← All research

Landscape scan

Field / Daily-Log / Site-Diary landscape — the commoditised capture layer (Raken-class)

Verdict capture is table-stakes and racing to free; the money layer is empty here too — confirms the wedge sits above this category, not in it Threat low Beatability high Collected2026-06-16

Dashboard: Dashboard · benchmark incumbent: raken/dossier · market grid: _MARKET-PROBLEM-MAP · lens: _OPPORTUNITY-LENS · cross-competitor: _CROSS-COMPETITOR · format sibling: _micro-entrants/cohort

This is a scan, not a teardown, of the field/daily-report layer Raken anchors (area 4 in _MARKET-PROBLEM-MAP) — the most commoditised band in the whole market. Five tools were asked for (SiteLogs, EasySiteLog, BuildPass, SmartBarrel, Fonn); a sixth pattern — a long tail of AI voice-to-report micro-apps — is flagged at the end because it is the loudest signal in the category. The question for each is not “is it good” but does it stay at capture, or move up-funnel toward the money (cost / change / claims / benchmarking)? and what is its price floor, since our entry wedge is a free, narrow front door and the floor here tells us how cheap “capture” already is.

The headline, established before any single profile: this category is bifurcating. At the bottom, capture has collapsed to $0–19/month, free-tier-with-watermark, web/PWA, voice-in-60-seconds — table-stakes, not a wedge. At the top, the only tool here that has actually moved up-funnel and shipped real agentic AI is BuildPass (AI Agent + an MCP server that plugs into Claude), and even it stops at safety/quality/RFI workflows, not money recovery. None of the five generate an entitlement narrative or reuse historical cost. The lesson from _micro-entrants/cohort’s ConstructionDailyReport.ai holds and hardens: capture is the price of entry; value has moved up-funnel; and the up-funnel money step is still unoccupied even by the tools that have started climbing.

Per-tool mini-profiles

SiteLogs (com.construction.diary, Play-Store-first) — thin

A genuinely tiny generic “Construction Diary” Android app: 1K+ installs, rated 3+, ad-supported with in-app purchases, by an anonymous developer. It does photos-on-records, multiple-site management, daily logs for workers/machinery/materials, basic task management and export. There is no web office side, no API, no AI, no commercial layer, and no findable vendor company — the App Store “SiteLogs” search collapsed straight to Procore, underlining how little brand surface this has. This is a paper-replacement micro-utility for a solo builder, not a platform. Price floor: effectively $0 (ads + IAP). It is a data point on how far down the floor goes, not a competitor. Distance to our wedge: maximal — it never leaves capture.

EasySiteLog (easysitelog.com) — thin-but-clean; the pure-capture floor

A single-purpose web/PWA daily-log app and the cleanest illustration of the commoditised floor. Free plan ($0): unlimited daily logs, GPS + weather auto-fill, voice input, 5 photos. Pro ($19/month, 50% off first month): watermark-free PDFs, unlimited photos, email/CSV export. Pitch is “feels like texting, ~60-second log, works one-handed in gloves, offline.” No native app, no AI beyond basic speech-to-text, no API, and — pointedly — no claims/cost/commercial tier at all. It is a near-clone of ConstructionDailyReport.ai’s capture layer minus the up-funnel claims tier, and cheaper. Read: the bare metal of “capture is free.” Distance to our wedge: maximal; it is the thing we treat as baseline, not target.

BuildPass (buildpass.ai, AU/US) — the up-funnel mover; the one to actually watch

The most substantial and most AI-forward tool in the set. An all-in-one site-management platform (AU-origin, US-expanding) that starts at safety/compliance/site-diary and climbs up-funnel into scheduling, punch lists, equipment, timesheets, and — on higher tiers — RFIs, submittals, custom workflows, API access and white-labeling. Buyer is GCs/builders/PMs with 5+ staff (explicitly not sole traders). Pricing is per admin user only (subs/workers free), unlimited projects, roughly Lite ~$175–219/mo, Standard ~$439–549, Pro ~$879–1099 — a real platform floor, an order of magnitude above the capture micro-apps. Crucially it has shipped agentic AI, not slideware: a BuildPass Agent, an AI Toolkit (summarise site diaries, record/flag defects), and a BuildPass MCP server that integrates with Claude. That MCP move is ahead of Raken’s capture-only AI. But it still stops short of the money — its AI summarises diaries and flags defects; it does not assemble cost/time/causation or quantify a claim. Distance to our wedge: nearest of the five, still a clear gap. Read: watch it; it is executing the “capture → workflow platform” climb fastest and could bolt a commercial/claims agent on sooner than the others.

SmartBarrel (smartbarrel.io) — really a time/labour (area 5) hardware play, not a daily-log tool

Mis-bucketed as field/daily: SmartBarrel’s core is a rugged biometric time-clock — hardware (solar/LTE TimeClock 4.0) plus software — that captures verified labour hours via facial recognition, kills buddy-punching, assigns cost codes, and feeds payroll/ERP. Daily-report auto-fill is a by-product of the labour data, not the product. Buyer: GCs/specialty/union contractors managing field crews; deep payroll integrations (Procore, Sage, Foundation, QuickBooks, Viewpoint, etc.), public API. Price floor ~$150/month per job site, hardware included. Its AI is computer-vision identity verification — not LLM-shaped, and not our fight. It does touch commercial faintly (earned-vs-burned hours, “documentation for billing disputes”), but as labour-productivity, not entitlement. Distance to our wedge: far on the daily-log axis; it overlaps Raken on time/labour, not on the commercial gap. Read: a labour tool that happens to emit daily logs.

Fonn (fonn.com → The Access Group) — a flat-fee platform, already acquired

A Norwegian cloud construction-management platform (field + office): daily logs, photos, document control, RFIs, submittals, safety checks, one-click auto-populated daily reports, tasks/comms — sold to main contractors, subs, homebuilders, civils, M&E and commercial fit-out (notably the same UK mid-market segment our thesis targets). Differentiator was flat-fee, unlimited users/projects/subs (reported ~£517–786/mo for a small company; quote-led, US sources cite a wide $49–$4,000+ range). Fonn was acquired by The Access Group (UK ERP/business-software consolidator) and its site/pricing now redirect into Access’s funnel — so it is now a feature of a larger suite, with “AI and analytics” pitched as a separate Access category, not Fonn-native shipped AI. Reviews flag exactly the gap we care about: users ask for stronger RFI, budgeting and programme tooling. Distance to our wedge: a capable capture+coordination platform that is thin on cost/commercial and now absorbed — a second small proof (after Pype) of the consolidation/build-on-top risk, this time an ERP roll-up rather than a design-tools giant. Read: an acquired platform, weak on the money layer; flank, don’t fight.

The AI voice-to-report tail (flagged, not profiled) — the category’s loudest signal

Beyond the five, a 2026 web sweep surfaces a dense long tail of near-identical AI voice-to-report micro-apps — Speakwise, BuildLog, SendStatus, InspectMind AI, plus CompanyCam’s report feature and the cohort’s ConstructionDailyReport.ai / Voice Log Pro. They all do the same thing (narrate end-of-shift → structured AI daily report with photos/weather/crew), and several lean on the “tamper-evident / court-ready record for delay claims” framing (BuildLog, Voice Log Pro) — i.e. they gesture at area 15 as a positioning line but ship only the contemporaneous record, not the recovery workflow. This tail is the single clearest evidence that capture-with-AI is now a crowded commodity racing to free.

Comparison

ToolBuildsBuyerPrice floorAI angleMoves up-funnel?Read
SiteLogsGeneric mobile construction diary (Play-Store micro-app)Solo builders/contractors~$0 (ads + in-app purchase)NoneNo — pure captureThin / irrelevant
EasySiteLogWeb/PWA daily-log + PDFContractors, small teams$0 free / $19 mo ProBasic speech-to-text onlyNo — capture only, no claims tierThin-but-clean; the pure-capture floor
BuildPassAll-in-one site-management platformGCs/builders/PMs, 5+ staff~$175–219/mo (admin users; subs free)Shipped: AI Agent + AI Toolkit + MCP (Claude); summarises diaries, flags defectsYes — into RFIs/submittals/workflows/QA (stops before money)Nearest of five; the one to watch
SmartBarrelBiometric time-clock (hardware + software)GC/specialty/union, field crews~$150/mo per job site (hardware incl.)Computer-vision facial verification (not LLM)Partly — labour productivity / billing-dispute docs, not entitlementReally a time/labour tool (area 5)
Fonn (acq. Access Group)Field+office PM platform; daily logs, docs, RFIsMain contractors, subs, fit-out, civilsFlat fee, unlimited users (~£517+/mo)None native; “AI” now an Access suite categoryPartly — RFIs/submittals/docs, thin on cost/commercialAcquired platform; flank, don’t fight
AI voice-report tailVoice → structured AI daily reportForemen, subs, solo trades~$19–49/mo, free tiersVoice/NLP report generation; some “court-ready claim record” framingNo — positioning only; ships the record, not recoveryCommodity tail; baseline, not target

What this category tells us

Sources & method

Per-tool pulls (App Store search/app/reviews US+GB, Exa web search + answer) under raw_<tool>/raw/ in this folder, via _research/pull.py. Vendor pages fetched directly: buildpass.ai/us/pricing + /plans, fonn.com/pricing (301 → theaccessgroup.com — the acquisition signal), smartbarrel.io, easysitelog.com. App-store search noise was discarded honestly: the “SiteLogs” query resolved to Procore and “EasySiteLog” to Buildertrend/Vyce — non-matches, so SiteLogs detail is taken from its real Play-Store listing via Exa, EasySiteLog from its own site. The AI voice-report tail is from a 2026 web sweep (Speakwise/BuildLog/SendStatus/InspectMind), flagged not profiled. No absence-claim is made without a checked surface (per _RESEARCH-METHOD): “no AI / no API / capture-only” reflect the vendor pages read, and thinness (SiteLogs, EasySiteLog) is stated as observed, not assumed.