Somia Data: measurement and intelligence for the Somia ecosystem

Somia Data is the measurement and intelligence platform for the Somia ecosystem. It is being designed to integrate aggregated information from its products, programs, experiences and settings, turn it into clear indicators and help organizations, funders and yeshcube teams make evidence-informed decisions.

The project is at the ideation stage. Its purpose, potential users and main deliverables have been defined. The technology architecture, shared data model and first integrations belong to the next phase of development.

From dispersed data to a shared view

A wellbeing solution may record sessions, resources used, locations, times or ratings. When each product organizes this information differently, understanding what is happening across the ecosystem becomes difficult.

This fragmentation particularly affects organizations that fund, deploy or maintain several solutions. Isolated data provides a partial picture and makes it harder to answer important questions:

• Which resources are used most often.
• In which settings they prove most useful.
• Whether people return after their first use.
• Which needs are being addressed.
• What differences appear between locations or periods.
• Which elements should be maintained, revised or expanded.
• Which results an organization can substantiate for its funders.

Somia Data aims to create a common framework for answering these questions through consistent definitions, aggregated information and measurement criteria adapted to each product.

What information it may integrate

Each Somia solution operates in a different setting. An acoustic booth, a family application and a voice experience generate different interactions. The platform will need to preserve those differences while establishing variables that support a combined view of the ecosystem.

The planned categories include:

• Participation and number of uses.
• Type of product, program or experience.
• Context and location.
• Resources used.
• Need associated with the interaction.
• Time of use.
• Continuity and recurrence.
• Perceived usefulness.
• Specific outcome indicators.
• Aggregated demographic characteristics where a defined purpose exists.

The final selection will depend on the product, its setting and the conditions that apply to each form of processing. Somia Data is being conceived around data minimisation. Every data point will need to answer a concrete question and serve a clear purpose.

Three levels of knowledge

Somia Data will distinguish three types of information so that results can be communicated accurately.

Activity and use

This layer describes how the service operates. It covers reach, sessions, frequency, recurrence, time of use, selected resources and distribution across settings. It can show whether a solution is reaching the intended people and how it fits into their everyday context.

Experience and perceived usefulness

This layer brings together ratings provided by participants, such as satisfaction, ease of use, relevance or perceived usefulness. These responses help teams understand the experience and identify opportunities for improvement. Their interpretation must preserve the context in which they were gathered.

Evaluated outcomes

This layer incorporates indicators defined through an evaluation methodology. It may include before-and-after measurements, validated instruments, comparisons between periods or studies developed with scientific partners.

Recorded activity shows use. Perceived usefulness describes people’s assessment. Demonstrating impact requires an appropriate evaluation design. Somia Data will preserve this distinction across its dashboards and reports so that each conclusion reflects the evidence that supports it.

How the platform is envisaged

The initial Somia Data architecture is organized around five components:

• Source integration, with connectors adapted to participating products, programs and implementations.
• Shared data model, defining concepts such as session, use, continuity, context, need, resource and outcome.
• Privacy and governance, with rules covering purposes, permissions, retention and aggregation levels.
• Analysis, producing indicators, trends, comparisons and signals that help explain each deployment.
• Dashboards and reports, adapted to clients, funders, internal teams and scientific collaborators.

Each organization will access only the information corresponding to its scope and agreed purposes. Permissions, indicators and levels of detail will need to be established for every deployment.

What an organization will receive

Somia Data may take the form of several services and deliverables:

• Monitoring dashboards.
• Periodic usage reports.
• Outcome and impact reports.
• Comparisons between periods or settings.
• Aggregated maps of use and need.
• Operational recommendations.
• Alerts about low use or incidents.
• Indicators for institutional reporting.
• Sector analyses.
• Specific studies developed with scientific partners.

An organization will be able to understand the reach of its deployment, identify patterns of use, assess continuity and decide where to focus resources. Recommendations will state which information supports them and what degree of certainty it allows. Analyses requiring specialist interpretation will include human review.

Shared infrastructure across Somia

Somia Within enables Somia capabilities to be integrated into first-party and third-party products and settings. Somia Atlas is being designed to locate those resources and facilitate access to the network. Somia Data will provide the measurement layer for understanding how they are used and how they evolve.

Products such as Somia Bloom, VVAVVE and Kiwikidoo may incorporate indicators suited to their experiences. Research and application programs may define specific variables linked to their questions and populations.

This architecture will make it possible to observe each implementation individually and build an aggregated view of the ecosystem where privacy and governance conditions allow it.

Useful information with privacy by design

Trust is a structural condition of Somia Data. Its development begins with several commitments:

• Clients and funders receive aggregated information.
• Reports exclude individual participant data.
• Personal data stays outside commercial activity.
• Information has no advertising use.
• Segments must reach a sufficient size before they are shown.
• Processing involving children requires enhanced safeguards.
• Every use of information must be tied to a defined purpose.
• Permissions and access must be documented.

The technical architecture will determine when information can be considered anonymised and when it remains pseudonymised. This distinction will form part of the infrastructure specification, together with retention rules, access controls and the assessments needed for each setting.

Value for clients and funders

Somia Data is conceived as part of the ecosystem’s recurring service. Its dashboards and reports will accompany a deployment throughout its lifecycle.

For a company or institution, this supports decisions about renewal and expansion. For a foundation or sponsor, it provides a way to understand the reach of a funded program and document its results. For a scientific team, it offers shared infrastructure on which to design specific evaluations. For yeshcube, it creates knowledge that can guide the evolution of products and experiences.

The infrastructure may be included in institutional licenses, recurring fees linked to Somia solutions, sponsorship agreements, reporting services and evaluation projects. Service levels, report frequency and pricing will be defined during commercial validation.

First phase of development

Somia Data is at ERL-0, initial hypothesis. yeshcube has defined its strategic function, main information categories and general responsible-use boundaries.

The first phase will focus on:

• Selecting one or two initial integrations.
• Defining a minimum shared taxonomy.
• Building the first data model.
• Designing an institutional dashboard prototype.
• Producing an initial report format.
• Establishing the privacy and permissions architecture.
• Validating clients’ priority questions.
• Developing a first pilot with a partner organization.

This phase will test the infrastructure’s technical feasibility, the quality of its indicators and the practical usefulness of its deliverables before its scope is expanded.

Collaboration and funding

yeshcube is seeking organizations interested in becoming design partners for Somia Data. The first collaborations will help define what information clients need, which indicators they understand and which formats support decisions.

The project is also open to specialists in data architecture, privacy, security, evaluation, visualization and applied research.

Initial funding will support the shared data model, first integrations, governance architecture, dashboard prototype and validation pilots. Companies, public authorities, insurers, foundations and sponsors can fund a particular phase, an integration or shared capabilities for the wider ecosystem.

Interested organizations can contact yeshcube or learn about the Allies scientific collaboration system.

Somia Data is intended to turn measurement into a tool for learning. Its development will help explain how Somia solutions are used, improve their design and provide clients and funders with information proportionate to the available evidence.