2026 Q2 Highlight: Qualification Providers record highest quality assurance and grading speed to date.

Q2 produced a marked change in professional qualification delivery. Across CIPD, CMI, ILM and ACCA programmes, 6,200 submissions entered the participating centre cohort. Centres using Sokros returned learner results within one day, against a wider operating norm of 7 to 14 days. Their quality measures rose at the same time, while the cost of marking each submission fell.

The comparison matters because the awarding-body standards did not change. The qualification levels, assessment criteria and quality assurance duties remained the same. The material difference was the marking model. As Sokros centres handled a larger share of the quarter's work, their results pulled the combined measures towards shorter queues, lower costs and grades that were more likely to be upheld.

Four awarding bodies moved together

CIPD supplied the largest cohort, but the pattern was not confined to people practice qualifications. CMI and ILM management programmes, and ACCA accounting programmes, all recorded higher Q2 volume through Sokros centres. Together, the four cohorts grew from 3,500 submissions in Q1 to 6,200 in Q2.

That breadth changes the reading of the quarter. A result from one provider can be explained by cohort mix or a quiet assessment window. Similar movement across four awarding bodies points to an operating change. More scripts were marked without a corresponding extension of the SLA or a weaker moderation result.

The SLA gap changed the average

The usual marking window for these qualifications is 7 to 14 days. It reflects a queue organised around marker capacity: submissions wait to be opened, checked, read and prepared for release. The Q2 Sokros cohort worked to a one-day learner results SLA. That measure includes assessor review and release, not only the time taken by the marking system.

The effect reached beyond the fastest centres. Once a material share of the quarter's submissions moved through a one-day service, the pooled turnaround measure fell. Sokros centres outperformed the wider range, and their volume began to change what a normal qualification marking window looked like.

Lower marking cost did not come from lower rates

Participating centres recorded a 58% reduction in cost per marked submission against their Q1 2025 operating baseline. The saving came from work around the judgement: opening and organising files, checking eligibility, reading a script again for moderation preparation, moving feedback into the LMS and chasing queue status. Assessor ownership of the final grade remained in place.

This distinction matters to the wider market. A lower marker rate can reduce cost for a quarter, but it does not alter the amount of work in the process. Removing repeated administration does. Centres could absorb higher submission volume without passing the increase into a longer queue or a larger marking team.

Quality improved as the queue shortened

Faster marking is useful only if the grade survives review. In the Q2 moderation sample, 98.6% of grades were upheld. A further 1.1% were adjusted and 0.3% required a re-mark. Grade agreement reached 96.4%, up 1.2 percentage points quarter on quarter and 4.1 points year on year.

The quality change came from a more inspectable first decision. Each grade carried the criterion, the applied rule and the relevant quotation from the learner's work. IQA and EQA reviewers could test the basis of a judgement directly. The moderation record indicates that shorter reading time did not transfer work or risk into quality assurance.

Learner satisfaction followed the wait

The four awarding bodies do not use one common learner-satisfaction question, so a single sector score would overstate what the research can show. The consistent operational finding was narrower: results arrived 6 to 13 days earlier at Sokros centres, status-chasing contacts fell, and referred learners could begin a resubmission while the brief and feedback were still recent.

This was most visible where a submission could not yet be marked. Eligibility checks identified missing signatures, incorrect briefs and word-count breaches before the work entered the marking queue. Learners received a reason promptly instead of waiting the full SLA for an administrative return. The improvement in satisfaction was associated with certainty and useful feedback, not speed in isolation.

CIPD provides the clearest test

CIPD is the largest and broadest cohort in the Q2 research: 2,400 submissions across Levels 3, 5 and 7, delivered by 34 centres in 20 markets. It also has the fullest quality record. Agreement reached 96.4%, 98.6% of sampled grades were upheld, and the learner results SLA was one day.

The scale makes CIPD the focal point without making it the whole story. Its centres carried enough volume to affect the quarterly average, while the range of levels tested whether the result held across different command verbs, evidence demands and moderation depth. It did. The cost index fell to 42 while the quality measures reached their highest point in the series.

What the market evidence supports

Q2 does not show that every qualification provider has reached a one-day SLA, or that technology alone determines learner satisfaction. It shows a measurable split within the same market. Centres using Sokros marked more work, at lower unit cost, with stronger quality outcomes and materially shorter waits than the wider operating norm.

As those centres take a larger share of delivery across CIPD, CMI, ILM and ACCA, the split is moving the aggregate result. The wider improvement is not separate from Sokros adoption. The Q2 figures indicate that it is being driven by it.