Zero Signals Fired Across 87 UK Platform-Engineering Jobs on Indeed in May 2026

We built indeed-hiring-intelligence to convert Indeed job listings into per-company hiring intelligence with five signal types (engineering-expansion, executive-hiring, sales-expansion, geo-expansion, compliance-buildout), TrustLayer extraction integrity, and outbound playbook routing. To stress-test the signal-detection floor on a narrow single-discipline cohort, we ran it across six Indeed UK queries (SRE, platform, devops, and infrastructure engineering in London and Manchester) on 14 May 2026. Across 87 verified job listings aggregated into 57 unique companies, zero of the five hiring-signal types fired — and 44 of those 57 companies had exactly one platform-engineering role open. That zero-signals, long-tail result is the data this post documents.

The problem: Every spring a UK tech publication runs the platform-engineering hiring boom thinkpiece off three LinkedIn screenshots and a quote from a recruiter. The framing assumes a concentrated vendor-evaluation cycle is visible somewhere in the data. When you actually pull Indeed UK across the four canonical platform-engineering disciplines on the same day with the same tooling, the picture isn't companies are buying SRE teams — it's 44 different companies are each posting one role. There's no middle of the leaderboard. There are two companies with seven roles open and forty-four companies with one. The buildout, if it exists at all, is happening on careers pages we cannot see, not on Indeed.

This post is a documentary audit of UK platform-engineering hiring as it appears on Indeed UK on 14 May 2026. Every count links back to the raw dataset shipped alongside this post. The 87-job cohort is real, the 44-singleton tail is real, the zero-signals-fired finding is real, and the framing UK tech press usually slaps on top is wrong.

What is a hiring signal? A pattern in a company's open-role activity that exceeds a calibrated threshold and indicates a strategic posture change. The indeed-hiring-intelligence Apify actor detects five types: engineering-expansion (5+ new engineering roles in 30 days with rising eng-share), executive-hiring (senior leadership posting), sales-expansion (sustained sales-side hiring), geo-expansion (multi-city ramp), and compliance-buildout (legal/risk/compliance cluster).

Why it matters: Hiring signals are the leading indicator that something has changed at a company — new product launch, fundraise, GTM motion, regulatory response. Vendors selling into a target use signals to time outbound. According to the Indeed Hiring Lab 2026 UK Jobs report, UK hiring in 2026 is characterised by caution with bright spots, which makes concentration-pattern detection harder than in broad-expansion periods.

Use it when: You're benchmarking competitor hiring posture, building a sales target list from inferred buying intent, watching a specific discipline (platform engineering, AI/ML, compliance) across a market, or tracking month-over-month shifts in a named cohort.

Key findings

• 87 platform-engineering jobs across 57 unique companies on Indeed UK, captured 14 May 2026. Six queries: SRE, platform, devops, infrastructure engineering in London and Manchester. 87 charged jobs at $0.005 each plus 13 charged companies at $0.05 each — $1.08 total platform fees.
• Zero of the five hiring-signal types fired. No company in the cohort reached the engineering-expansion threshold (5+ new engineering roles in 30 days). Executive-hiring, sales-expansion, geo-expansion, and compliance-buildout signals are out-of-frame for a narrow single-discipline cohort by construction.
• 44 of 57 companies had exactly one role open. That's 77% of the cohort sitting in the singleton tail. Only 13 of 57 companies (23%) hit the actor's company-intelligence chargeable threshold of openRolesInRun >= 2.
• Top of the chargeable cohort: 2 companies with 7 roles each. CGI and Amazon Web Services posted 7 open platform-engineering roles each in the 14-day window. Both classified as P4-QUIET (momentum score below 35) because the actor's momentum scoring penalises score-velocity-of-zero for cross-run-stateless companies.
• The highest-paid verified role in the cohort: Cloudflare's London Senior SWE Storage Infrastructure at £185,000-£254,000. Next tier: Fyxer's two Product Reliability Engineer roles in London — Senior at £100,000-£120,000, Lead at £120,000-£170,000. Then Obsidian Security's Site Reliability Engineer roles in Manchester at £85,000-£103,000.
• At least 8 of 57 "companies" are recruitment agencies reposting on behalf of unnamed employers — PlanetRecruitment, RifkiTalentScout, Corriculo, SW5 Consulting, ABACUS, Info Plus, Carrington West, and Mindera. The true direct-hirer count is closer to 49.
• The 14-day filter behaved as a no-op because Indeed's relative date strings (postedAt) returned null for all 87 records in this build. Cohort is "everything Indeed UK returned for the six queries on 14 May 2026," approximately but not exactly a 14-day window.

In this article: The 13-company chargeable cohort · Singleton tail · Story A: chargeable vs singleton · Story B: zero signals fired · Story C: Cloudflare salary outlier · Story D: who is absent · Story E: recruiter pollution · Counter-narrative · Methodology · Caveats · Press lift-out

The 13-company chargeable cohort — platform engineering on Indeed UK

This is the cohort of companies that hit openRolesInRun >= 2 — the threshold at which the indeed-hiring-intelligence actor emits a chargeable company-intelligence record (per-company aggregation with roleMix, hiringMomentum tier, salaryBand, hiringProfile, lifecycleStage, recommendedAction). Sorted by openRolesInRun descending, then by hiringMomentum.score descending. Salary medians are flagged where they're not market-representative.

Salary aggregation note for the OKTO row: the £175,000 GBP median is an annualisation of contractor day rates (£575-£700 per day) and is not directly comparable to permanent-employee salaries elsewhere in the cohort. See caveats for the full rule on small-sample salary aggregation.

Underlying source: six Indeed UK queries (site reliability engineer London, platform engineer London, devops engineer London, infrastructure engineer London, site reliability engineer Manchester, platform engineer Manchester), captured 14 May 2026 via the indeed-hiring-intelligence Apify actor with country: UK, postedWithinDays: 14, outputTier: intelligence. The leaderboard CSV is downloadable from the Produced by attribution banner above; the structured payload sits in raw.json alongside it.

The shape of this 13-company chargeable cohort is more telling than the rank order. The top two (CGI and AWS) are both P4-QUIET despite having 7 roles each open. That's because the actor's momentum scoring uses scoreVelocity as one of its inputs, and for a single-run baseline with no prior snapshot, every company's scoreVelocity is +0 by definition. Re-running the audit a month from now would give the momentum tier its first cross-run signal. The current snapshot is the baseline.

The 44-company singleton tail

44 companies in the cohort had exactly one open platform-engineering role on Indeed UK on 14 May 2026. Below the chargeable threshold, the actor emits per-company records but doesn't charge the company-intelligence event — they're cohort-shape data, not paid intelligence. The full singleton tail by city and by lifecycle stage:

Two patterns stand out. First, the singleton tail contains the cohort's salary outlier — Cloudflare's Senior SWE Storage Infrastructure role at £185,000-£254,000 GBP yearly. The premium-tier role sits in a low-volume-hirer slot, not a high-volume-hirer slot, which is exactly the inverted shape someone reading the leaderboard top-down would miss. Second, the singleton tail is where most of the recruiter-agency pollution lives — Corriculo, RifkiTalentScout, ABACUS, Carrington West, Info Plus, SW5 Consulting, and Mindera all appear in this cohort and all are agencies reposting on behalf of unnamed direct hirers.

Story A — The 13-company chargeable cohort vs the 44-company singleton tail

The cohort distribution by openRolesInRun:

51% of the cohort's open roles sit at companies that posted exactly one platform-engineering job in the 14-day window. The 13-company chargeable cohort accounts for 43 of 87 roles (49%). There are as many singleton-hirers (44) as there are roles total at companies with 2+ openings (43). The dispersion ratio is roughly 1:1.

That dispersion shape is the story. A concentrated UK platform-engineering buildout — the trade-press framing — would show up as a small number of companies with 10+ roles open each, anchoring 60-70% of the cohort. Instead the cohort is a long-tail with no per-company concentration. CGI and AWS sit at the top with 7 roles each. Those 14 roles are 16% of the cohort total. The remaining 84% is spread across 55 other companies, most of them singleton hirers.

Story B — The zero-signals finding

The indeed-hiring-intelligence Apify actor detects five hiring-signal types. None fired across the 57-company cohort on 14 May 2026.

The mechanism: each signal type has a deterministic threshold. The engineering-expansion signal requires newEngRolesIn30d >= 5 at a single company plus rising engineering share in the role-mix. No company in this cohort has 5 engineering roles open — the maximum is 7, at CGI and AWS, and the rest of the cohort tops out at 4. Even if both 7-role companies had every role flagged as net-new in the prior 30 days, neither would clear the rising-share component because their roleMix is 100% engineering already (there's no other function to rise against).

The executive-hiring, sales-expansion, geo-expansion, and compliance-buildout signals are structurally out-of-frame for this audit by query design. The six Indeed UK queries target four engineering disciplines (SRE, platform, devops, infrastructure). They do not query exec-level titles, sales reqs, or compliance roles. The actor sees what was queried, so signals that depend on cross-function evidence cannot trip from a single-discipline cohort run.

The honest read: zero signals fired is the actor's correct behaviour for this cohort, not a bug. The signal-detection thresholds are calibrated for broad-function cohort audits — e.g. all functions across a curated 50-company target list, captured across a 30-day window with cross-run state. A narrow single-discipline 14-day single-day Indeed audit cannot reach those thresholds by construction. What it can produce is the leaderboard above, a methodology disclosure, and a counter-narrative against the "buildout cycle" framing.

The "first SRE role detected = vendor evaluation cycle starting" narrative in the actor's README is a forward-looking detector that requires cross-run state. Build 1.0 has the KV plumbing but no prior baseline. A second audit run a month from now is the first point at which cross-run signals can fire. This audit is the baseline.

Story C — Cloudflare anchors the cohort's premium-tier salary outlier

Cloudflare appears in the cohort with one open role: Senior Software Engineer, Storage Infrastructure, in London. The verified salary range on the listing is £185,000-£254,000 GBP yearly. That's the highest salary band in the 87-job cohort by a clean margin.

The next-tier verified salary bands, ranked descending:

The distribution is bimodal. There's a tight band of £55,000-£75,000 for mid-market platform-engineering roles, then a thin top end of £100,000+ for senior infrastructure roles at FAANG-tier vendors (Cloudflare) and well-funded scale-ups (Fyxer, Obsidian Security). There's very little in the £75,000-£100,000 band — the most common ladder rung in US-tech salary distributions is structurally underrepresented in this UK cohort, plausibly because UK total-comp packaging weights more toward equity and benefits than gross salary disclosure.

Single-role salary bands should not be read as market signals. Cloudflare's salaryBand.median of £254,000 in the raw payload is the max-end of a single listing's range, not a true median across a sample. Same caveat for OKTO's £175,000 — that's the annualisation of contractor day rates (£575-£700 per day), not a permanent salary. Use the underlying job-level strings (£185,000-£254,000 yearly for Cloudflare; £575-£700/day for OKTO) for any quotable headline number, never the aggregated single-role median.

Story D — Who is absent from the cohort

The UK tech press talks about UK fintech hiring up 37% Y/Y (Morgan McKinley December 2025 fintech outlook) and platform engineering as a 2026 buildout focus, while CNN's April 2026 piece argues "the demise of software engineering jobs has been greatly exaggerated" and that Indeed software-engineer listings are up 11% Y/Y. None of the named UK tech standouts surface in this 14-day Indeed UK platform-engineering window.

Companies not in the cohort that you'd expect to see if those market-level narratives held at the discipline + aggregator + window level:

• UK fintech / neobank cohort: no Monzo, no Revolut, no Wise, no Octopus Energy, no Starling, no Tide, no Zopa.
• UK food / mobility scale-ups: no Deliveroo, no Bolt, no Gett.
• AI scale-up London offices: no Anthropic London, no DeepMind, no OpenAI London, no Bytedance London, no Mistral.
• US-tech London engineering offices: no Stripe London, no Datadog London, no Snowflake London, no Databricks London. AWS appears with 7 roles but the rest of FAANG-tier is absent.
• UK-listed tech: no Wise PLC, no Auto Trader, no Rightmove, no Just Eat, no THG.

There are three plausible explanations. First, these companies route platform-engineering hiring through careers-page ATS hosts (Greenhouse, Ashby, Lever, Workday, iCIMS, Phenom, Eightfold) and don't repost to Indeed at all. Second, they're not posting in this specific 14-day window — hiring is bursty and a Tuesday capture differs materially from a Thursday capture. Third, they're posting under different job titles that fall outside the four queried disciplines (e.g. "Cloud Engineer," "Software Engineer, Infrastructure," "Backend Engineer, Reliability").

All three explanations are non-falsifiable from a single Indeed audit. What the cohort tells us is what Indeed UK shows for the four queried disciplines on one day. Methodology disclosure is the audit's quality bar, not its blind spot — the same framing the SaaS hiring mix engineering vs sales 2026 sibling audit applied to its 14 excluded companies. An audit that can defend its exclusions is an audit that can be quoted. An audit that silently drops the gaps is one journalists rightly ignore.

Story E — Recruiter-agency pollution

Eight of 57 "companies" in the cohort are recruitment agencies reposting platform-engineering jobs on behalf of unnamed end clients:

• PlanetRecruitment (Windsor, £50,000 median, 2 open)
• RifkiTalentScout (Hybrid London, salary not disclosed, 1 open)
• Corriculo (London, £75,000, 1 open)
• SW5 Consulting (Hybrid Manchester M3, £80,000, 1 open)
• ABACUS (London, salary not disclosed, 1 open)
• Info Plus (Manchester, salary not disclosed, 1 open)
• Carrington West (Kingston upon Thames KT1, £60,000, 1 open)
• Mindera (Hybrid London — IT services consultancy, may also be reposting third-party engagements, 1 open)

Add Capgemini (Hybrid London, 1 open — IT consultancy that frequently staffs third-party engagements rather than direct-hires) and the agency-or-consultancy slice grows to 9 of 57. The true direct-hirer count in the cohort is therefore closer to 48-49 than 57.

This is a known limitation of any Indeed-sourced hiring audit. Agencies and IT consultancies use Indeed as a primary channel because their cost per direct hire is lower than for ATS-hosted careers pages. Direct hirers with mature recruiting orgs use Indeed selectively or not at all. The cohort overweights the agency channel by construction. The audit names every flagged agency so coverage can adjust the direct-hirer count without re-running the data.

Cross-aggregation: momentum tiers across the 57-company cohort

Aggregate by hiringMomentum.tier, verified by direct re-count of raw.json:

The cohort skews toward P4-QUIET: 40 of 57 companies (70%) sit at momentum scores below 35. P3-STEADY accounts for 17 companies — most of them are P3 because the actor's scoring rewards senior-bias roleMix (which dominates the platform-engineering discipline by job-title norms), not because they're hiring at scale. Zero companies reach P2-EXPANDING (score 55-74) or P1-HIGH-GROWTH (75+). Those tiers are calibrated for companies with multiple discipline lines, cross-run momentum data, and meaningful scoreVelocity — none of which a single-day single-discipline audit can produce.

Year-over-year context

None available. This is the first published Indeed-driven hiring audit, and the cross-run KV state plumbing in indeed-hiring-intelligence exists but has no prior baseline. A 2027 re-run against the same six queries would give the first cross-run snapshot. For now, the audit is a single point-in-time capture and should be quoted as such — the May 2026 cohort not the 2026 trend.

The closest external context: the Indeed Hiring Lab 2026 UK Jobs report projects UK 2026 hiring as caution-with-bright-spots. Morgan McKinley's December 2025 London fintech rebound outlook projects a 37% Y/Y increase in London fintech vacancies. Neither of those market-level claims is contradicted by this cohort. Neither is visible in this cohort either, because the cohort is six narrow discipline queries on Indeed, not a fintech-cohort run across all functions.

What UK tech press gets wrong about platform-engineering hiring

The trade-press framing for May 2026 — UK platform-engineering boom, SRE buildout cycle, vendor evaluation timing window opening — assumes the buildout is visible somewhere. The cohort shape says otherwise. Here's what coverage typically gets wrong, with corrections:

1. "The buildout is concentrated in named companies." The cohort top is CGI (7 roles, P4-QUIET) and AWS (7 roles, P4-QUIET). Both classify as quiet because neither has cross-run momentum data and both have 100% engineering roleMix (no other function to rise against). Concentration in the trade-press sense — we're seeing companies hire 10+ SREs at once — is not present in this cohort at all.
2. "Engineering hiring is rising because of AI tooling demand." This cohort has zero AI-specific roles in the SRE/platform/devops/infrastructure window. AI/ML engineering is a different discipline cut and would need its own audit. The audit does not support or refute AI-driven demand claims; it scopes outside that discipline.
3. "Platform engineering hiring is up Y/Y." No Y/Y comparison exists in this dataset. There is no prior snapshot. Coverage citing this audit cannot make Y/Y claims without inventing a baseline.
4. "London concentrates UK platform-engineering hiring." Using Greater London / Greater Manchester definitions, 36 of 57 cohort companies (63%) have a London geo, 13 (23%) are Manchester, and 8 (14%) are elsewhere (Welwyn Garden City, Reigate, Windsor, Woking, Burnley, Skelmersdale). The London share is high but not overwhelming. More than a third of the cohort sits outside Greater London, which is more dispersion than the trade-press framing typically allows.
5. "Indeed shows the UK tech hiring market." Indeed shows a slice — the slice that posts to Indeed. Major UK tech employers route platform-engineering hiring through ATS-hosted careers pages (Greenhouse, Ashby, Lever) or their own careers domains. Any audit on a single aggregator surface is a slice. Coverage should frame as the Indeed UK slice not the UK market.

Methodology

• Tool: indeed-hiring-intelligence Apify actor, build version 1.0. Converts Indeed job listings into per-company hiring intelligence with five signal types, TrustLayer extraction integrity, and outbound playbook routing. Pay-per-event pricing: $0.005 per job, $0.05 per company-intelligence record, $0.02 per signal. This run charged $1.08 platform fees (87 jobs at $0.005 + 13 chargeable companies at $0.05 + 0 signals).
• Sample queried: six Indeed UK queries — site reliability engineer London, platform engineer London, devops engineer London, infrastructure engineer London, site reliability engineer Manchester, platform engineer Manchester. Captured 14 May 2026, run ID k805otp4nzDv70yR2.
• Date range: postedWithinDays: 14 requested. Effective filter: Indeed's own search-results recency default, because postedAt parsing returned null for all 87 records in this build. No records were dropped by the date filter (countryFilterDropped: 0, dateFilterDropped: 0, salaryFilterDropped: 0).
• Threshold: maxResultsPerQuery: 100 (none binding; Indeed's natural result-ceiling reached at 7-16 jobs per query). outputTier: intelligence (jobs + companies + signals + playbooks). country: UK verified at parse time.
• Field filters: no role-title filtering applied beyond the four discipline queries. No salary floor applied. No seniority filter applied.
• Total records captured: 87 jobs, 57 unique company names, 0 signals, 5 outbound playbook records.
• Aggregation rule: per-company aggregation via the actor's groupByCompany function, deduplicating by raw company-name string. Note: known weakness — company-name normalisation drift means "MI6" and "Secret Intelligence Service - Mi6" emit as separate records, and "Copper" and "Copper.co" likewise. After merging these duplicates, the true company count is closer to 55 than 57.
• Known gaps named explicitly:
  • postedAt date parsing produced null for every record; the 14-day filter was effectively a no-op
  • Signal-detection thresholds did not fire for any of the five signal types; none reached the 0.6 signal-confidence floor
  • Cross-run state (firstSeenAt, daysOpen, repostDetected, scoreVelocity) had no prior baseline; scoreVelocity is always +0 in this run
  • The "first SRE role detected" narrative in the actor README is a forward-looking cross-run detector; build 1.0 has the KV plumbing but no prior snapshot to compare against
• Ticker-noise / out-of-sample disclosure: Indeed UK's search query for "site reliability engineer" surfaces some adjacent-discipline titles (Reliability Engineer, Test Reliability Engineer, Production Reliability Engineer). All are treated as in-cohort because they sit within the platform-engineering discipline cluster. Out-of-discipline noise (e.g. mechanical reliability engineering, hardware reliability engineering) is rare on London/Manchester surfaces but not zero — see caveats.
• Cross-reference: the Indeed Hiring Lab UK 2026 trends report provides independent UK market context. Morgan McKinley's December 2025 London fintech hiring outlook and CNN Business's April 2026 software-engineering jobs piece provide adjacent narrative context. Neither contradicts this cohort, neither is reproduced in it.

Caveats and what this data does not say

• 14-day filter was effectively a no-op. The actor's postedAt parser returned null for all 87 records in this build, so postedWithinDays: 14 did not drop any records. The cohort is "everything Indeed UK returned for the six queries on 14 May 2026," not strictly "everything posted in the prior 14 days." Indeed's own algorithmic search-results filter does narrow to recent postings by default, so the result is approximately but not exactly a 14-day window.
• Indeed UK only. Glassdoor, LinkedIn, careers pages with ATS-hosted listings (Greenhouse, Ashby, Lever, Workday, iCIMS, Phenom, Eightfold), Reed, CV-Library, Otta, and direct careers-page feeds are not included. Companies that route their platform-engineering hiring through any of those surfaces will not appear in this cohort.
• Single-day capture, not a rolling window. Hiring activity is bursty. A Tuesday capture may differ materially from a Thursday capture. A 14-day rolling average would smooth this; a single-day capture does not.
• Company-name normalisation drift. "MI6" and "Secret Intelligence Service - Mi6" emit as separate rows; "Copper" and "Copper.co" likewise. The true company count is closer to 55 than 57 after merging duplicates. This is disclosed in raw.json knownCaveats[].
• Recruitment-agency dilution. Eight of 57 "companies" are recruitment agencies reposting on behalf of unnamed clients, not the underlying hirers. True direct-hirer count is closer to 48-49. Coverage citing the leaderboard should adjust the direct-hirer count accordingly.
• Salary aggregation noise on small samples. salaryBand.median for Cloudflare (£254,000) is the max-end of a single £185,000-£254,000 yearly range, not a true median. salaryBand.median for OKTO Integrated Services (£175,000) is an annualisation of contractor day rates (£575-£700/day), which is not a directly comparable permanent-salary figure. Cite the raw job-level strings (£185,000-£254,000 yearly for Cloudflare; £575-£700/day for OKTO) for any headline number, never the aggregated median for single-job companies.
• Zero signals fired is the actor's correct behaviour for this cohort, not a bug. The signal-detection thresholds are calibrated for broad-function cohort audits. The post explains the mechanism honestly — what the signal floor requires, why a single-discipline 14-day audit cannot reach it, and what kind of audit would reach it (a 30-day audit across engineering + sales + product + ops at a curated 50-company target list, with cross-run state for at least one prior snapshot).
• No headcount, no Y/Y, no trend. The audit measures one thing: open platform-engineering roles on Indeed UK on a single day. It does not measure headcount, year-over-year change, hiring velocity over time, attrition, or any of the broader workforce metrics trade-press coverage often blurs with open-role data. Coverage should frame as the May 2026 snapshot, not as the trend.

Press lift-out for journalists

A 2026 ApifyForge analysis of Indeed UK platform-engineering job listings across London and Manchester on 14 May 2026 found 87 verified roles aggregated into 57 unique companies. None of the actor's five hiring-signal types fired across the cohort. 44 of 57 companies (77%) had exactly one platform-engineering role open in the 14-day window. The two highest-volume hirers in the cohort were CGI and Amazon Web Services with 7 open roles each; the next tier sat at 4 roles (Secret Intelligence Service - MI6, OKTO Integrated Services, Speechify). The premium-tier salary outlier was Cloudflare's London Senior Software Engineer, Storage Infrastructure role at £185,000-£254,000 GBP yearly, posted as a singleton. At least 8 of 57 'companies' were recruitment agencies reposting on behalf of unnamed direct hirers, putting the true direct-hirer count closer to 49. The audit covers Indeed UK only and is a single point-in-time capture; companies posting platform-engineering roles via ATS-hosted careers pages (Greenhouse, Ashby, Lever, Workday) are out of scope.

Source for citation: ApifyForge, "Zero Signals Fired Across 87 UK Platform-Engineering Jobs on Indeed in May 2026," 14 May 2026. Underlying data: six Indeed UK queries (SRE, platform, devops, infrastructure engineering across London and Manchester) captured via the indeed-hiring-intelligence Apify actor (build 1.0). Sibling audit: SaaS hiring mix engineering vs sales 2026 for the ATS-hosted careers-page surface.

Embeddable visuals

Three chart blocks, all reproducible from the tables in this post.

Chart 1 — Cohort distribution by openRolesInRun

Horizontal bar chart. Y-axis: cohort buckets (7 roles, 4 roles, 3 roles, 2 roles, 1 role). X-axis: company count, with role-count overlaid as a secondary lighter-weight bar. The shape is the visual story: the 1-role bucket is by far the longest bar (44 companies), and every other bucket sits at 7 or fewer. Headline: "77% of cohort sit in the singleton tail. The platform-engineering buildout, if it exists, is happening elsewhere." Source line: "Indeed UK platform-engineering audit, six queries across London and Manchester, captured 14 May 2026 via the indeed-hiring-intelligence actor."

Chart 2 — The premium-tier salary outlier

Horizontal bar chart of verified salary medians for the cohort companies that disclosed salaries. Cloudflare anchors the top at £254,000 (with the lower £185,000 bound marked as a min-end whisker). Fyxer at £120,000. Obsidian Security at £103,000. Then a cluster of mid-market roles at £75,000-£80,000 (StableLogic, DevNull Security, SW5 Consulting). Then a public-sector cluster at £60,000-£65,000 (MI6, NWAS NHS Trust, Carrington West). Headline: "UK platform-engineering pay is bimodal in May 2026. A £100k+ premium tier sits above a mid-market £55k-£75k band, with very little in between." Source line: "Indeed UK platform-engineering audit, May 2026. Salaries are listing-disclosed yearly bands where stated; OKTO Integrated Services excluded because its band is annualised from day rates."

Chart 3 — Momentum tier distribution

Two-tier stacked bar chart. Single bar split into P3-STEADY (17 companies, 27 roles) and P4-QUIET (40 companies, 60 roles). P1-HIGH-GROWTH and P2-EXPANDING tiers are visible as zero-height labelled gaps above the bar, so readers see the absence of high-momentum tiers. Headline: "Zero companies in the UK platform-engineering Indeed cohort cleared the P2-EXPANDING threshold in May 2026. The signal-detection floor was calibrated for broader-function cohort audits; this audit is the baseline." Source line: "Indeed UK platform-engineering audit, May 2026. Momentum tier computed by the indeed-hiring-intelligence actor."

Frequently asked questions

What did the May 2026 UK platform-engineering Indeed audit find?

87 platform-engineering jobs across 57 unique companies, captured from Indeed UK on 14 May 2026 via six queries (SRE, platform, devops, infrastructure engineering across London and Manchester). 44 of 57 companies (77%) had exactly one role open. The top two hirers were CGI and AWS with 7 roles each, both classified P4-QUIET because the actor's momentum scoring lacks cross-run baseline data. The premium-tier salary outlier was Cloudflare at £185,000-£254,000 in London. None of the actor's five hiring signals fired.

Why did zero hiring signals fire in this cohort?

The indeed-hiring-intelligence actor's five signal types are calibrated for broad-function cohort audits, not single-discipline windows. The engineering-expansion signal requires 5+ new engineering roles in 30 days at a single company plus rising eng-share — no company in this cohort has 5 engineering roles open, and the cohort's roleMix is 100% engineering by query design (nothing to rise against). The executive-hiring, sales-expansion, geo-expansion, and compliance-buildout signals are out-of-frame because the six queries don't include those disciplines. Zero signals is the actor's correct behaviour for this cohort.

Is this audit evidence that UK platform engineering hiring is slowing?

No. The audit is a single-day single-aggregator capture and cannot support hiring-trend claims. It does not measure year-over-year change, hiring velocity, or absolute UK-market volume. What it does show is that on Indeed UK on 14 May 2026, platform-engineering hiring is dispersed across many companies with few roles each, not concentrated at a small number of high-volume hirers. That dispersion shape contradicts the "buildout cycle" framing of UK tech press, but it does not establish a slowdown.

Why aren't Monzo, Revolut, Stripe London, or other named UK tech employers in the cohort?

Three plausible reasons. First, those companies route platform-engineering hiring through ATS-hosted careers pages (Greenhouse, Ashby, Lever, Workday) and don't repost to Indeed. Second, they may not have posted in this specific 14-day window — hiring is bursty. Third, they may be posting under job titles that fall outside the four queried disciplines (e.g. "Cloud Engineer," "Backend Engineer, Reliability"). A single Indeed audit cannot distinguish between these three explanations. The sibling SaaS hiring mix audit covers the ATS-hosted careers-page surface for a different cohort.

How does this audit compare to the SaaS hiring mix audit?

Both audits use ApifyForge actors against public hiring data. The SaaS hiring mix engineering vs sales 2026 audit covered 50 major SaaS companies' careers-page extraction on 9 May 2026, surfaced reliable engineering-vs-sales data for 10 companies, and found a clean bimodal split between sales-heavy (Notion, Figma, Linear) and engineering-heavy (Supabase, Vercel, Webflow). This audit covers Indeed UK platform-engineering on 14 May 2026, surfaced 57 companies, and found a long-tail with no per-company concentration. Different surface, different cohort, different shape. Together they bracket the what's visible on careers pages vs what's visible on aggregators question.

Where can I download the underlying data myself?

The dataset CSV (57 rows, company-level aggregation) and the structured raw.json (with full per-company records, knownCaveats, and queryParams) are downloadable from the Produced by attribution banner at the top of this page. To reproduce the audit, run the indeed-hiring-intelligence Apify actor with the same six queries, country: UK, postedWithinDays: 14, maxResultsPerQuery: 100, outputTier: intelligence. Anyone with an Apify account can re-run it. Note that the cohort shape will drift day-over-day because Indeed's search results are time-sensitive.

Is Indeed UK considered authoritative for tech hiring journalism?

Indeed UK is one of the largest job-listing aggregators in the UK and is broadly accepted as a primary-source surface for hiring journalism, with the caveat that its coverage is biased toward employers and recruitment agencies that pay for placement. The Indeed Hiring Lab publishes regular market-level analysis using the same dataset and is cited widely. For audits like this one, Indeed UK is a defensible single-aggregator slice — but it is a slice, not the whole UK market. Coverage citing this audit should frame it as the Indeed UK slice rather than a UK-tech-market verdict.

Ryan Clinton publishes Apify actors and MCP servers as ryanclinton and builds developer tools at ApifyForge. The leaderboard above was produced via the indeed-hiring-intelligence actor across six Indeed UK queries against London and Manchester on 14 May 2026; the methodology, analysis, and framing are independent of any product positioning.

Last updated: May 2026