Data Science > Web Analytics > Predictive Segmentation
Predictive segmentation is the ability to programmatically identify and generate significant visitor segments which are first of all fully compliant with current regulations such as the General Data Protection Regulation (GDPR) and others, and which then may respond in a certain way to specific events or conditions.
Before the introduction of forecast segmentation, the analysis, identification and segmentation of Internet traffic was done manually, carefully and slowly. The routines required marketers to crunch and analyze the data derived from analytics, and then to analyze each separate segment separately. This was necessary to discover useful information or a typical behavior associated with primary KPIs.
This slow method of segment identification is done using an approach in which traffic is segmented from the bottom to the top, where each available data point is sized-up in contrast to the primary KPI. This is the opposite of a top-down model, in which visits are segmented by KPI performance and only then reduced according to the captured data.
Available GDPR-compliant data points include:
Regulatory Compliance Data Restrictions (ie: GDPR)
Operating System data
New guest visitors
Lifetime Value (LTV)
Any data points that do not comply with current regulations are excluded.
With data that is first of all compliant with all relevant regulations, smart data-driven marketers scour the available segmented traffic data to find the most important or worst performing elements. Once that is applied to the data, then data-driven marketers take action on the data by targeting each audience segment to give a custom experience for each audience segment. Innovative and best of breed marketing processes A/B/C tests on the audience segments to ensure that the correct experience is produced for each segment.
Predictive segmentation eliminates labor-intensive repetitious work by automatically detecting and deploying high-traffic or potentially large audience segments to target. That said, as an independent force, it is not possible to determine the right experience to adapt to each identified sector. This means that predictive segmentation promotes the position of the marketing work process. It also gives them the specific areas to work on, even as the corresponding experiences stay as assumptions that should be verified and confirmed.
The final hurdle of the process concerns new tech developments as well as regulations such as GDPR, including “predictive targeting” on content personalization optimization and eCommerce ROI optimization platforms such as Personali (learn more about predictive analytics).