UK Procurement Portals Ranking 2026: Where each platform stands

The Tenderline Portal Index weighs what the notices actually show. In-tend leads at 77.9 of 100, followed by ProContract (74.1) and MultiQuote (68.4). Positions are relative to this 18-portal cohort and this edition. The chart shows exactly how every total is built, so a reader can disagree with the weights and still use the parts.

What separates the leading portals?

One index, four stories behind it

A single total also hides differences worth knowing. The same 18 portals reorder when you ask a different question: procedure volume, buyer reach, published value or SME value share. Follow a portal’s line to see where the index agrees with your own priority, and where it does not.

Compare every portal on the question that matters.

The explorer opens on the overall index with its four criterion scores. Every other view is an independent comparison: activity can be ordered by a primary sector; award values retain their monetary sample; SME shares retain their classification rules; competition uses completed Open-procedure lots. Procurement timing is context about the buying process. Missing measurements remain visible.

Overall indexActivityAward valuesSME outcomesBids per lotPublished informationProcurement timing

Your primary sectorAll sectorsAgricultural, farming, fishing, forestry (03)Petroleum, coal, oil and gas (09)Mining, basic metals and related products (14)Food, drink, tobacco (15)Agricultural machinery (16)Clothing, footwear, bags (18)Leather and textile fabrics (19)Printed matter and related products (22)Chemical products (24)Office and computing machinery (30)Electrical machinery (31)Radio, television, communication (32)Medical equipments, pharmaceuticals (33)Transport equipment and products (34)Security, fire-fighting, police (35)Musical instruments, sport, games (37)Laboratory, optical and precision (38)Furniture, furnishings, domestic appliances (39)Collected and purified water (41)Industrial machinery (42)Mining, quarrying and construction machinery (43)Construction structures and materials (44)Construction work (45)Software package and information systems (48)Repair and maintenance services (50)Installation services (except software) (51)Hotel, restaurant and retail trade services (55)Transport services (except waste transport) (60)Supporting and auxiliary transport; travel (63)Postal and telecommunications services (64)Public utilities (65)Financial and insurance services (66)Real estate services (70)Architectural, construction, engineering (71)IT services: consulting, software, internet (72)Research and development services (73)Administration, defence and social security (75)Services related to the oil and gas industry (76)Agricultural, forestry, horticultural services (77)Business services: law, marketing, consulting (79)Education and training services (80)Health and social work services (85)Sewage, refuse, cleaning and environmental (90)Recreational, cultural and sporting (92)Other community, social and personal services (98)

18 portals · portal index (0–100), highest first

All 18 portals. Each view ranks a separate measure; sector selection switches to activity. Values are published award amounts, not payments.

Portal	Portal index (0–100) ↓ i Portal index (0–100). This table retains all 18 portals and marks missing observations explicitly.	Market reach ↕ i Market reach. This table retains all 18 portals and marks missing observations explicitly.	Competition & bid depth ↕ i Competition & bid depth. This table retains all 18 portals and marks missing observations explicitly.	SMEs & new suppliers ↕ i SMEs & new suppliers. This table retains all 18 portals and marks missing observations explicitly.	Published data quality ↕ i Published data quality. This table retains all 18 portals and marks missing observations explicitly.
1In-tend	77.9rank #1 of 18	97.8	62.3	67.5	80.7
2ProContract	74.1rank #2 of 18	95.8	60.8	66.6	67.4
3MultiQuote	68.4rank #3 of 18	51.5	98.1	58.9	68.3
4Atamis	67.3rank #4 of 18	87.2	48.5	55.1	76.2
5Delta	67.2rank #5 of 18	83.5	51.3	64.9	65.5
6myTenders	64.7rank #6 of 18	55.1	73.7	65.2	67.3
7EU-Supply	64.4rank #7 of 18	68.1	52.8	60.0	78.9
8Sell2Wales	64.1rank #8 of 18	67.8	34.5	77.8	78.4
9PCS	62.2rank #9 of 18	80.7	50.0neutral: no comparable PA 2023 sample	64.8	46.5
10Panacea	60.9rank #10 of 18	17.4	78.2	84.5	75.0
11JAGGAER	60.7rank #11 of 18	74.2	44.3	57.3	65.0
12Litmus TMS	56.5rank #12 of 18	25.0	75.0	72.1	61.0
13eTendersNI	56.4rank #13 of 18	61.3	18.6	77.8	69.5
14London Tenders	50.1rank #14 of 18	17.5	50.0neutral: peer ratio below sample gate	80.3	61.4
15SAP Ariba	49.1rank #15 of 18	36.9	50.0neutral: peer ratio below sample gate	46.5	69.7
16East Midlands	45.5rank #16 of 18	25.5	50.0neutral: peer ratio below sample gate	64.4	46.4
17OPEN-UK	45.2rank #17 of 18	30.0	55.3	48.2	51.4
18MOD DSP	36.9rank #18 of 18	18.5	0.0	64.8	75.6

The Overall index view shows the Tenderline Portal Index (recipe and weight sensitivity in the next chapter); every other view measures one separate question. Public Contracts Scotland keeps a neutral competition score because its notices sit outside the comparable PA 2023 sample. Procurement timing describes the buying process, not website speed, and is not an index input. Award values and activity cover the 24-month cohort; bid counts cover completed Procurement Act 2023 lots. Missing observations stay visible. Sector selection orders activity; switching view returns to all sectors.

Why the two largest channels deserve separate descriptions

ProContract appears on 4,687 procedures and carries £10.90bn of published award value, the highest totals in this sample. In-tend reaches 466 distinct buyers against ProContract’s 315. Together they appear on 8,790 procedures, 34.9% of the ranked sample.

The difference matters: repeated buying by a smaller set of organisations can create more procedures, while a wider buyer network can expose a supplier to more institutions. Neither pattern alone establishes a better service. Their strengths become clearer when the volume, value and buyer columns remain separate.

An overall score is an editorial choice about what to value, not a hidden statistical discovery. We publish the weights, the inputs and the transformation below, so a reader can disagree with those choices and use the unweighted measurements in the previous chapter instead.

Four criteria, adding to 100 points

For each portal the total is 0.30 × reach + 0.25 × competition + 0.25 × openness + 0.20 × publication. Calculation retains full precision; the article shows one decimal. Exact ties use the permanent portal slug. If a cohort-relative input is identical across all 18 portals, it receives a neutral 50. A missing required observation stops the calculation rather than receiving an invented value.

Why some portals keep a neutral competition score

Received-bid counts support a narrower comparison: completed, contested lots under the Procurement Act 2023, excluding frameworks and dynamic markets. Public Contracts Scotland publishes under the Scottish procurement regime, so its activity cannot be judged as if it were a PA 2023 sample. Three small channels (SAP Ariba, London Tenders, East Midlands Tenders) hold measured lots, but fewer than the 150 completed procedures the study requires before computing a peer-matched ratio. All four keep every other criterion and a visible neutral 50 for competition, rather than being quietly excluded or scored on incomparable data.

What changes the order, and what the data cannot rate

Weights express priorities, so the article publishes how the result moves when they change:

In-tend holds first place in every scenario we tested; the runner-up changes with the weights. A one-point difference is not a statistically established gap: several measures correlate, and the index is a descriptive comparison, not a causal model.

The deadline-to-award interval is deliberately not an index input. It is mainly a feature of each buyer’s process and publication practice; relabelling it “platform speed” would go beyond the evidence. It stays in its own chapter as context. Likewise, a high SME share can reflect local buying, smaller lots or a service mix, and a complete structured field can still contain an error.

Search quality, account reuse, document access, upload reliability and page response times require a separate hands-on product audit. Commercial review-site scores often describe an entire enterprise suite, so copying them into a UK tender-portal league table would not make the comparison more rigorous.

The attributable single-supplier awards attached to the 18 portals carry £53.37bn of published net GBP value. ProContract leads at £10.90bn, followed by In-tend at £9.24bn and Atamis at £7.67bn. These are recorded award values over contract terms, not invoices paid during the study period.

Value is available for 11,464 of 19,120 attributable award rows (60.0%). Multi-supplier awards, gross-only amounts and amounts failing the shared attribution rules are outside this monetary sample. The catalogue includes utility and other regulated buyers as well as public authorities, so its total is not synonymous with government or taxpayer spending.

A portal can be broad, specialised or dominated by large awards.

The scatter plot separates the number of distinct buyers from the median award value. Circle area adds the total published value, allowing a large monetary footprint to be distinguished from a large typical award. This is why adding the pounds to a count of tenders would conceal more than it explains.

In-tend’s 466 buyers give it the broadest observed reach, while ProContract’s 4,687 procedures make it the activity leader. Atamis carries £7.67bn across a different institutional footprint. Ariba’s smaller set of valued awards sits much higher on the median-value axis. The plot describes those market positions; it does not identify why buyers chose each system.

Two large awards worth opening before you quote them

Why the total is a published-value total, not an audited spend figure

One of the highest entries, LondonEnergy’s mobile-plant procurement, carries a £830.72m award alongside a £750,000 tender estimate and a separate later £850,700 award entry. The available fields do not establish how those entries should be reconciled. It would be misleading to promote the £830.72m as a confirmed purchase without that check.

The published-value total retains this unresolved entry, making it traceable rather than applying an unreviewed exclusion. That one entry represents 1.6% of the total. This uncertainty is another reason to use activity and sector relevance for a shortlist, and to inspect the source before using monetary ranks for a commercial decision.

SAP Ariba illustrates the effect of a small, high-value sample: 308 observed procedures lead to 76 valued award rows and £3.50bn of published value. That does not mean Ariba has a correspondingly large share of all UK opportunities. A few long-term utility awards can move the total substantially.

SME access matters to suppliers and to public buyers. Counting awards answers how often identified small and medium-sized businesses appear among winners. Counting value answers how much of the classified published value is attached to those awards. Neither is the probability that an SME submitting a bid will win: we do not have all losing bidders and their sizes.

On EU-Supply, SMEs account for 74.2% of classified attributable awards by count, but 21.8% of classified value. On Atamis, the comparable figures are 63.5% and 20.1%. This suggests a different distribution of award sizes. It does not establish discrimination, poor software or a deliberate barrier to small firms.

Across the ranked portals, £14.72bn is attributed to SME-classified winners and £28.47bn to large-classified winners. The binary value comparison covers 80.9% of the published-value total; the remainder includes grouped or unclassified winners. Treat that coverage as part of the result.

The procurement route changes the comparison too.

Below-threshold competitions, Open procedures and competitive flexible procedures serve different buying needs. Their observed winners differ. It is useful to examine the split, but the notice data does not prove that changing the procedure alone causes the observed SME share.

A new supplier can use this information to ask better questions about lot size, minimum capacity, turnover requirements and purchasing-system admission. It should not be used as an instruction to pursue every below-threshold notice or avoid a complex procurement that genuinely matches the firm’s strengths.

For competition, the lot is the commercial unit. A five-lot procurement with different services is not one five-bid observation. We use the number of tenders received for each eligible completed lot, keeping it apart from the number left at the final evaluation stage.

In-tend’s base ratio is 0.67×: its 8.3% observed single-bid share compares with 12.4% expected from matched peers. ProContract is at 0.75×. The stricter analysis adds buyer type and nation and uses a different matched sample; for In-tend its interval crosses 1. That change is a reason to distinguish a descriptive base result from a confident claim about the platform itself.

Sell2Wales illustrates why separate criteria matter. Its strong SME value share sits alongside a higher base single-bid ratio. Strong classified SME participation can coexist with thinner competition in a particular completed-lot sample. Combining those facts into an unqualified “best portal” label would hide an important trade-off.

The headline ratio needs the bid distribution beside it.

A single-bid rate focuses on one end of the distribution. It does not tell you whether the remaining lots usually receive two bids or seven. The next chart keeps that distinction visible, using received bids rather than the number left in a final evaluation stage.

The chart shows the recorded distribution for Open-procedure lots. It makes a practical point: the middle number of bids can differ markedly between channels. A supplier should then ask whether that reflects the type of work, the buyers, the region or incomplete reporting. A raw difference is not a portal’s ability to attract competition.

How many suppliers actually win work through each channel?

Competition is not only bids per lot. The supplier pool behind a channel tells another part of the story: how many distinct firms win, how often a winner is new to that buyer, and how many suppliers win exactly once in the window.

ProContract and In-tend carry the widest winner networks (2,678 and 2,246 distinct suppliers). At the other end, Litmus TMS concentrates a small catering panel with repeat wins, while most of SAP Ariba’s small UK sample are one-award suppliers. A wide pool is not a portal virtue claim: it follows from volume, sector mix and lot design.

Unlike some other jurisdictions’ auction systems, UK notices do not record a price reduction against a starting price. Where both a net estimate and a usable award value exist, the median award-to-estimate ratio sits close to 1.0 on every portal, so a “price saving” criterion would mostly measure estimate conventions rather than competition. We therefore leave it out and keep the bid-based measures above.

Northern Ireland illustrates the attribution problem.

In the measured cohort, 28.6% of Northern Ireland buyer lots received one bid, against 10.8% of Great Britain buyer lots. This covers 887 NI lots in 578 procedures from 80 buyers. Equal weighting of procedures gives 27.3% and 10.6%, respectively.

The pattern is visible across several channels used by NI buyers, not just eTendersNI. Removing insurance lots in a separate sensitivity calculation leaves 22.7% in NI and 10.9% in GB. That supports caution about blaming a single system. It does not identify a cause, and fewer bids do not by themselves establish worse outcomes or improper procurement.

The appendix goes further, matching lots on procedure, sector and value, then adding buyer type and nation. On the strict comparison, several intervals include the peer benchmark. That is an uncertainty result, not proof that the platforms are identical. For suppliers, the more immediate question is whether the actual lot fits their capacity and whether a credible bid is worth preparing.

Plan for time after submission, but do not call it software speed.

The interval from a deadline to the first recorded award activity is commercially useful context. In-tend’s middle interval is 50.4 days across 2,005 measured procedures; Atamis records 61.3 days across 1,053. These are completed observations, excluding non-positive intervals and intervals over a year. Work still waiting for an award is not represented.

The timestamp can reflect an award event or a publication fallback, rather than a clean measurement of evaluation work. Contract complexity, procedure choice, buyer practice and reporting all matter. Preparation and post-deadline medians also use different samples, so adding them does not produce a valid typical end-to-end timeline.

Publication completeness helps a reader assess the opportunity.

Publication can be compared through four observable fields: usable award values, net estimates, structured documents and tender-specific links. A complete monetary field lets a reader judge scale; a net estimate avoids treating a gross-only amount as directly comparable; linked documents and deep links reduce the distance from a notice to its underlying tender.

These are notice-field measures, not a test of the supplier experience. A missing structured document can coexist with a full pack behind a login. A root link can lead to the correct workspace after authentication. An HTTP string can redirect securely. Conversely, a populated field can contain an error, as the large-value examples demonstrate. Completeness and accuracy must not be confused.

The appendix publishes all 18 portals’ field coverage and sample counts. Buyers can use those differences to review their publication workflow, while suppliers can use them to understand what may require a second check in the pack. Neither conclusion requires claiming that the platform vendor caused the observed result.

The same edition, split into two equal periods.

The original study’s year-on-year comparison is useful because a large historical footprint need not describe the current direction of activity. The chart below restores the previous and latest 12-month counts for every portal. A percentage is easiest to interpret beside its starting count: a small channel can show rapid percentage growth without becoming one of the largest.

Panacea is a useful small-base example: 45 procedures in the previous period and 205 in the latest, a 355.6% rise. That is a material change for the channel, while its full-period volume remains much smaller than ProContract’s. The paired counts make both facts visible.

Changes in publication requirements, the mix of notices and the transition between procurement regimes can affect this comparison. A larger count does not, by itself, prove that organisations migrated to a vendor or that more money was spent. The chart preserves those questions rather than inventing an explanation for the movement.

Specialisation is a substantive result, not an afterthought.

Catering, medical equipment and IT services were useful distinctions in the original article. They belong alongside the national comparisons because they can change which portal matters to a reader. The matrix brings them back with construction, health/social work and environmental services, using complete division-level classifications.

Litmus TMS is #3 in catering and hospitality within this sample, with 65 of 468 procedures (13.9%). Atamis leads medical equipment with 210 of 584 (36.0%). In IT services, In-tend leads with 148 of 951 (15.6%). These narrower positions explain why volume alone misses useful specialist channels.

Each column has its own denominator across the same 18 portals. This lets a smaller specialist channel stand out without pretending it dominates all UK purchasing. The topical NHS, social-care and policy examples remain later in the article and use their own, narrower definitions.

Frameworks and dynamic markets: where flagged activity concentrates.

A tender notice is not always a new open competition. Two vehicle types change what a supplier is looking at: framework agreements, where buyers call off work from a pre-selected panel, and dynamic purchasing systems, known as dynamic markets under the Procurement Act 2023, where suppliers join an admitted pool over time and buyers run mini-competitions inside it. In this cohort 4,245 procedures on the 18 portals carry one of these flags.

Public Contracts Scotland carries the largest flagged share (29.6% of its procedures, overwhelmingly frameworks), while explicit dynamic-market notices remain rare everywhere: no portal records more than 60. For a supplier this is practical. A framework call-off may never appear as an open tender, and admission to a dynamic market is a separate registration decision, not a bid. Neither count says anything about software quality; both explain why watching notices alone can miss a channel’s real purchasing flow.

Portal brands, suppliers’ accounts and software are different layers.

A tender notice, a regional portal, a vendor’s software and a supplier account are different things. The central digital platform publishes notices and supplier information; a tender’s instructions can then take a supplier to a separate workspace. A family of familiar brands does not establish a common login or identical buyer configuration.

Among tracked buyers with at least five procedures, 188 of 1,056 (17.8%) appear across more than one hub in the study period. That may reflect migration over time, different workstreams or different purchasing arrangements. It is not proof that those buyers use multiple systems simultaneously.

The practical consequence follows from the evidence: a portal-level ranking is a view of distribution channels, while an individual organisation’s publication history remains a separate source of intelligence. Buyer counts overlap across portals and cannot be added together into a national total.

Nor do the notice data establish vendor market share, ownership concentration, supplier fees or abandonment rates. We have not audited every software relationship or run usability trials. Those questions require product and corporate evidence beyond the scope of this ranking. A common vendor alone would not prove a common bidding experience.

A league table becomes practical when it answers a narrower question. Where should a homecare provider register? Where do NHS organisations buy? Where do retrofit contractors need to look? The comparisons below use the same study window, but explicitly define each market rather than treating a topical keyword as a complete sector.

NHS: Atamis matters, and a health contract still needs its own route check.

Among procedures from budget holders whose canonical name includes “NHS” or “National Health Service”, Atamis appears on 1,796 of 2,386 procedures in the ranked portal sample (75.3%). That gives a health supplier a concrete reason to understand the Health Family system. NHS CPC describes Atamis as its procurement and contract-management system, with a supplier portal for participating organisations.

This is not a claim that Atamis handles that share of all NHS purchasing. The full cohort contains 21,934 procedures matching the buyer-name definition, many outside these 18 tracked channels. Nor is every health notice a new contest: in England, relevant healthcare-service procurements can follow the Provider Selection Regime. Goods, software and care services need the notice’s actual route checked individually.

Social care: follow local commissioning, not just the NHS brand.

Social care is a large and locally varied market. DHSC’s England finance return records £34.5bn of adult social-care expenditure in 2024/25. That is expenditure context, not a figure we add to this portal study. Our narrower discovery slice identifies procedures with a primary social-work classification, covering both adult and children’s services.

ProContract and In-tend lead the identifiable social-care slice with 292 and 244 procedures respectively. A care provider should then check geography, service-user group and access mechanism. A supported-living placement, a homecare competition and admission to a purchasing system can have very different commercial consequences, even when a portal lists them together.

The specific social-care classification is available only where the retained catalogue has sufficient four-digit detail. Across the 18 portals, that detail exists on 20,370 of 25,178 procedures (80.9%). Broader division-level sector coverage remains complete. The care count is an identifiable slice, not a census of all social-care purchasing.

Environment and net zero: look across waste services, construction and equipment.

Waste collection, cleaning and environmental services create a recurring procurement market. In-tend and ProContract are close in this sample, at 279 and 275 procedures in the environmental-services division. But a decarbonisation opportunity may sit under construction, heating equipment or engineering instead. An “environment” filter alone can miss the work you do.

For example, the retained records include 00350-AST Whole House Retrofit Installation Works 2026-2030 through Delta, and LCR Public Sector Decarbonisation Scheme Phase 4 through ProContract. These are examples found in titles, not a complete net-zero market estimate. Carbon Reduction Plan policy can also affect participation in relevant major central-government contracts; check the policy’s scope and the actual tender conditions.

Construction: broad reach still needs a local lens.

The construction division contains 4,210 procedures across the ranked portals. ProContract leads with 1,074, followed by In-tend with 774 and Sell2Wales with 446. For regional contractors, the devolved and local channels can therefore deserve attention alongside the largest national brands.

Do not use the value of a whole framework as your bid budget. Inspect the actual lot, contract period, geographic coverage and whether the notice admits suppliers to a future buying arrangement or awards a specific piece of work.

AI is a purchasing question, not a CPV sector of its own.

The IT and software divisions are a useful starting point: they contain 1,461 procedures across the 18 portals. In-tend, ProContract and Atamis lead that combined slice. It would be wrong to describe all of those purchases as AI.

Title-level examples illustrate the variety: Provision of Artificial Intelligence Schedule Risk Analysis Tool concerns project-schedule risk; Chest X-Ray Artificial Intelligence Solution and Radiology AI Platform concerns clinical imaging; Machine Learning Proof of Concept is a proof of concept. Some records concern training rather than buying an AI product. Read the specification before deciding that an opportunity fits your capability.

For central-government procurement teams, PPN 017 addresses transparency and proportionate controls around AI use. It is policy for its defined central-government audience, not evidence that every UK tender has an AI requirement. A supplier’s use of AI to prepare a bid is also a different question from a buyer procuring an AI system.

Asylum and resettlement: a real market, with very different services inside it.

Accommodation, resettlement support, clinical care, interpretation and case-management technology are distinct contracts. Combining them into a single “migrant spending” number would hide the service being purchased and the period it covers. The NAO’s May 2025 report examines seven regional accommodation-and-support contracts and a national support service. A June 2026 parliamentary report discusses a projected £15.3bn over the accommodation contracts’ ten-year term. That is external, multi-year context, separate from this study’s 24-month cohort.

Within our retained cohort, examples include HAMPSHIRE RESETTLEMENT SUPPORT SERVICE through In-tend and Supported Accommodation for Asylum Seeking Children (Age 16+) through ProContract. The 715903478 - Afghan Resettlement Programme (ARP) - Provision of Accommodation. Venue Find and Ancillary Services has no mapped link to one of the 18 ranked portals. This is a useful reminder to follow the buyer and the official publication as well as portal alerts.

A £674,627 service can look like £674m in a numeric field.

The notice for Enhanced Primary Care Services for Asylum Seekers illustrates why the original document matters. Its description states a 12-month value of £674,627.42; its structured monetary field states £674,627,042. The fields conflict. We cannot responsibly present the larger figure as verified expenditure, or silently rewrite the source to the smaller one.

This example sits outside the ranked-portal value total. It is included here as a documented data-quality warning, not an estimate of asylum-related costs. The practical lesson for a reader is to check the specification, the value basis and the service period before quoting a large number.

This edition was revised on 15 November 2026 after checking the displayed claims against the builder, retained observations and local catalogue. Percentage-unit errors are corrected. The Tenderline Portal Index v3 is defined in Chapter 2, with its weight sensitivity; the separate criterion comparisons remain in Chapter 1. No portal, procedure, competition lot or attributable award row was removed from the retained study populations.

What the ranking means

The fixed cutoff is 2026-11-14T18:19:02.095Z. The retained catalogue’s latest Find a Tender release is earlier than that cutoff; this is not a claim of live coverage through every minute of 14 November. A date filter cannot reconstruct old source state from a later compiled catalogue. Reproduction needs the retained edition inputs as well as the scripts.

The market cohort spans previous and current procurement regimes, devolved arrangements and relevant healthcare routes. The PA 2023 competition subset is narrower. Scottish activity remains in the market ranking without being forced into the PA 2023 competition comparison. Primary sector comes from the shared primary-only classification projection; additional codes do not add duplicate weight.

Primary sources and further reading

• Cabinet Office: the central digital platform ↗
• NHS CPC: using the Atamis Health Family system ↗
• DHSC: adult social care finance, England, 2024 to 2025 ↗
• NHS England: Provider Selection Regime guidance ↗
• Cabinet Office: Carbon Reduction Plans (PPN 006) ↗
• Cabinet Office: transparency of AI use in procurement (PPN 017) ↗
• National Audit Office: asylum accommodation contracts, May 2025 ↗
• Public Accounts Committee: analysis of the asylum system, June 2026 ↗

Policy references were checked on 15 November 2026. Their jurisdictions, audiences and dates differ. They supply context and practical guidance; they are not inputs to a portal quality score.