Estimated reading time: 17 minutes
Social value reporting has matured fast, but the data underneath it hasn’t always kept pace. In a recent Thrive webinar, our team explored why more organisations are becoming exposed to scrutiny they aren’t ready for, what good social value data actually looks like in practice, and how real-time validation helps teams build defensible data from the point of entry rather than fixing problems months later.
This recap covers the discussion, the audience poll results, and answers to the questions raised on the day about evidence, GDPR, supply chains and getting started.
From measurement to proof
Until recently most organisations were still working out how to measure social value at all. Today that data feeds procurement decisions, contractual commitments, ESG and annual reporting, and Cabinet Office returns.
But as Danni Aberg, Head of Social Value at Thrive, explained setting the scene, the sector has matured quickly – and that is a good thing. Nevertheless, as expectations around social value have risen, expectations around the underlying data have risen with them.
Most people working in the space experienced opening a report and not being sure where a number came from, or whether they could defend it if challenged.
This isn’t because people are gaming the system. but because measurement has matured faster than the assurance, governance and operational controls that should sit around the data. Organisations have understandably focused on getting reporting established first. The conversation is now shifting to whether that data is consistent, defensible and properly evidenced, Danni explained.
Confidence in the outputs, not always the inputs
Benjamin Fellowes, Director of Projects and Innovation, framed the core problem: organisations often have confidence in their outputs without the same confidence in the data underneath them.
Externally, social value reporting has become sophisticated. But internally, many organisations still rely on spreadsheets and manual interpretation, with any checking done retrospectively against fragmented evidence. While that used to be tolerated because scrutiny of social impact reporting wasn’t deep, Benjamin warned, that is changing.
Part of the issue, the panel agreed, is that the market has tended to equate validation with “whether the right fields are filled in”, rather than whether the data is accurate, attributable, evidenced and defensible.
The risk is a false sense of security: data that passes a submission but wouldn’t stand up to any deeper examination. As procurement teams, auditors and increasingly investors become more demanding, the gap between data that is easy to submit and data that is genuinely defensible is becoming a real exposure.
Michael Bole, Product Lead at Thrive, brought the operational view. The most common ways social value data fails a quality check are familiar:
- Missing evidence
- Inconsistent interpretation of framework requirements across a business
- Duplicate claims where something is recorded twice or applied across several metrics when it should sit against one
These problems are rarely deliberate. They happen because the people entering data aren’t given enough support to understand what good looks like.
Teams and suppliers are often expected to grasp complex frameworks instinctively or after a cursory training session without embedding the rules in the organisation’s own processes:
- Does spend count when the invoice is raised or when payment leaves the account?
- Is training genuinely additional, or mandatory and therefore not claimable?
When those rules live in people’s heads rather than being defined and consistently applied, variation is inevitable, and six months on someone is left working out why two projects reported the same thing in completely different ways. Michael suggested.
When a frustration becomes a commercial risk
Asked when poor data moves from an operational frustration to a commercial risk, Benjamin was direct: it is the moment an organisation relies on its numbers externally.
Once social value data feeds contract commitments, annual reports, public disclosures and investor conversations, the organisation is no longer just reporting activity, he warned. It is making claims it may need to defend.
He pointed to research Thrive commissioned reviewing 900 ESG and impact reports across the UK, US, Australia and other markets. It found that social disclosures are now independently audited and published more than half the time, a significant shift from a few years ago. More telling, it also showed organisations dropping impactful case studies and social data from their reports because their teams couldn’t easily assure it. Real impact is being left out, not because it didn’t happen, but because the data behind it couldn’t be found or backed up.
The regulatory backdrop is hard to ignore. Social value is embedded in central government procurement through PPN 026 and PPN 002 and the Social Value Models, and sustainability claims face growing scrutiny through measures such as the FCA’s anti-greenwashing rules.
Indeed, environmental reporting offers a useful template: a decade ago that landscape was fragmented and largely unassured, before the conversation moved toward assurance, traceability and auditability, Benjamin argued.
Social reporting is following the same path. The honest test is simple: if a client, investor or auditor challenged one of your figures tomorrow, could you explain where it came from, how it was calculated, what evidence supports it, and whether the rules were applied consistently?
What the audience told us
Midway through, we polled attendees on exactly that. The results matched what we see across the market:
- Only around 9% felt very confident in their processes and evidence
- Close to half were broadly confident but aware of gaps
- The rest split between those finding consistency and evidence a challenge and those still early in their journey
The small group at the top is broadly the group already doing some form of assurance, mirroring the global research almost exactly.
The most common challenges, consistency and evidence, tend to come from decentralised structures, varied governance across business units, and the difficulty of extracting data from a supply chain. A recurring early-stage trap is trying to do twenty things at once rather than ten things well.
What good data looks like, operationally
So what does good social value data mean day to day?
Michael set out five qualities of “good” data:
- Evidenced
- Specific
- Consistent
- Attributable
- Plausible
In practice, someone reviewing the data later should understand who delivered something, what happened, when, and how it aligns to the framework, with the evidence to support it, without needing the original person beside them to explain it.
Much reporting comes unstuck precisely here, leaning on the institutional knowledge that the person entering the data has and assumes everyone else shares. The discipline, as Michael put it, is taking yourself out of the equation and asking whether someone without your knowledge would understand it.
That diagnosis is where Thrive’s validation approach began.
Rather than expecting people to instinctively know the framework rules, the system helps them understand what good looks like at the point data is entered. It is built from the ground up around the framework logic itself, with more than 1,000 rules that flag potential over-claiming, duplicate submissions, missing evidence or declarations, inconsistent application of metrics, and figures that don’t look plausible.
Michael showed the new system in action, trying to input a record deliberately seeded with errors, which returned feedback immediately on submission – not in a report months later or an email from an auditor.
The demo was running in “advisory mode”, which flags issues but still lets the record through, because the aim isn’t to let perfect be the enemy of good. However, a stricter mode (Full Data Entry) built into the platform can prevent submission until everything is right. Both can be configured at project level alongside an internal review step.
As Benjamin drew out, validation is most when it is embedded in the reporting process itself. If it happens at the end, or six months later, that is exactly where organisations come unstuck.
Validation, verification and audit
Thrive uses a three part schema for the wider assurance journey: validation, verification and audit.
They sit under one umbrella but solve three different problems:
- Validation is the foundation, confirming at the point of entry that a claim aligns with the framework rules, includes the required declarations, has evidence attached, looks plausible, and isn’t contradicted by issues like over-claiming or double counting
- Verification builds on top, moving from “does this data appear robust” to “how confidently can we rely on it as true,” with the consulting team looking critically at whether the evidence actually supports the claim. This stage matters most for organisations disclosing social value externally
- Audit is the highest level of scrutiny, examining governance, controls, systems and processes holistically, and typically sits with organisations in highly regulated environments
Businesses can enter this journey at very different stages, and most UK organisations currently sit between validation and verification.
As scrutiny matures we anticipate a steady move toward more formal assurance, but nobody needs to operate at the highest maturity level tomorrow. The task is to understand where you are now and take sensible next steps.
Small practical steps, whatever platform you use
Improving data quality in practice tends to focus on small operational changes rather than huge transformation programmes, and starting early avoids the painful retrofitting many organisations went through when CSRD landed.
For SMEs in particular, less is more: do a couple of things well, focus on your most challenging metrics, and look at how to reduce the burden of the largest, most line-item-heavy data. Once a few metrics are right, expanding is easier.
This isn’t only for Thrive customers. If you are still working in spreadsheets, you can start by choosing a subset of metrics and creating a centralised document for your teams setting out the metrics you are focusing on and the specific evidence requirements for each, accounting for both the framework you use and the nuance of how your own business operates. That groundwork is where some of the best-prepared organisations begin. The whole industry, we believe, needs to move in this direction, and we will keep publishing guidance to help, customer or not.
Find out more about validation or get in touch to arrange a demo and see it for yourself. If you’d like a recording of the webinar itself please contact us and we’ll send it over!
Questions from the webinar
We didn’t have time to answer every question live, so here are the main ones raised, with our responses. Where several people asked variations of the same question, we have combined them.
Context matters, but at an organisational level, where resources are limited and social value data is readily available from core systems, we recommend selecting a limited set of key metrics with a regular reporting cadence. You can then supplement those on specific procurements where a client has specified particular requirements. We can advise if you like based on our experience the best, most impactful metrics you can start with.
The practical steps covered in the session are deliberately designed for organisations at any stage. Better evidence discipline, more consistent data capture, and mapping what you report against the standard you follow are all things any team can start on without significant resource. Moving in the right direction, however incrementally, is what matters. Part of what makes this manageable for smaller teams is having clarity on exactly what data needs to be captured in the first place, so resource is focused precisely rather than spread thinly. Paired with a consistent set of measures reported on regularly, that creates a cadence that holds up to scrutiny without placing undue burden on the team.
Social value claims need to be substantiated to be credible and defensible. The Impact Evaluation Standard sets out the evidence requirements for each metric, and Thrive’s in-platform validation is built around these, giving organisations a structured, auditable way to demonstrate the social value they deliver. Organisations should familiarise themselves with those requirements. Typically, it would suffice to enter an Employee ID number rather than a name. Where evidence involves personal data, the organisation is responsible for having the appropriate reporting governance and data protection frameworks in place. In practice, those responsible for social value data are often not the owners of the underlying evidence, so it is common, and valuable, to collaborate with internal departments. Your HR team, for example, may be able to provide apprenticeship agreements pre-redacted. It is also possible to use unique identifiers, such as an employee number rather than a name, when uploading data and evidence to the system.
The Thrive ruleset covers the full breadth of the Impact Evaluation Standard across community, wellbeing and service user outcomes and metrics, so there is likely more coverage there than you might expect. If you have specific measures in mind, we would be glad to explore them with you after the session.
It depends on your setup. If validation is enabled for your organisation, Full Data Entry is required, because that is what allows the rules to run. Within that, there are two modes: strict, which holds submission until all rules are met, and advisory, which flags rules not yet met but still allows the record to be submitted. The default is set at organisational level and can be adjusted per project. If validation isn’t enabled, your administrator can configure whether Full Data Entry, Basic Data Entry, or both are available. On evidence: claims need to be substantiated on a per-metric basis to be credible and defensible. The IES sets out the evidence requirement for each metric in-platform, and we recommend familiarising yourself with these early. New customers can also opt to receive an Onboarding Consultancy Report, developed with them and based on factors including the availability of data and evidence, which provides bespoke metric prioritisation across the immediate, near and long term. Your Customer Success Manager can confirm exactly how your platform is configured.
The same principle applies: claims need evidence to be credible and defensible, and the IES sets out the requirement for each metric in-platform. Because social value specialists are often not the owners of the underlying data and evidence, meaningful collaboration with internal departments is usually necessary to enable IES-compliant claims. Familiarising those departments with the requirements early in the journey makes that collaboration far smoother.
Spending time with other departments to define and agree a reporting cadence is the key to reducing double or triple handling. The platform also has API capability, which can significantly reduce manual data handling by drawing data from source systems rather than needing teams to re-enter it.
This is a very common challenge. The platform is designed to flex around where you are: not every field is mandatory by default, so you can capture what your supply chain can realistically provide and build from there as those relationships mature. The aim is to make progress with what you have rather than wait for a perfect dataset. It is also worth knowing that data collectors don’t need platform access or logins at all. They can be assigned to specific metrics and receive automated prompts in line with your reporting cadence, submitting via a simple link, which shifts responsibility to the people who hold the data and reduces the burden on your team. We are also working on functionality to make bulk submission more straightforward. Your Customer Success Manager can help you think through a practical approach to improving supply-chain data capture incrementally.
No. Evidence such as images sits in the storytelling and reporting side of the platform rather than being a data-quality requirement, so there is no obligation to upload the same files multiple times. Once it is added against one record it is in the system and accessible, and it will surface across dashboards and reporting without being duplicated. If you prefer to add it against multiple records, that is fine too. It comes down to your own workflow.
The real-time nature of the guidance is designed to do much of that heavy lifting. Users are pointed toward what is needed at the point of entry, rather than having to learn the full ruleset upfront. Your Customer Success Manager can also walk you through how other organisations have approached the transition and what has worked well.
Yes. You can do both in the platform. When records are submitted they are validated, but even validated records stay pending until someone internally approves them, so your own review step sits alongside the automated checks.
In short, Thrive Validation is a real-time data quality layer that checks records against the requirements of the Impact Evaluation Standard as they are entered, so gaps or inconsistencies are flagged immediately rather than discovered later. In practice that means data that is more robust, more consistent, and better positioned to stand up to scrutiny. There are two modes: one that guides your team through the rules without blocking submission, and one that holds submission until all checks pass. Whether validation is part of your subscription depends on your current tier, so your Customer Success Manager is best placed to walk you through what is available and what getting set up would look like.
The Impact Evaluation Standard has recommended evidence requirements published in-platform, and key information for these metrics forms part of validation. For volunteering, evidence might include event volunteer logs, photos of the activity, or confirmation from the volunteering beneficiary.
We are using social value in the broad sense, including ESG, CSR and procurement-driven social value.
Asset-based approaches are built on relationships and strengths, and there is a real tension in making sure data capture serves that rather than detracting from it. A couple of things tend to help: focusing on a consistent, meaningful, smaller set of measures rather than trying to capture everything, so collection becomes a light-touch habit rather than a burden; and distributing responsibility through the organisation by appointing people at team or programme level who own data for their area, rather than centralising it. Both reduce friction and keep data capture close to the work itself. We would be happy to explore what that could look like in your specific context.
Alan Boyce
Senior Researcher More articles by Alan >
