Saudi Arabia's Digital Experience Maturity Index reached 87.06% in 2026 and every top-ten platform passed 91%. Where the last points hide, and who owns them.
In August 2026 the Digital Government Authority (DGA) announced that Saudi Arabia's Digital Experience Maturity Index for Government Services (DXMI) had reached 87.06%, at the Advanced level. The cycle covered 59 platforms, and more than 805,500 people answered the user survey that feeds it. Absher led with 94.38%, Etimad a fraction behind at 94.36%, then Senaei, Balady and Tawakkalna. The Ministry of Tourism portal was tenth at 91.16%, so the whole top ten sits above 91%.
The national result is 7.32 percentage points below the leader. Every entity whose platform is still short of a score in the nineties faces two questions. What separates 87 from 94? And why do the first points come quickly and the last ones slowly?
The middle is our term, not a DGA level: in 2025 the national result and all ten leading platforms were rated Advanced alike. It is a stage anyone who has run a digital service will recognise. Services exist and work, and users usually finish their transactions, yet the experience is uneven. Steps are clear in one service and confusing in the next. A complaint arrives, and the reply does not. A platform can sit here with a score above 80. So can a company.
Five years of index results suggest this stage is far quicker to reach than to leave: the national curves climbed fast, then flattened. Our reading is that the first points come from building and publishing services. The last come from work that sits between departments and that no one owns by name: following a complaint after it arrives, making services usable for people with disabilities and older users, measuring the experience as it happens, and keeping evidence current between cycles.
Two indices, two different questions
The DXMI reads a service through the eyes of the person using it. It groups 20 themes under four perspectives: Beneficiary Satisfaction, User Experience, Complaints Handling and Technologies & Tools. A large user survey feeds into it, and DGA revises the methodology every year.
The Digital Transformation Measurement Index, known as Qiyas, looks at the entity from the inside. Its 2025 cycle assessed 240 entities against ten perspectives, 23 axes and 95 standards. Entities upload evidence to the Qiyas system, DGA auditors verify it over two rounds, and results place each entity in one of five stages: Building, Availability, Optimisation, Integration and Innovation. In 2022 DGA changed what this index measures, moving it, in its own words, "from measuring maturity to measuring the extent of adherence to standards".
Maturity is what users feel while trying to finish a transaction; compliance is what the entity can prove with documented evidence. An entity can meet and document every requirement of a standard while its users still stall at step three. We read the experience through the DXMI and use Qiyas to look inside the entity.
A curve that climbs, then slows
Saudi entities have been in the literal middle before. More than a decade ago, the Yesser programme's fifth e-government transformation measurement averaged 53.65% across 152 entities. It was a different instrument from another era and cannot be plotted alongside today's indices, but the average entity then stood near the midpoint of its scale.
In 2022 the national DXMI result was 77.26% across 12 platforms. It rose to 80.68% in 2023, 85.04% in 2024 and 86.71% in 2025, before reaching 87.06% in 2026. The yearly gains were 3.42 percentage points, then 4.36, 1.67 and 0.35.
The Qiyas digital transformation index traces the same curve. Its results reports show the national result rising from 69.39% in 2021 (DGA's restated figure for that year) to 88.30% in 2025: 11.57 points in the first year, then 4.57, 1.61 and 1.16.
Read these series with care. DXMI coverage widened from 12 platforms to 59, and the 2022 cycle used three perspectives before Technologies & Tools arrived in 2023. Qiyas went from 125 standards in 2023 to 96, then 95, and news coverage puts the 2026 cycle at 89. The platform map shifted too: the Vision 2030 Annual Report 2025 records 327 government platforms merged or closed and 490 still operating. Results are not like-for-like across years, and the slowdown may partly reflect new platforms joining each cycle and the shrinking distance to 100.
The national plateau does not hold every platform back, though. In the year the national result added only 0.35 points, Senaei climbed from tenth place at 89.58% to third at 92.83%.
The nineties remain the preserve of a few. In the 2025 Qiyas results, 26 of 240 entities reached the Innovation stage, about 11%, all of them scoring above 90%. In 2022 there were seven.
Maturity is what users feel as they try to finish their transaction.
The last points are won in the spaces between departments.
Where do the last points hide?
The DXMI grades results on five ascending levels: Emergent, Developed, Competent, Advanced and Exceptional. In DGA's infographic of the 2025 results, three perspectives reached Advanced: User Experience at 88.53%, Beneficiary Satisfaction at 87.09% and Technologies & Tools at 86.64%. Complaints Handling, at 80.05%, was the only perspective still at Competent. The Digital Inclusion sub-index, which measures whether people with disabilities and older people can use services on their own, stayed at that level in both years: 75.30% in 2025 and 76.98% in 2026.
In Qiyas, a small gap is enough to lose full compliance. Under the Core Digital Transformation Standards, an entity earns full compliance on a standard when it meets every requirement, partial compliance when it misses one, and non-compliance when it misses two or more. A single requirement is easy to miss when two departments share it.
According to DGA's reports, the entity that led Qiyas in 2022 at 92.38% was 48th by 2025, at 86.49%. Another, fourth in 2024 at 94.09%, was 42nd a year later at 87.85%. Meanwhile the Ministry of Human Resources and Social Development rose from 91.76% in 2022 to 95.24% in 2025, taking first place. The reports do not explain each decline, and changing standards may account for some of it. Even so, a high score in one cycle guarantees nothing about the next.
The Ministry of Energy, the most improved entity of 2022, added 70.14 points in that cycle, yet the leap landed in Integration, at 87.47%. By 2025 it stood fourth, at 94.15%. The other big leaps also stopped short of the top: from 2022 to 2025, none of the five most improved entities in any cycle went straight into Innovation. All landed in Integration or Optimisation.
We read the jump as the product of building, and the points above 90 as coming later, from the spaces between departments. User Experience, the strongest perspective, sits largely with one team that can design and test screens and service steps. A complaint crosses departments. It starts with the user and ends, if it ends at all, with the department that runs the service, and none of what it teaches may reach the service's designers. Digital inclusion needs testing with people with disabilities and older users, plus content and technical decisions beyond any design team, and it may not be assigned to anyone by name. And when a project team that disbands after the cycle gathers the Qiyas evidence, that evidence goes stale before the next one.
Enterprises sit in the middle too
No national announcement tells a company each August where it stands. That makes its problem harder: it can stay in the middle for years without ever naming it.
Research finds a similar slowdown in business. MIT's Center for Information Systems Research (CISR) asked firms how much of their digital transformation they had completed: on average 33% in 2017 (400 firms), 50% in 2019 (1,311) and 55% at the end of 2022 (721). These are self-assessments, and CISR calls the latest stage a stall. The causes executives named include a lack of top-management ownership, slow decisions, competing investments, moving targets and difficulty proving value. None is a technology problem, and the first, ownership, echoes the gap we read in the government numbers.
In its 2019 survey, CISR classed 22% of firms as "future ready", and those firms reported revenue growth 17.3 percentage points above their industry average and net margins 14.0 points above it. That links maturity and financial performance in self-reported data; it does not prove causation.
Boston Consulting Group's 2020 study gives some of the clearest odds: 30% of digital transformations met or exceeded their target value, 44% created some value but missed the target, and 26% created little value. That split serves a decision-maker better than the familiar line that 70% of change initiatives fail, which a peer-reviewed review by Hughes in 2011 found to lack valid empirical evidence.
In the Gulf, a 2024 BCG study placed 25% of the region's organisations in the top two levels of digital maturity, against 31% globally. At home, figures from the General Authority for Statistics show that 33.1% of establishments in the Kingdom used AI in 2025, up from 27.6% in 2024.
As the tools spread, they set fewer companies apart. For a transformation leader in a large Saudi enterprise, the difference is whether each part of the customer experience has an owner who can show what changed for the customer. Showing it means measuring it, so the company has to build a digital transformation maturity assessment of its own and accept what it finds, welcome or not. It need not start from a blank page: the four perspectives behind DGA's index work as an internal frame for customer and supplier journeys, each with an owner and a number.
How the last points are won
If the next DXMI results arrive in August, as in the past five cycles, an entity has less than a year. Start from your platform's score on each perspective if you can see it, or your internal measure if not; your weakest perspective may not be the weakest nationally.
Take a hypothetical platform that scores 92 on User Experience, 88 on Beneficiary Satisfaction, 70 on Complaints Handling and 85 on Technologies & Tools. For the 2025 cycle, DGA published perspective weights of 50% for User Experience, 25% for Beneficiary Satisfaction, 15% for Complaints Handling and 10% for Technologies & Tools. Applied directly at perspective level, those weights give the platform 87.0. Raising Complaints Handling alone from 70 to 90 lifts it to 90.0, a gain of 3.0 points. Raising User Experience alone from 92 to 96 adds 2.0 points.
This is a simplified illustration; DGA's exact calculation sits in its methodology guide. The complaints move adds more, but it needs twenty points against four on User Experience, and a User Experience point weighs more than three complaints points. Complaints are the shorter route only if those twenty take less than one and a half times the effort of the four, which sit near User Experience's ceiling. That may well be so, because the fix starts with organisational questions: who receives a complaint, and who can change the service because of it. The general rule: multiply each perspective's weight by the distance you can realistically cover in it this cycle, start with the largest, and remember that weights can change between cycles.
The practices below suit entities and companies alike. In a company they fit best where little is measured today, such as onboarding a supplier or approving an invoice, and in any service the public uses directly, from a banking app to an appointment booking portal. None guarantees a result, but each targets a place where the numbers show points being lost.
First: give every perspective an owner and a number. Digital inclusion gets one too. Index results are annual, so the number the leadership meeting reviews each month is an operational indicator from the entity's own service data: first-attempt completion for User Experience, time to close a complaint for Complaints Handling, the share of services tested with people with disabilities and older users for inclusion. Once a number has an owner, "why did we stay at 80?" has an answer.
Second: measure in the moment of use. An entity that waits for the annual DXMI to learn where it stands learns late. Begin with data that needs no question: where users drop out, and what makes them call support. Then add a short question at the end of the transaction, placed so it does not compete with DGA's survey, and review answers weekly with the team that can change things. The annual figure then confirms what the entity already knows.
Third: treat complaints as design input. Map complaint categories to the service steps where they arose, and whenever complaints lead to a change, tell users what changed. A company can test its handling of corporate clients' complaints with one question: who owns a complaint from the client's first message to the last change it causes in the service?
Fourth: build inclusion into the first release. Bring people with disabilities and older users into testing before a service launches, and make accessibility compliance a condition of every release approval. Adding inclusion after launch means reopening an approved design and waiting for another development cycle.
Fifth: keep evidence current all year. Standards change between cycles, so make evidence a by-product of the work: file the minutes of a decision the day it is taken, and assign each standard to a department that owns it all year. In a company, draw experience indicators straight from operational data, so nobody builds a special presentation before a board meeting.
When the number becomes the target
In 2006, Bevan and Hood studied performance targets in English public healthcare, documented how they were gamed, and cited Goodhart's law: "any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes." The lesson travels: when an assessment or ranking rests on a number, part of the effort shifts to the number.
A good index protects itself with the user's voice and independent verification. In the UN E-Government Survey, where Saudi Arabia rose from 43rd place in 2020 to sixth in 2024, two researchers assess each country's online services independently. For the two Saudi indices, a user survey feeds the DXMI, and Qiyas goes through two verification rounds.
Stating limits is the other safeguard. When the World Bank published its GovTech Maturity Index 2025 update, which gave Saudi Arabia 0.9964, the second-highest score among 197 economies, it noted that the index "is not a ranking" and that most of its 2025 data was self-reported by governments.
Entities and companies share the risk that a cycle becomes a season for improving the number, with rushed evidence and cosmetic fixes before the survey. The number may rise once that way, but the user who returns a month later finds the service unchanged. The score is a reading of the work, never a substitute for it.
The question before the number
Before the next figures for either index, every transformation leader should put one question to their team: which part of our experience does nobody own yet?
It might be a blind user never invited to a test, or a complaint category that no department reports on. Give it an owner and a number before the next cycle opens.
