Memo 01
CivilGrid — the map is a commodity; the planning desk might not be
The map is a commodity. The planning desk might not be.
1. Thesis
CivilGrid gathers records of what is buried under US streets from the asset owners themselves, and puts them on one map that engineers use before they design P2P18. The press frames it as "Google Maps for the underground" P2. I think the map itself is turning into a commodity. The more interesting product is the planning layer on top: deciding which of a utility's hundreds of projects collide with a fresh repaving job, a moratorium or another owner's dig P21R2. That layer sits right next to a record utility capex cycle M12. Whether CivilGrid becomes a company or stays a feature depends on whether a second large utility shows what PG&E has shown.
Confidence: 5/10. All of the customer-side evidence is n=1: PG&E, the founder's former employer R1F12.
2. Variant perception
- Consensus thinks: This is a mapping race about preventing strikes, the accidents where an excavator hits a buried line. The "Google Maps of the subsurface" line belongs to three companies at once: CivilGrid, 4M Analytics and Exodigo P38M25M26. Most of the capital has gone to physical sensing, where Exodigo has raised $214M M26inference.
- I think: Strikes are the wrong frame for a records platform. In CGA's 2024 data, only 2.16% of damages trace to an incorrect facility record or map. Failing to call 811 (24.54%) and locator errors (11.94% + 8.58%) dominate R4. A desk-side records tool can't do much about those. It can change what happens months earlier, in capital planning. PG&E's own headline number is about paving conflicts across a portfolio of planned projects, not about strikes P5R2.
- Why I might be right: The best evidence comes from the customer's own newsroom, and it is denominated in planning money: "$60M+ in paving conflicts eliminated through strategic portfolio optimization", "40% reduction in field research and survey activities" P5. CivilGrid's product has moved the same way. Its site went from "Construction Intelligence Platform" in 2022 P35 to a "civil design platform" with Plan and Design modules in 2026 P14. The Plan module covers paving moratoria, "Dig Once" bundling and ranking a portfolio against budget P21. CivilGrid says PG&E runs 300+ projects a year through it P21company claim. inference The map gets people in the door. The planning desk is what keeps them.
3. The company in one paragraph
CivilGrid was founded in 2020 in San Francisco F1. It requests, digitizes and standardizes utility, right-of-way, permit and environmental records directly from asset owners, and keeps a chain of custody for each record that it describes as ASCE 38-22 compliant P18P19. Its FAQ says it does not infer utility locations from satellite imagery, scans or AI models P18. Buyers are utilities, civil engineering firms, DOTs and municipalities P2P14. Pricing is not public, and the motion is demo-first enterprise sales P17. On funding, it disclosed a $2.2M pre-seed raised over 15 months to July 2022 F13 and a $26M Series A led by Spark Capital on 2026-08-27, with Afore, A, Ford Street Ventures, SNR and Energy Impact Partners P1F5. Cumulative funding is not verifiable*: aggregators give figures from ~$25M to $51.2M, and no Form D was found under the company name R5R6R7. Headcount is unresolved; Built In says 26 P27, and Simplify says 51–200 with −14% over two years P28. Revenue and customer count: not in public sources.
4. Why now
- Utility capex is at a record. Investor-owned electric utilities were on pace for ~$208B of capex in 2025, the "highest amount ever", with >$1.1T planned for 2025–2029. EEI cites AI data-center demand as a key driver M12. PG&E alone plans $73B through 2030 M13.
- Damage costs were re-priced. CGA's August 2026 study puts the annual US economic cost of buried-utility damage at $83.2B, about 3x its 2019 benchmark of $30B M8.
- The data standard exists. ASCE 38-22 and 75-22 (July 2022) define how utility data should be investigated, recorded and exchanged M15. The FHWA says their use is not a federal requirement M16.
- The headwind. IIJA program authorities expired on 2026-09-30, and the stopgap law does not extend the Division J advance appropriations M10M11. Public-sector buyers face a funding cliff while utility buyers are at a peak inference.
- Caveat inference: No source links capex growth directly to spending on records or planning software. The why-now is a set of plausible conditions, not a proven pull.
5. Product
| Item | Score | Key evidence |
|---|---|---|
| Problem | 4 | PG&E survey-basemap waits of 2–4 months (gas transmission) and 7–9 months (electric) before CivilGrid P9company-published; San Jose Water handles 500+ data requests a year P24 |
| Solution | 3 | PG&E-confirmed: 40% less field research and survey, up to 4 months faster P5. Larger claims are vendor-only P23P33 |
| Why now | 2 | Conditions only, no direct link (§4) |
| Moat | 3 | Owner relationships plus a human research team P28; counter-positioned against AI inference P18R8. Unproven |
| Distribution | 3 | Sales-led into utilities and engineering firms P27; utility innovation program into enterprise use P6 |
| PMF signals | 3 | PG&E confirms daily use by gas and land teams R1; no reviews, no changelog P41 |
| Competition | 2 | 4M Analytics and Exodigo are better funded and use the same metaphor P38M26 |
Problem. inference For each project, engineers ask every owner in a corridor for records. CivilGrid cites 1–8 weeks per request as the status quo P15company claim, then designs wait on field survey. The direct cost is engineering time. The costly failure comes when a project is permitted and the street was repaved last spring, or another utility digs there next quarter.
Solution. The PG&E-confirmed figures (40% less field research, up to 4 months faster P5) amount to roughly 1.5–2x on specific steps inference. It is meaningful, but it is not 10x. CivilGrid's own case studies claim more: 40h → 5h of research per project at Mark Thomas P23. Neither has a published method.
The $60M, read carefully. CivilGrid's case study describes a 2026 desk analysis. It mapped the boundaries of 1,600+ planned gas distribution projects against city paving plans, moratoria and right-of-way data, and "identified" $60M+ of avoidable paving costs R2. PG&E's newsroom calls the same number "eliminated" but gives no method, period or audit R1. estimate That works out to ~$37.5K per planned project ($60M ÷ 1,600, assuming an even spread). It is a one-time screen; no public source shows whether it recurs R3.
Moat. The bet is counter-positioning. CivilGrid says it uses owner-sourced records with provenance and no AI inference P18, while 4M pairs records with remote AI detection and says it delivered Minnesota DOT lines "100% remotely" R8R9. Provenance only wins if buyers pay for chain of custody, and no source shows that they do inference. There is also a cost question. Owner cooperation and a records-research team in Fort Mill, SC (growing from 10 to 15+ P28) suggest coverage cost grows with each new state inference. Whether CivilGrid can reuse owner records across customers is not in public sources. That one fact decides whether this is a data network or a services business.
6. Founders & team
| Item | Score | Key evidence |
|---|---|---|
| Founder–market fit | 5* | ~10 years as a PG&E engineer, $500M+ of projects managed F1F8F18 |
| Execution speed | 3 | ~6 years from founding to Series A on a small disclosed pre-seed F13F5 |
| Clarity of thought | 4 | "One map" story consistent since 2022–23; scope widened without a pivot F13F14F7 |
| Talent magnetism | 2 | Senior GTM titles exist; no public prior employers for them (data gap) |
| Track record | 3 | Strong corporate operator record; first-time founder |
| Team completeness | 3 | Domain + data/ops co-founders; engineering led by hires inference |
| Integrity of claims | 4 | Core claims confirmed by PG&E's own channel R1; bio details vary (§5 of founders.md) |
* Red-team caveat: the PG&E relationship scores both founder fit and customer proof, so the two are not independent signals.
Earned secret inference. The records already exist inside thousands of owners. CivilGrid claims California has 3,099 agencies with underground assets versus 1,238 registered with 811 P20company claimunverified. The bottleneck is finding, requesting and reconciling them before design. Josh Mackanic did that job from inside a utility: "I managed $500 million in infrastructure projects at PG&E. I was the customer." F8 He recalls: "I had a job that got shut down because a pipe was in the excavation." F3 He left because the problem was "better solved from the outside in than the inside out" F2. Co-founder Brandon Cohen is described as a former water-utility manager F18.
The scene. In September 2023, about three years after founding, CivilGrid was among the organizations invited to PG&E's first Innovation Pitch Fest, in the Undergrounding category F20. It earned a spot to co-develop with PG&E teams. Two years later the platform is "used daily by PG&E's gas and land teams" F12R1. The founder pitched his former employer as an outsider, and his former employer became the anchor customer.
7. Market & competitive structure
Size estimate. US software spend on pre-design subsurface and site data comes to ≈ $68M / $191M / $518M per year (low/base/high). Formula: Σ over segments (buyers × share that buys × ACV), across IOUs, public power and co-ops, ENR Top 500 design firms, engineering establishments, state DOTs, and counties and cities M44. Buyer counts are sourced. Every ACV and every adoption share is a hypothesis, because no company in the category publishes pricing. Red-team flags that the IOU row assumes all 168 IOUs buy, when one is confirmed. Treat the base case as a ceiling, not a forecast.
| Category | Player | Funding (sourced) | Threat |
|---|---|---|---|
| Records + remote AI | 4M Analytics | $45M Series A total, last round Sep 2022 M25 | High: announced ≥6 state-DOT deployments (vendor-side) R9 |
| Physical sensing | Exodigo | $214M raised, 50+ agency/utility customers M26 | Medium: complements desk research today |
| GIS system of record | Esri ArcGIS Utility Network | — M32 | High: could absorb multi-owner sharing as a feature inference |
| Public maps | NYC 3D Underground | $10M HUD grant, users expected early 2028 M18 | Medium in dense cities inference |
| Substitute | SUE consultants, manual records research | — | Default behavior and also a channel |
8. Traction signals (verifiable only)
| Signal | Date | Source |
|---|---|---|
| PG&E: deployed 2025, daily use by gas and land teams, SVP quoted | 2026-08-27 | R1 |
| PG&E: "$60M+ in paving conflicts eliminated" (no method stated) | 2026-08-27 | R1 |
| San Jose Water: data partner since June 2021, 8 users | n.d. | P24 |
| Arizona and Nevada expansion announced | 2025-07-23 | P33 |
| $26M Series A led by Spark Capital | 2026-08-27 | P1 |
| Other named customers (Atmos, Mark Thomas, GHD): not verifiable customer-side | — | R11 |
9. What has to be true
- A second large utility deploys CivilGrid and says so on its own channels.
- The planning layer (Plan module) is bought as recurring portfolio work, not as a one-time screen.
- CivilGrid can reuse owner records across customers, so the cost per mile covered falls as coverage grows.
- Buyers value provenance (chain of custody) enough to choose it over a vendor with broader AI-inferred coverage.
- Utility demand carries the company through a weak public-sector budget year.
10. Bear case & pre-mortem
Bear case (steelman). The one customer-side proof point is the founder's former employer, and its headline number is a one-time desk screen with no stated period, recurrence or realization R1R2R3. The model depends on owner cooperation and a human research team, so costs grow with each new state P28inference. 4M already combines records with remote AI inference that doesn't need owner cooperation, and says it is deployed at state DOTs R8R9. Utilities already hold their system of record in Esri M32. If buyers see the shared map as a commodity, CivilGrid is left serving a small software pool with services margins.
| It's 2029 and CivilGrid is dead. Why? | Early indicator to watch now |
|---|---|
| Market: the value was real but one-time. Utilities stayed at a few seats; DOT budgets stalled after the IIJA cliff | Expanded PG&E scope or a second IOU in procurement or CPUC records; seat counts in new case studies |
| Execution: research-ops cost per mile never fell; the A ran out before national coverage | Ratio of research-ops to engineering job posts; headcount trend; data-reuse terms |
| Competition: coverage beat provenance. 4M won DOTs and engineering firms; Esri, Autodesk or Bentley added the feature | DOT RFP awards; incumbent product announcements; state ASCE 38 adoption |
11. Key risks & what would retire them
| Risk | Severity | What would retire it | When we'd know |
|---|---|---|---|
| n=1 customer proof | High | A second IOU confirming deployment on its own channels | 12–18 months |
| Services-like cost structure | High | Evidence of owner-data reuse across customers; ops headcount growing slower than coverage | Next round disclosures |
| 4M / incumbent coverage | High | Public DOT or engineering-firm wins over 4M; ASCE 38 provenance in RFPs | Ongoing |
| Funding opacity | Medium | Form D or a company statement on total raised | Any time |
| Public-sector budget cliff | Medium | Multi-year surface transportation reauthorization | FY2027 |
12. What I'd ask the founders
- Is the PG&E $60M a one-time screen, or does it repeat each planning cycle? How much of it was realized?
- Can you reuse a record that one owner shares across every customer working in that corridor?
- What does it cost you to add a new state, and how does that cost change over time?
- Where have you lost to 4M Analytics, and why?
- How many paying utilities do you have besides PG&E, and how many seats does a typical utility buy?
- Do buyers name ASCE 38 chain of custody in RFPs, or is provenance your argument rather than theirs?
- What does a utility contract look like (per seat, per project or enterprise), and who signs it?
- Can you reconcile total funding raised to date?
13. Verdict & kill criteria
Verdict. The pain is real. The founder lived it, and the founder's former employer vouches for the product on its own channel, which is rarer at Series A than it should be. But the public record shows one utility and one large, one-time number. The interesting bet is not the map. It is whether CivilGrid becomes the place where utilities plan their capital portfolio around everyone else's digs. The planning layer is the part that is easy to explain and hard to replace once it holds a portfolio's history. The map is the part that three competitors and New York City are all building.
Kill criteria (I change my mind if): - No second IOU confirms deployment on its own channels by end of 2027. - Research-ops headcount keeps scaling one-for-one with new states. - 4M or an incumbent wins a major DOT or engineering-firm account that CivilGrid competed for. - PG&E's next public mention describes the $60M as a one-time exercise with no follow-up program.
Falsifiable prediction: By 2027-12-31, at least one more US investor-owned utility, besides PG&E, will publicly confirm using CivilGrid for project planning on its own channels (newsroom, rate case or procurement record). If that doesn't happen, my thesis is wrong.
Sources (53)
- P2TechCrunch: founded 2020 by Josh Mackanic after ~10-year engineering career at PG&E; product combines utility assets, property ownership and environmental regulation data into a single view ("Google Maps for what's underground"; headline: "Google Maps for the underground"); customers = governments, civil engineering firms, utilities; PG&E named as customer · 2026-08-27
techcrunch.com/2026/08/27/this-former-pge-engineer-is-building-a-googl - P18Data page: data sourced "directly from the owners" via automated records requests with tracking/logging; FAQ states CivilGrid is not a replacement for 811 and does NOT infer utility locations from satellite imagery, scans or AI models; extracts/standardizes map data from PDF scans
www.civilgrid.com/product/data - P21Plan page: ROW conflicts, paving moratoria, long-lead permits, "Dig Once" bundling, jurisdiction/permit analysis, site risk scoring, portfolio prioritization vs budget; describes "automated analysis/recommendations" without the term AI; claims PG&E runs 300+ projects/yr through platform
www.civilgrid.com/product/plan - R2Re-fetched CivilGrid PG&E case study: headline "Identified $60M in Avoidable Costs"; method = mapping 1,600+ planned Gas Distribution project boundaries against city paving plans, moratoria and ROW data (2026 desk analysis); conflicts "would otherwise have emerged" — i.e. flagged, not banked savings · n.d. (analysis 2026)
www.civilgrid.com/case-studies/pge - M12EEI: investor-owned electric utilities on pace for ~$208B capex in 2025 ("highest amount ever"); >$1.1T projected 2025–2029; AI data-center demand cited as key driver · 2025-10-09
www.utilitydive.com/news/investor-owned-utilities-spending-more-than-e - R1Re-fetched PG&E newsroom (Fred Han, 2026-08-27): "$60M+ in paving conflicts eliminated through strategic portfolio optimization"; "Since deploying CivilGrid in 2025"; "used daily by PG&E's gas and land teams"; SVP Christine Cowsert quoted. No methodology, baseline, measurement period or audit stated for any figure · 2026-08-27
www.pge.com/en/newsroom/currents/future-of-energy/from-pitch-fest-winn - F12PG&E newsroom: CivilGrid pitched at PG&E's 2023 Innovation Pitch Fest and earned a spot to co-develop with PG&E teams; two years later in daily use by PG&E gas and land teams; Mackanic titled "CEO and Co-founder", "nearly a decade" at PG&E, managed $500M+ projects · 2026-08-27
www.pge.com/en/newsroom/currents/future-of-energy/from-pitch-fest-winn - P384M Analytics: Series A total $45M after $30M extension (Sept 2022); its blog uses "Google Maps of the Subsurface" · 2022-09-20
siliconangle.com/2022/09/20/4m-analytics-raises-30m-map-u-s-s-undergro 4manalytics.com/blog/4m-analytics-secures-30-million-in-series-a-exten - M254M Analytics: $30M Series A extension (Sep 2022) led by Insight Partners and ITI VC, total Series A $45M; prior $11M round led by Viola, $3M seed led by F2; product 4Map compiles utility data into one dashboard · 2022-09
4manalytics.com/blog/4m-analytics-secures-30-million-in-series-a-exten - M26Exodigo: $96M Series B (Jul 2025) co-led by Zeev Ventures and Greenfield Partners; $214M raised since 2021; 50+ transit agencies, DOTs, municipalities and utilities as customers (Amtrak, LA Metro, National Grid, MTA…); AI + multi-sensing platform · 2025-07-23
www.geoconnexion.com/news/exodigo-brings-ai-efficiency-to-infrastructu - R4CGA 2024 DIRT (as reported by Underground Construction, 2025-08-29): "facility not marked due to incorrect facility record/map" = 2.16%; failure to notify 811 = 24.54%; locator not marked 11.94%; marked inaccurately 8.58%; ~197K reports; top 10 causes = 85%. CGA's own DIRT site returned 403 to direct fetch · 2025-08-29
undergroundinfrastructure.com/news/2025/august/cga-report-shows-utilit - P5PG&E newsroom (author Fred Han): "$60M+ in paving conflicts eliminated through strategic portfolio optimization"; "40% reduction in field research and survey activities"; "up to 4 months faster" scoping/planning/design · 2026-08-27
www.pge.com/en/newsroom/currents/future-of-energy/from-pitch-fest-winn - P35Wayback CDX: civilgrid.com captures from 2017; 2019 capture is an unrelated Chinese-language chemicals site; 2022-01 capture is CivilGrid (Squarespace) "Construction Intelligence Platform" combining utility, environmental, geotechnical data · 2022-01-03
web.archive.org/web/20220103224035/http://civilgrid.com/ - P14Homepage: civil design platform consolidating utility and site data from 100,000+ asset owners; modules Data / Plan / Design; targets utilities, civil engineering firms, DOTs/municipalities; roles SUE, survey, civil engineer, PM, portfolio planner
www.civilgrid.com - F1Josh Mackanic founded CivilGrid in 2020 after a ~10-year engineering career at PG&E · 2026-08-27
techcrunch.com/2026/08/27/this-former-pge-engineer-is-building-a-googl - P19Data page: chain of custody per record (source, request history, metadata) documented as ASCE 38-22 compliant; SOC 2 Type II; exports DWG, DGN, SHP, KMZ; no named direct Esri/Civil 3D integration
www.civilgrid.com/product/data - P17No public pricing on site; CTAs are "Talk to our team"/"Book a live walkthrough"/"Try CivilGrid"; self-serve registration exists at map.civilgrid.com/register
www.civilgrid.com - F13PR Newswire release: CivilGrid launched 2020 by founder & CEO Mackanic, "14 years in the utility and construction industry"; $2.2M pre-seed raised over the prior 15 months (investors not named); PG&E and San Jose Water working with CivilGrid · 2022-07-20
www.prnewswire.com/news-releases/digital-infrastructure-mapping-initia - P1CivilGrid raised a $26M Series A led by Spark Capital; Afore, A*, Ford Street Ventures, SNR participated; Energy Impact Partners also invested (EIP has utilities among its LPs) · 2026-08-27
techcrunch.com/2026/08/27/this-former-pge-engineer-is-building-a-googl - F5$26M Series A led by Spark Capital; Afore, A*, Ford Street Ventures, SNR and Energy Impact Partners participated · 2026-08-27
techcrunch.com/2026/08/27/this-former-pge-engineer-is-building-a-googl - R5SEC EDGAR: company search "civilgrid" (Form D) → "No matching companies"; full-text search for "CivilGrid", "Civil Grid", "Agora Maps" → 0 hits. Legal entity name may differ, so this does not establish absence of filings
www.sec.gov/cgi-bin/browse-edgar?company=civilgrid&type=D&action=getco - R6TechCompanyNews (2026-09-01): "PitchBook data indicates total capital raised of approximately $31.7 million" across earlier seed/accelerator rounds, and calls the Series A the largest raise to date. Caplight: total $51.2M; lists two Series A entries (2026-05-17, amount locked; 2026-08-27, $26M); "Valuation information … is not publicly available" · 2026-09-01 / n.d.
www.techcompanynews.com/civilgrid-raises-26-million-in-series-a-fundin www.caplight.com/company/civilgrid - R7Verdict on cumulative funding: not verifiable. "Total capital raised ≈ $31.7M" reads equally as cumulative incl. the $26M A ($31.7M − $26M = $5.7M prior, closer to the disclosed $2.2M pre-seed) or as prior-only; Caplight $51.2M − $26M = $25.2M = Employbl's $25.2M, suggesting aggregators mix cumulative and prior totals. No primary source (Form D, company) found · 2022–2026
derived (see claim) - P27Built In: 26 total employees, founded 2020, HQ SF; 5 open jobs: AE Engineering Firms (CA), Sr/Staff PM (Boston), Strategic CSM (remote), Director/Sr Director PM (Boston), BDR Government (San Diego)
builtin.com/company/civilgrid - P28Simplify listing (inactive) for Utility Research Operations Specialist, Fort Mill SC hybrid, $85.6–100.5K + equity; research/outreach team growing "from 10 to 15+"; About: data from "3,000+ utilities, agencies, and municipalities", "dozens of leading engineering firms"; Simplify shows size 51–200, headcount −14% over 2 yrs, total funding $28.2M · n.d. (~2026-06)
simplify.jobs/p/2bc3528b-ca5e-41b4-9fc1-1c421e961e44/Utility-Research- - M13PG&E plans $73B capital expenditures through 2030 · 2025-09-30
amp.insurancejournal.com/news/west/2025/09/30/840989.htm - M8CGA "Between the Lines: The True Cost of Damages to America's Buried Utilities": $83.2B/yr economic cost; 668,999 damage incidents modeled for 2025; ~3x the $30B 2019 benchmark; direct repair = $4.9B (5.9%) · 2026-08-13
www.wateronline.com/doc/new-report-underground-utility-damage-costs-am - M15ASCE published ASCE/UESI/CI 38-22 (investigating & documenting existing utilities; replaces 38-02) and ASCE/UESI/CI 75-22 (recording & exchanging utility infrastructure data) in July 2022 · 2022-07
undergroundinfrastructure.com/news/2022/july/american-society-of-civil - M16FHWA: ASCE 38/75 provide technical guidelines; their use is not a Federal requirement. Colorado statute (eff. 2019-01-01) requires engineers on public projects to meet ASCE 38 and strive for Quality Level B (per search summary of FHWA/state docs — verify statute text)
www.fhwa.dot.gov/construction/dabs/library_hif24094.pdf - M10IIJA program authorities expire 2026-09-30; Division J advance appropriations cover FY2022–FY2026 only · 2026-08-04
www.naco.org/news/iija-authorities-expire-september-30-naco-urges-cong - M11Continuing Appropriations and Extensions Act of 2027 (P.L. 119-416) enacted; continues many IIJA surface transportation programs but not Division J advance appropriations (e.g., Bridge Formula, SS4A cut) · 2026-09-02
www.naco.org/news/federal-funding-extension-creates-transportation-fun - P9CivilGrid PG&E case study (company-reported, no methodology): 3–4 months saved per project; $20–30K engineering time saved per project; 30–40% lower survey costs; 50% fewer potential design changes; pre-CivilGrid survey basemap waits of 2–4 months (gas transmission) and 7–9 months (electric)
www.civilgrid.com/case-studies/pge - P24San Jose Water case study: data partner since June 2021, 8 users, CivilGrid Plan; SJW handles 500+ data requests/yr to other utilities; 2–5% project cost savings attributed to consolidated GIS data (company-reported)
www.civilgrid.com/case-studies/san-jose-water - P23Mark Thomas case study (company-published): utility research 40h → 5h per project (88%); 75% lower cost/project; >$500K annual savings at 100+ projects/yr; quote from Francisco Rojas, SUE Program Manager; no dates or methodology
www.civilgrid.com/case-studies/mark-thomas - P33InBusiness PHX: CivilGrid announced expansion into Arizona and Nevada; claims up to 90% faster pre-construction; CEO cites waitlist of NV/AZ firms; GHD's Caisey/Casey Raines quoted · 2025-07-23
inbusinessphx.com/?p=187328 - R84M Analytics homepage: compiles "millions of scattered public utility records into one easy-to-use map" plus AI computer-vision detection of above-ground assets from satellite imagery; "As easy as Google Maps"; logo strip incl. GDOT, CDOT, MDOT, WSB, MasTec, Stantec
4manalytics.com - P6PG&E newsroom: Mackanic pitched at PG&E's 2023 Innovation Pitch Fest; PG&E deployed CivilGrid in 2025; platform "used daily by PG&E's gas and land teams"; SVP Christine Cowsert quoted in support · 2026-08-27
www.pge.com/en/newsroom/currents/future-of-energy/from-pitch-fest-winn - P41No independent user reviews found (G2/Capterra/Reddit/HN searches returned only job listings and press)
derived (see claim) - P15Homepage stats (company-reported): 4,200+ projects completed; 1,500+ data sets digitized; 2.8M+ miles of utilities mapped; 200K+ hours of records research saved; 1–8 weeks per data request cited as status quo
www.civilgrid.com - R3Verdict on "$60M+ PG&E savings": weaker than stated. PG&E-side confirmation of deployment and daily use is solid; the $60M is a one-time portfolio screen of planned projects (identified), with no public period, recurrence or realization data. PG&E's "eliminated" wording is not supported by a stated method · 2026
derived (see claim) - R94M press releases (company-published, no DOT staff quoted): MnDOT ~12,000 highway miles, lines generated from network topology "100% remotely" (2026-05-11); Michigan DOT (2026-05-18); Alabama DOT (2026-05-27); other DOT clients named: TxDOT, GDOT, CDOT · 2026-05-11 / 2026-05-27
www.4manalytics.com/blog/4m-delivers-utility-intelligence-for-minnesot www.4manalytics.com/blog/4m-analytics-supports-project-development-and - F8Company site lists Josh Mackanic as "CEO and Founder" and Brandon Cohen as "Co-Founder"; Mackanic quote "I managed $500 million in infrastructure projects at PG&E. I was the customer." · n/a (live page)
civilgrid.com/about-us - F18ENR: Brandon Cohen is co-founder and a former water utility manager; Mackanic managed $500M+ over 10 years at PG&E; quotes "It's not that locators are bad. We're putting too much weight on their shoulders" and (on San Bruno) "Having that in our DNA as a company has been helpful"; data from 3,000+ owners in CA, FL, TX · 2026-09-03
www.enr.com/articles/63611-underground-conditions-platform-civilgrid-r - F14CivilGrid in Imagine H2O 2023 Accelerator cohort; Demo Day 2023-07-26; Mackanic: "The guidance they've given us has been invaluable."; described as "the Google Maps equivalent of the underground" · 2023-07-20
www.wateronline.com/doc/get-civilgrid-makes-waves-in-the-water-industr - F7Mackanic: "Right now, we're providing engineers the data to be able to make a decision"; plans to move toward automating decisions and permitting · 2026-08-27
techcrunch.com/2026/08/27/this-former-pge-engineer-is-building-a-googl - P20Data page claims 3,099 California agencies vs 1,238 in 811, and 3,936 Georgia agencies vs 873 in 811; source-count figures on the page are inconsistent (35,000 / 100,000+ / 3,000+)
www.civilgrid.com/product/data - F3Mackanic quote: "I had a job that got shut down because a pipe was in the excavation" · 2026-08-27
techcrunch.com/2026/08/27/this-former-pge-engineer-is-building-a-googl - F2Mackanic quote on leaving PG&E: "I felt like this was a problem that was going to be better solved from the outside in than the inside out" · 2026-08-27
techcrunch.com/2026/08/27/this-former-pge-engineer-is-building-a-googl - F20CivilGrid among organizations invited to PG&E's first Innovation Pitch Fest (Sept 12–14, 2023), Undergrounding category · 2023-09-11
www.renewableenergyworld.com/news/pge-selects-54-organizations-to-part - M44US bottom-up SAM for subsurface/site-data software ≈ $68M (low) / $191M (base) / $518M (high) per year — see market.md formula · 2026-11-02
derived (see claim) - M32Esri ArcGIS Utility Network models gas, water, electric, communication, district heating networks
www.esri.com/arcgis-blog/products/arcgis-enterprise/data-management/vi - M18NYC announced 3D Underground (3DU) citywide utility/subsurface mapping platform; $10M HUD disaster-recovery grant; user launch expected early 2028 · 2025-11-24
admin.undergroundinfrastructure.com/news/2025/november/nyc-launches-3d - R11Verdict on named customers beyond PG&E/SJW (Atmos, Mark Thomas, GHD): not verifiable. Web and domain-restricted searches (ghd.com, markthomas.com, atmosenergy.com) returned no customer-side mention; only lead: a Geosyntec staff LinkedIn post (search snippet, not opened)
derived (see claim)
Disclosure. Not investment advice. This memo is based solely on public information as of 2026-11-02. Disclosure: I'm a student at Minerva University, not a professional investor or analyst. I write these memos to learn how to evaluate companies, in public. I hold no position in CivilGrid or its competitors, have no relationship with the company or its investors, and no one paid for this memo. I will get things wrong; corrections are welcome at [email protected].