.

+1. Average qualifying or completion rate, trended by month
+2. Percent of paid completes rejected per month/project, trended by month
+3. Percent of members/accounts removed/quarantined, trended by month
+4.Percent of paid completes from 0-3 months tenure, trended by month
+5. Percent of paid completes from smartphones, trended by month
+6. Percent of paid completes from owned/branded member relationships versus intercept participants, trended by month
+7. Average number of dispositions (survey attempts, screenouts, and completes) per member, trended by month (potentially by cohort)
+8. Average number of paid completes per member, trended by month (potentially by cohort)
+9. Active unique participants in the last 30 days
+10. Active unique 18-24 male participants in the last 30 days
+11. Maximum feasibility in a specific country with nat rep quotas, seven days in field, 100% incidence, 10-minute interview
+12. Percent of quotas that reached full quota at time of delivery, trended by month