# Electrical Contractor Estimators
## Who they are
**Primary target: Estimators and Chief Estimators at small-to-mid commercial/industrial electrical contractors (11–200 employees).** These are the people inside an EC firm who pull bid documents off a portal, sit down with a PDF plan set, and count every receptacle, light fixture, conduit run, panel, and piece of gear so the firm can put a price on the job. They are the entry point to every dollar the company earns.
Their workflow is fundamentally upstream of the personas we already target — distributors quote *after* the contractor's BOM exists; manufacturers quote *after* the contractor specifies the gear. The EC estimator is the one who actually builds that initial BOM from the drawings. They're the supply chain's *demand side*.
**Titles seen on LinkedIn:**
- Electrical Estimator / Senior Electrical Estimator
- Chief Estimator (at firms with 5+ estimators under them)
- Estimator / Senior Estimator (at electrical firms — disambiguate via company keyword)
- Estimator–Project Manager / PM–Estimator hybrid (common at <50-employee ECs)
- Preconstruction Manager (often the estimator wearing a fancier hat)
- Director of Estimating / Estimating Manager (10–50 person estimating teams)
**Company profile:**
- Commercial/industrial electrical contractors — Division 26 (the CSI MasterFormat code for electrical; older "Division 16" still floats around in conversation)
- 11–200 employees is the sweet spot. Below 11 = service-heavy / residential and the owner does their own takeoffs. Above 200 = enterprise procurement and the estimator becomes one of 20 specialists rather than a hands-on bidder
- $5M–$150M revenue range (ENR Top 600 median is $128M — most of our target sits below that)
- Examples of the shape we want: Foster Electric (CO), All Industrial Electric (IL), HSH Contracting Services (ON), Medina Electric (NC). NOT what we want: Cache Valley Electric, ArchKey, Guarantee Electrical (ENR Top 50 — too big)
- IBEW/union shops, merit shops (IECI/ABC), and independents are all in scope — the estimator's daily pain is identical across labor model
**What their day looks like:**
- Reviews bid invitations from GCs and project owners — usually 3–10 per week, each with a Friday deadline
- Downloads the plan set and Division 26 specifications. Spends the first hour just figuring out scope and what the engineer actually wants
- Spends the bulk of the day on **quantity takeoff**: counting symbols, measuring conduit and wire runs, tabulating fixtures, devices, gear. A typical commercial bid: ~12 hours of takeoff work. Larger projects easily run 30–80 hours
- Cross-references the takeoff against panel schedules, one-lines, and lighting schedules to make sure nothing is missed
- Loads quantities into estimating software (Accubid, McCormick, ConEst, PataBid, TurboBid) or — surprisingly often — a personal Excel file
- Applies labor units (typically NECA Manual of Labor Units) and material pricing
- Adds overhead (industry avg 19.16% of sales), markup, and bid bonds
- Phone rings constantly while doing all of this. At smaller firms, the estimator is also fielding submittal questions, change-order pricing, and purchasing calls
- Win rate: ~9% on individual bids submitted (industry avg, 1 in 11). Firms that respond within 24 hours win 25% more
**Who they are NOT:**
- Pure residential electricians (different work — service calls, panel upgrades, no plan-set takeoffs)
- Service & maintenance estimators (T&M and flat-rate work, not blueprint takeoffs)
- Field foremen / superintendents (run installs, don't bid them)
- Estimators at ENR Top 100 ECs (work in 10–50-person estimating departments; software decisions go to a Director of Preconstruction and tooling is enterprise-procured)
- Owners at <10-employee shops (do their own takeoffs but won't trial a tool — see Bay Area sub-owner campaign at 0.8% reply / 0 conversions)
## Problems we solve
- **Manual takeoff is the single biggest time sink in the bid process.** A mid-size EC estimating team averages 12 hours per bid, and the takeoff is the bulk of that. Manual symbol counting on a 40-page plan set has a 5–10% fatigue error rate; humans get worse the longer they stare at the page ([Procore Estimating Guide](https://www.procore.com/library/estimating-electrical-contractors), [Trimble](https://www.trimble.com/blog/construction/en-US/article/how-to-master-the-takeoff-in-construction-estimating)).
- **The phone never stops.** EC Mag's "Day in the Life of an Estimator" is literally titled *Overwhelmed* — "Most estimators in small-to-medium companies are required to do more than just prepare estimates—many are also project managers, purchasing agents and truck drivers." Interruptions destroy takeoff accuracy because you lose your place on the page ([ECMag](https://www.ecmag.com/magazine/articles/article-detail/your-business-overwhelmed-day-life-estimator)).
- **Slow turnaround loses winnable bids.** Contractors who get the proposal back inside 24 hours win 25% more jobs. With a ~9% baseline win rate on commercial bids, every hour of takeoff speed is directly revenue ([Blackridge Research via BuildOps](https://buildops.com/resources/electrical-bidding/)).
- **Errors compound through the quote.** A miscount of 5% on fixtures cascades into wrong labor units, wrong material cost, wrong overhead allocation. Underestimating indirect costs is the #1 reason electrical bids lose money. With commercial margins at 10–15%, a 5% takeoff error eats half your profit ([FieldEdge](https://fieldedge.com/blog/electrical-contractor-pricing-strategies-for-commercial-projects/)).
- **Can't hire enough estimators.** 82% of construction firms struggle to hire salaried roles. NECA estimates 10k electricians retire vs 7k entering the trade each year — and experienced estimators are even scarcer because they take a decade in the field to grow. Throughput-per-estimator is the only lever most ECs can pull ([HBI Labor Report Fall 2025](https://hbi.org/wp-content/uploads/2025/10/Fall-2025-Final-Construction-Labor-Market-Report-Update.pdf), [Raiven](https://raiven.com/blog/demand-for-electricians-is-soaring/)).
- **Legacy estimating software is expensive and dated.** Trimble Accubid Classic costs $2,500 upfront + $800/yr, requires LiveCount as a separate add-on for graphical takeoff, and is entirely manual — no AI. Accubid Anywhere (the cloud version) runs $15k+/yr per user. McCormick and ConEst are similar in price. Many small ECs simply can't justify the cost and stay on spreadsheets ([Software Advice](https://www.softwareadvice.com/construction/mccormick-electrical-estimating-profile/vs/trimble-accubid-anywhere/), [ECN forums](https://www.electrical-contractor.net/forums/ubbthreads.php/ubb/printthread/Board/20/main/21985/type/thread.html)).
- **Excel-built personal workflows are everywhere but don't scale.** Many SMB estimators have built their own Excel templates over years of bidding and are proud of them. They work for repeat project types but fall apart on novel scopes and can't be shared across a team ([Electronate](https://www.electronate.app/blog/estimating-software-vs-excel)). Our All Industrial Electric replier in the 2026-01 campaign said exactly this: "built an excel file that I bid my jobs really fast using." Jason Wagner at North Valley Electric runs a similar setup — *"custom spreadsheet with formulas: listed price, labor, taxes."*
- **Proposal writing is a 4x multiplier on top of estimating.** Direct quote from Jason Wagner (North Valley Electric, call 2025-10-10): *"Estimating takes 2 hours, but writing the proposal takes 8 hours. Every electrical scope is different. Equipment packages alone require 2 pages of detailed specs (phase, indoor/outdoor, amperage, etc.)."* Most takeoff tools stop at quantities; the proposal/spec sheet is where the real time goes.
- **Drawing quality is degrading.** Dan McNeillie (Kress Electric, 20+ years estimating): *"Poor drawing quality getting worse over time. Architects/engineers lack time, may bid just to get the contract. Symbol inconsistencies (e.g., receptacle symbol exists but a note says all receptacles need to be type X — no symbol for that)."* AI tools that depend on clean symbology hit this wall constantly.
- **The bid/no-bid decision is harder than the bid itself.** Dan McNeillie: *"Job selection (to bid or not to bid) is the biggest challenge, not job details."* Estimators don't just need faster takeoffs — they need triage to decide which RFPs deserve takeoff effort at all.
- **Material price volatility makes estimates perishable.** Chris Low (JE Dunn / Rosendin, large EC) explained Rosendin uses 9-month commodity forecasting for copper because *"the estimate is only good for the day it's made."* This is a known industry condition that punishes slow estimators — by the time you submit, your materials cost is wrong.
## How Gink helps
| Problem | Gink capability | Fit level |
|---|---|---|
| Manual symbol counting eats 12 hours/bid | Upload PDF, get instant quantity takeoff for fixtures, devices, panels, and gear. Minutes instead of hours. | **Strong** — current core capability |
| 5–10% manual error rate destroys margin | AI counts are consistent and auditable — every symbol's location is preserved and reviewable | **Strong** |
| Phone interruptions ruin takeoff accuracy | One pass, one result. The AI doesn't lose its place when the phone rings. | **Strong** |
| 24-hour turnaround wins 25% more bids | Compresses the longest single step of the estimate from hours to minutes; same-day proposals become realistic | **Strong** |
| Can't hire more estimators | One estimator with Gink covers the throughput of two without it. Solves the staffing pinch without payroll growth. | **Strong** |
| Excel-built personal workflows | Gink hands you the quantities; estimators paste them into the workflow they already have. Augments instead of replacing. | **Strong (positioning matters)** |
| Accubid/McCormick/ConEst integration | Quantities export to common formats; users still use their existing software for labor units, pricing, and proposal generation. Gink only replaces the takeoff step. | **Partial** — manual paste works today; deeper integrations are roadmap |
| Conduit length measurement | Gink counts symbols accurately; conduit length runs along path geometry. Quality varies by drawing detail. | **Partial** — improving |
| BOM export for vendor pricing | One of the highest-confidence features from call feedback (Christopher Randall, Emerald Electric: *"BOM download for vendor quoting is important"*). Same need from Dan McNeillie ("partial BOM download is valuable"). | **Strong (validated on calls)** |
| Switchgear / panel detection | Dan McNeillie: *"Switch gear detection is most reliable feature — prioritize it."* Chris Low: *"As an estimator, doesn't look at lighting schedule. Needs to quickly count fixtures (A1, A2), then send counts to supplier."* Count accuracy on gear is core value. | **Strong (validated)** |
| NEMA ratings & panel specs extraction | Christopher Randall flagged *"concerns about extracting specific panel specs like NEMA ratings."* Right now Gink counts panels and reads labels; structured spec extraction (NEMA type, bus rating, AIC, mains) is a deeper feature gap. | **Partial → Future** |
| One-line / single-line diagram interpretation | Chris Low (Rosendin/JE Dunn): *"Nobody knows how to translate a single line to plan with 50 panels. Point-to-point on big feeders (from/to, length, type of conduit) = 60% of electrical estimating value."* This is the single most valuable future capability — gating requirement for large-EC adoption. | **Future (high-leverage)** |
| Proposal generation from takeoff | Jason Wagner's 2h-estimate / 8h-proposal pain. Gink could feed structured BOM + spec data into proposal templates. | **Future** |
| Mechanical equipment counting | Dan McNeillie: *"Mechanical equipment typically supplied by mechanical contractor (99% of the time)."* — out of scope for EC estimator workflow. Don't waste effort here. | **Out of scope** |
**Honest assessment of current fit:** EC estimators are the strongest current fit for Gink among the personas we've researched. Their core pain — count every symbol on every page — is the exact problem Gink was built to solve. Unlike manufacturers (one-line interpretation gap) or distributors (mostly counting from already-extracted BOMs), the EC estimator's workflow maps directly onto today's product.
**Positioning vs. competitors:**
- **PataBid Quantify** ($115/mo per user): AI-enhanced graphical takeoff inside a full estimating suite. ~80% AI accuracy on symbol counting per their own marketing. Direct competitor. Differentiation: Gink is *just* the extraction layer (lightweight, fast, no estimating UI to learn) and integrates with whatever estimating software the contractor already uses.
- **Beam AI (iBeam.ai)**: Service-model AI takeoff — estimator uploads PDF, expert reviews, 24–72 hr turnaround at "±1% of in-house accuracy." Competes on accuracy + human-in-loop. Differentiation: Gink is self-serve and minutes, not hours.
- **Trimble Accubid / McCormick / ConEst**: Legacy estimating suites with manual takeoff (Accubid) or graphical takeoff add-ons (LiveCount, McCormick PT). No AI. Differentiation: Gink replaces the most painful step (extraction) while the contractor keeps using their existing labor/pricing/proposal engine.
- **TurboBid** ($99/mo): Cheap, simple, mostly manual. Different market — small residential shops. Not a real competitor.
The strategic insight: the legacy estimating-suite incumbents own the labor-unit and pricing layer, which is where their moat lives (decades of cost data). They are weakest on takeoff extraction. Gink should slot in as the takeoff layer that feeds their tools, not as a replacement for the whole suite.
## Objections we'll hear (and how to handle them)
These are real objections we've heard on calls and in campaign replies. Listed in rough order of frequency.
### "I tried AI estimating, it didn't work for electrical."
**Heard from:** Jason Wagner (North Valley Electric) — *"used it in 2017 at A1 Electrical, gave up. Electrical is field-installed after everyone else — you don't know how many I-beams are in the way, if things will move, ceiling height varies. Always estimate worst case."*
**Response approach:** Acknowledge the historical baggage. AI in 2017 was symbol-matching templates and was bad. Today's foundation-model-based extraction is a different generation. Don't argue — offer a free trial on one of their own drawings. The product speaks louder than the pitch. Frame Gink as *takeoff extraction*, not *estimating* — we don't claim to know the field unknowns; we give you accurate counts so you can apply your own field judgment.
### "Last AI tool I tried was 90% accurate — that's not good enough for competitive bids."
**Heard from:** Dan McNeillie (Kress Electric) — *"Drawer.ai: autocount is 90% right but the remaining 10% still needs manual correction. For competitive bids, details matter. If it's not competitive, the precision requirement goes up."*
**Response approach:** This is the most credible AI-skeptic objection because it's grounded in real product experience. Two angles: (1) accuracy depends on drawing quality and standardization — let them test on a *standard* commercial project (where Gink hits 95%+) before judging on edge cases; (2) emphasize uncount/correction tools and reviewable provenance — the value isn't 100% autonomy, it's getting from 0% to 95% in minutes vs hours, then correcting the last 5%. Reference Christopher Randall's feedback that the cleanup feature ("uncount" like Trimble LiveCount right-click) is essential.
### "We use Accubid / McCormick / our own Excel."
**Heard from:** Dan McNeillie's journey was *"Paper → Accubid → Bluebeam → drawer.ai."* Jason Wagner runs a custom Excel with formulas. All Industrial Electric replier from 2026-01: *"built an excel file that I bid my jobs really fast using."*
**Response approach:** Lead with augmentation, not replacement. Gink hands you quantities; you keep using Accubid for labor units, McCormick for pricing data, your Excel for proposal logic. Specifically NEVER position as a replacement for the labor-unit and pricing data — that's where the legacy tools win. The takeoff step is where Gink wins. This positioning was validated on Juan Pimentel's 2026-03-27 call: *"Should only provide factual data and high-level summaries. Should NOT make bidding decisions for the contractor."*
### "Online supplier pricing doesn't reflect my real supplier relationships."
**Heard from:** Jason Wagner — *"Software he tried used online pricing which doesn't reflect real supplier relationships. Electrician A gets $400, B gets $360, C gets $260 for the same product. Relationships drive pricing."*
**Response approach:** Don't claim Gink prices anything. Gink delivers quantities + product specs; the contractor's existing pricing engine (Accubid TRA-SER, McCormick, or supplier-direct quotes from their reps) applies their negotiated rates. Avoiding the pricing layer is a feature, not a bug.
### "We need single-line / one-line diagram understanding, not just floor plan symbol counting."
**Heard from:** Chris Low (JE Dunn / Rosendin) — gating requirement before Rosendin will engage further. *"Single line diagram translation. Point-to-point on big feeders (from/to, length, type of conduit) = 60% of electrical estimating value."*
**Response approach:** Be honest about the current gap. SLD interpretation is on the roadmap; it's the same gap called out in the manufacturer persona. For large ECs (Rosendin scale), this is a blocker — don't push them to adopt until the feature lands. For small/mid ECs whose primary pain is floor-plan symbol counting (the 90% of bid hours), Gink's current capability already solves the bigger pain.
### "I'm not a tradesperson / I don't do takeoffs."
**Heard from:** Bay-area owners (2025-12 campaign). One owner: *"I'm not a tradesperson, nor do I have any on site experience, so I'm probably not the best fit."* Stefan Demeyere (Pleasant Hill Electric, residential): *"Solo electrician, prices on phone, doesn't really use software."*
**Response approach:** Targeting failure — should never have reached this lead. Use Sales Nav exclusions (Owner, President, residential keywords) to prevent the bleed. If accidentally targeted, politely close: "Sounds like Gink isn't a fit for your work today — appreciate the directness."
## Incentives
### Personal
- Get home before 7pm during bid weeks (estimating is famously a "two-week sprint into the deadline" job)
- Hit fewer phone-call interruptions during heads-down takeoff work
- Look modern and competent to ownership — "I found a tool that lets us bid 3x more jobs without hiring" is exactly the kind of win that gets noticed at a 50-person EC
- Fewer post-award arguments about scope misses (a missed receptacle in the takeoff becomes a $500 fight at the job-cost meeting)
- Build internal credibility for the next promotion (Senior Estimator → Chief Estimator → Director of Preconstruction)
- At small firms (11–50 employees), the estimator who improves bid throughput often gets equity or profit share. Owner sees the bid pipeline directly.
### Company
- Win more bids by responding faster (24-hour proposals win 25% more)
- Improve win rate by reducing scope misses and bid errors (commercial margins are 10–15% — accuracy is survival)
- Grow revenue without hiring (NECA labor shortage means hiring isn't an option for most firms)
- Bid on more project types — speed lets the firm say yes to RFPs they used to skip
- Compete against larger ENR-ranked ECs who have full preconstruction departments
## Sales Navigator filters
- **Job titles:**
- Inclusion: "Electrical Estimator", "Senior Electrical Estimator", "Chief Estimator", "Estimator" (with company keyword filter), "Preconstruction Manager", "Director of Estimating", "Estimating Manager", "Project Manager Estimator", "Estimator Project Manager"
- Strongest signal: "Electrical Estimator" exact title at a company with "electrical" in the name
- **Industries:** "Construction" (primary), "Civil Engineering" (some ECs misclassify), "Renewables & Environment" (solar ECs)
- **Company size:** 11–200 employees. Strong preference for 11–100. Allow 100–200 with caution.
- **Keywords (company):** "electrical contractor", "electrical contracting", "electrical construction", "commercial electrical", "industrial electrical", "Division 26", "Div 26", "design-build electrical", "electrical services" (filter — many service-only shops use this; manually qualify)
- **Keywords (profile):** "takeoff", "estimating", "blueprint", "Accubid", "McCormick", "ConEst", "Division 26", "NECA labor units", "MLU", "bid", "preconstruction"
- **Geography:** US-wide with weighting. Over-index: **Texas, Florida, Ontario, Quebec, Tennessee, North Carolina, Georgia, Indiana**. Under-index but include: Midwest industrial belt, mid-Atlantic. **Exclude Bay Area** entirely (Bay Area sub-owner campaign: 0.8% reply, 4% accept, 0 conversions).
- **Active on LinkedIn filter:** Strongly recommended. Prior estimator-PM-active campaign had 40% acceptance rate vs ~10% baseline — the highest connection rate we've ever seen.
- **Exclusions:**
- Companies with 1000+ employees (enterprise procurement kills startup trials — proven across 6 campaigns)
- Titles: "Owner", "President", "CEO" (2.8% reply rate, 0 conversions historically — they don't do takeoffs)
- Titles: "Foreman", "Superintendent", "Field Supervisor", "Apprentice", "Journeyman", "Electrician" (install side, not bid side)
- Titles: "Service Manager", "Service Technician", "Service Coordinator" (T&M work, no plan-set takeoffs)
- Pure residential ECs (filter by company keyword: avoid "residential electrician", "home electrical", "panel upgrade", "EV charger installation" as primary company descriptor)
- Estimating services firms (already covered by the 2026-03-estimating-services persona)
- Distributors and manufacturers (already covered by `electrical-distributor` and `electrical-manufacturer` personas)
## Outreach angles
### 1. Founder story + bid throughput pain ask — HIGH CONFIDENCE
Open with founder credibility ("former electrical apprentice turned dev, spent a decade building PlanGrid and Autodesk") and end with an open question about manual takeoff time. Under 60 words. This is the exact formula that hit 13.7% reply rate and 42% positive sentiment in the 2026-03 estimating-services campaign — and EC estimators are a tighter version of that audience.
Example DM#1 skeleton (under 60 words):
> Hey {{firstname}}, I'm an electrical engineer who spent the last decade building tools at PlanGrid and Autodesk. Now I'm building AI that pulls quantity takeoffs straight from PDF plan sets — for Division 26 specifically.
>
> Curious — how many hours does a typical commercial bid eat at {{companyName}} right now?
**Rationale:** Founder-story-first wins by 5.4x over product-first across all our data. Open-question CTA wins by 12x over "15-minute call." "Division 26" is the insider signal that signals trade knowledge in the first sentence. Length is in the proven 55–60 word sweet spot.
### 2. "Bid more without hiring" labor-shortage angle — HIGH CONFIDENCE
Frame around the labor crisis. "Most ECs I talk to can't hire estimators fast enough — we built Gink so one estimator can cover the throughput of two." Lands especially well with chief estimators and owners-who-estimate, because they feel the hiring squeeze directly. Cite NECA's 10k-out / 7k-in retirement gap.
**Rationale:** The labor shortage is the #1 universal anxiety in the EC industry right now. The 2026-04 and 2026-05 distributor campaigns both confirmed throughput-per-headcount is a stronger angle than pure efficiency. Untested at EC estimators specifically, but adjacent persona signal is strong.
### 3. "Augment your Excel, don't replace it" anti-displacement — MEDIUM-HIGH CONFIDENCE
Many SMB estimators have a personal Excel workflow they're proud of. Direct quote from All Industrial Electric estimator in 2026-01: *"built an excel file that I bid my jobs really fast using."* Lead with "Gink hands you the quantities — you keep using the workflow that already works for you" rather than "replace your spreadsheet."
**Rationale:** Acknowledged in the 2026-01 retro as a clear positioning fix. Estimators are technically minded and territorial about their tooling. Anti-displacement framing reduces the "we already have X" objection seen in the playbook.
### 4. "Built by an EE who spent the decade at PlanGrid" — HIGH CONFIDENCE for email subject
Use the dual-credibility hook for email subject lines: *"Former electrical apprentice & Autodesk dev"*. This drove 54% email open rate in 2026-03 — best of any campaign. Body of email is a longer version of the founder story (90–100 words is fine for email, unlike LinkedIn DMs).
**Rationale:** Direct port of the proven email subject from the gold-standard campaign. Should perform similarly for EC estimators because the audience is the same shape (technical, trade-credible, software-curious).
### 5. "Your contractors are losing bids to faster competitors" — MEDIUM CONFIDENCE
Cite the 24-hour-proposal-wins-25%-more stat and frame Gink as the way to compress takeoff time. Better as a DM#2 follow-up than a cold open — it's data-heavy and works better once the recipient knows who you are.
**Rationale:** Strong, citable industry stat. Risk: sounds like marketing copy if it leads. Use as a supporting argument once the founder story has earned attention.
### 6. "We are the missing upstream step for Accubid users" — USE SELECTIVELY
If discovery reveals the EC already uses Accubid or McCormick, position Gink as the takeoff layer that feeds those tools rather than competing with them. "Accubid prices the BOM beautifully — we just hand you the BOM to start with."
**Rationale:** Same positioning logic as Gink-vs-Parspec in the distributor persona. Reduces the "we already have software" objection. Only useful after discovery — never lead with it cold.
## Persona-specific learnings
### Voice from active CRM electrical contractors
These are real ECs we've talked to (1+ calls each in `apps/internal/data/calls/`). Their quotes drive a lot of the framing in this doc. Listed by relevance to the persona, not chronology.
**Juan Pimentel — NAX Electrical Services (founder/estimator, MA)**
- 4-year-old EC; first year $3.5M in contracts. Apprentice → general foreman → owner. Eager early adopter: *"AI everything. What tool can I have?"*
- Already uses **Draw AI** (competitor) on lighting takeoffs. Saves 50% of time. Accuracy ~60% on complex (circular/spread-out fixtures), 90–92% on standard office TI and multi-residential corridors. Worth the cost despite imperfection.
- Hates existing software: *"poor, rarely update, constantly raise prices."*
- Wants downstream: procurement, change orders, PM.
- UX bar: *"ask your phone about a panel rating instead of searching through drawings."*
- Strong product principle from this call: *"Should only provide factual data and high-level summaries. Should NOT make bidding decisions for the contractor."* — informs our positioning.
**Christopher Randall — Emerald Electric (estimator + LinkedIn Learning author, demo done)**
- Created a CSI MasterFormat estimating course on LinkedIn Learning (480+ high ratings). Deep domain expert; high accuracy bar; aligned mental model with Alexei on first call.
- 2026-03-26 demo: improved results picked up *"circuits, breakers, distribution panels, transformers."* Impressed Gink picked up shaded daylight-harvesting zones.
- Top feature requests: **uncount feature** (Trimble LiveCount-style right-click), **BOM download for vendor quoting**, NEMA ratings / panel specs extraction, better sheet display/navigation.
- Industry-condition context: *"3-month lag between aggressive bidding and project start, causing current slow period."* — feeds the "win more bids" angle, but also the bid/no-bid triage need.
**Dan McNeillie — Kress Electric (electrician → estimator, 20+ yrs, 50-employee EC, ON)**
- Software journey: Paper → Accubid → Bluebeam → drawer.ai. He represents the technically-fluent senior estimator we want as champion at every 50-person EC.
- Hiring pain: *"Biggest hiring issue is electricians, not work. Can get the work but can't staff it."* Confirms labor-shortage angle.
- *"Time is the #1 problem. Everything is a race."* — best one-line summary of the persona's pain.
- Drawing-quality decay: receptacle symbol exists but a note says "all receptacles type X" — no symbol for that. The kind of edge case that breaks naive symbol matchers.
- 90% accuracy ceiling on drawer.ai is the credibility bar Gink has to clear for him.
- Bid/no-bid triage is harder than takeoff: *"Job selection is the biggest challenge, not job details."*
- Switch gear detection is the highest-leverage feature; mechanical equipment is 99% out of scope.
**Jason Wagner — North Valley Electric (SMB owner/estimator)**
- The "rough-estimate in 10 minutes from $/sq ft" archetype: office TI = $18/sq ft, apartment = $14k/unit, 2BR = $15.5k.
- Custom spreadsheet with formulas (listed price + labor + taxes). Detailed and well-built.
- **The proposal-writing pain quote** (most important sales artifact in this doc): *"Estimating takes 2 hours, but writing the proposal takes 8 hours. Every electrical scope is different."* — the 4x multiplier.
- Tried AI estimating in 2017 (A1 Electrical), gave up. The "I already tried AI" objection has a real history here.
- Value engineering is his edge: spec equivalent cheaper components before the engineer can re-spec. 9.5/10 times the GC copies his contract verbatim.
- Insight on engineers: *"Electrical engineers over-engineer 100%: spec expensive lights ($400) when identical alternatives exist at $100. Engineers are wined and dined by specific manufacturers."*
**Chris Low — JE Dunn (GC) + Rosendin (large EC, $1B+ revenue)**
- Largest-EC voice in the dataset. Different shape from the 11–200 sweet spot — but his feedback shapes the *future* roadmap and explains why we're targeting smaller ECs first.
- Rosendin uses Accubid with 9-month commodity forecasting on copper. *"Estimate is only good for the day it's made."*
- Estimator's #1 pain at this scale: *"What is this project worth?"* — defending the number is hard; most default to $/sq ft.
- *"As an estimator, doesn't look at lighting schedule. Needs to quickly count fixtures (A1, A2), then send counts to supplier for pricing."* — count accuracy >> schedule parsing.
- **Gating requirement for Rosendin adoption:** single-line diagram → feeder schedule conversion. *"Point-to-point on big feeders (from/to, length, type of conduit) = 60% of electrical estimating value. Nobody knows how to translate a single line to plan with 50 panels."* Until that exists, large ECs stay out of reach. Use small/mid ECs to build the wedge.
**Stefan Demeyere — Pleasant Hill Electric (solo residential, Bay Area)**
- Not a target, but useful boundary case. Solo residential electrician, no software, prices over the phone. Confirms the residential exclusion from filters.
- East Bay union work struggling against non-union saturation. Reinforces the Bay-Area de-prioritization.
### Past campaign signals (consolidated, ordered by relevance)
**2026-01-estimator-pm-electrical** (33 leads, 6.1% reply, n=2 positive) — the cleanest historical EC-estimator targeting we've run, now re-labeled to this persona.
- Both positive replies: *"Electrical Estimator"* at small ECs — Foster Electric Corp (CO, replied "Yes") and All Industrial Electric (IL, the *"built an excel file"* replier).
- Estimator titles converted at 22% (2/9); PM titles at 0% (0/8). **Estimator > PM** for this persona, even though PMs lead the cross-campaign average.
- The 30-minute phone-call CTA hurt; switch to open question (validated in 2026-03).
**2026-02-estimator-pm-active** (347 leads, ~9% reply, engagement data missing) — re-labeled to this persona, but with a caveat:
- Audience leaked across trades. Companies in the leads list: Hirschi Masonry, Architectural Wood Design, Custom Wood Services, Nashville Waterproofing, Wayne Brothers. Only a minority were pure ECs. Filter was too broad.
- 40% connection acceptance — the highest acceptance rate we've ever seen, driven by the "Active on LinkedIn" filter. Keep that filter.
- *"Div 26"* insider language in the DM was a credibility signal worth preserving (better when used at actual EC estimators, not the cross-trade audience here).
- 232 of 347 titles were generic *"Project Manager"* — confirms PM-only targeting pulls the wrong people. Estimator-in-title filter is the right cut.
**2026-03-estimating-services** (139 leads, 13.7% reply, 42% positive) — different persona (estimating-services firms), but the DM template is the template:
- The DM#1 formula (founder story + open question + <60 words + company personalization) hit best-in-class reply rate. Port it directly.
- Repliers were heavy-civil, forensic, restoration, woodwork — almost none were pure EC estimators. The audience is meaningfully distinct from this persona. Don't conflate.
- Texas (4 positive) and Ontario (2 positive) over-indexed. Carry into our geo weighting.
**2025-12-electrical-sub-owners-bay-area** (240 leads, 0.8% reply, 0 positive) — wrong role within the right company type.
- Targeted owners of 20–500-employee electrical subs in the Bay Area. 4% connection acceptance, 2 replies (both negative).
- Confirms: **owners don't engage with cold LinkedIn outreach on takeoff tools.** They delegate estimating; the product pitch doesn't resonate with their daily work.
- Geographic constraint cut the pool. Bay Area further skewed toward residential/tech-sector work.
- The retro's recommendation — *"Target estimators and PMs instead of owners"* — is the seed of this persona doc.
**2026-02-lighting-quotations** (119 leads, 5% reply, 1 positive) — adjacent persona, but the strongest single quote is the cross-reference back to this persona:
- FRM Lighting & Controls Quotations Manager replied: *"I don't use that type of software... Might be better on the EC side."* The lighting quotations world *itself* says ECs are the right target.
- Confirms: quotation specialists are adjacent to the pain (they receive RFQs) but they don't do takeoffs themselves. The EC estimator does.
- Holiday timing killed this campaign (December launch). Don't repeat.
### Pre-campaign hypothesis bets
Predictions to verify on the first dedicated EC-estimator campaign. These go in the first retro.
1. **EC estimators out-reply estimating-services.** Estimating-services people work across trades; pure EC estimators feel the EC-specific pain harder. Predicted: 15–18% reply rate (vs 13.7% baseline).
2. **Chief Estimator out-replies Estimator.** Chiefs see throughput pain across a team and have buying authority. Predicted: Chief Estimator 20%+, Estimator 10–12%.
3. **Texas + Ontario + Tennessee out-perform national baseline by 2x.** Same pattern as estimating-services and distributor campaigns.
4. **Bay Area exclusion holds.** 6 campaigns of data agree — no need to re-test.
5. **"Augment your Excel" DM#2 variant beats generic "trial the platform" DM#2.** Worth A/B testing given the Excel-pride signal across multiple calls (Jason, All Industrial Electric).
6. **DM#1 mention of switchgear/panel detection (not just fixtures) lifts reply rate at Chief Estimator title.** Dan McNeillie called gear the most-reliable feature; Chris Low called it the count that matters. Worth testing as a higher-signal opener for senior estimators.
### Updated after retros
*(empty — fill in after first dedicated EC-estimator campaign)*