Every site I build teaches the next one.
WireRef is a free NEC electrical reference with 545+ pages, 6 interactive calculators, and code lookup across all 50 states. It is the fourth site I built and the second reference site after GageRef. Every pattern that works here was tested on the three that came before it.
WireRef did not start from zero. It started from GageRef. GageRef was the first reference site I built: welding electrode specs, amperage charts, rod comparisons, and hydraulic fitting thread data, all sourced from AWS and SAE standards. That project proved the model. Take scattered public data from industry standards, organize it around the questions practitioners actually ask, and structure it for the way search systems retrieve answers. WireRef is GageRef’s architecture scaled up with everything I learned from CheckMyTap and UpOrbit layered on top.
Each project taught me something different. CheckMyTap proved that public data, organized well, could outperform established sites. UpOrbit proved that free tools drive return visits in ways content alone cannot. GageRef proved that the reference-site model specifically works for trade professionals: structured lookups, spec tables, comparison pages. WireRef is the project where I did not have to discover the architecture. I already knew what worked. The question was how far I could push it.
The National Electrical Code is public. The ampacity tables, the GFCI requirements, the conduit fill charts, the state adoption data are all available through NFPA, state licensing boards, and government databases. But it is scattered across PDFs, paywalled code viewers, and state agency websites built for regulators, not for the electrician standing in a crawl space who needs to know what wire to pull. That gap is the same one GageRef fills for welders looking up electrode specs. I recognized the pattern immediately because I had already built the system once.
What Carried Forward
Every project in this portfolio follows the same Get Found, Get Understood, Get Chosen framework. But the execution gets sharper each time because the mistakes are already documented and the wins are already measured. WireRef inherited the most from GageRef because it is the same type of site: structured reference data for a specific trade. CheckMyTap and UpOrbit contributed the layers that GageRef did not have. Each of those projects has its own case study that goes deeper into what worked and what did not.
None of these sites are finished. They are all still being improved, and I am still learning from each one. But the pattern is clear: CheckMyTap proved that normalizing scattered public data into location-level pages creates genuine information gain at scale, though the state-level strategy was an afterthought and the site launched without interactive tools. UpOrbit proved that free tools with no account wall drive daily usage and organic return visits, though the content taxonomy came after the articles instead of before them. GageRef was a turning point. It took the public-data model and applied it to a structured reference format for trade professionals: electrode specs, amperage charts, fitting standards. It proved the reference-site architecture works. WireRef then took GageRef’s proven model, added the localization that CheckMyTap validated, the interactive tools and retention features that UpOrbit validated, and started with the taxonomy defined before writing the first page. It is still a work in progress, but the foundation is stronger because of what came before it.
- The reference-site model: Structured spec lookups organized by the questions practitioners ask on the job. GageRef did this for welders. WireRef does it for electricians. Same architecture, different domain
- Comparison page strategy: GageRef’s rod-to-rod comparisons (E6011 vs E6013, E7018 vs E7024) became WireRef’s wire comparisons (12 AWG vs 10 AWG, copper vs aluminum). The format maps directly
- Answer-first layout: GageRef’s quick-spec boxes at the top of every electrode page became WireRef’s NEC Quick Answer boxes. Same pattern: put the answer before the explanation
- URL taxonomy for reference content: GageRef’s clean URL patterns for specs and comparisons directly informed WireRef’s 14-type URL structure
- Affiliate integration: Product recommendations on tool and material pages with clear disclosure. GageRef tested this approach. WireRef refined it with calculator output tie-ins
- State-level architecture (CheckMyTap): CheckMyTap proved city-level pages work when each has genuinely distinct data. WireRef applied that to 50 state profiles with unique NEC editions, licensing rules, and compliance scores
- Interactive tools as content (UpOrbit): UpOrbit’s daily-use tools proved that interactivity drives return visits. WireRef launched with 6 calculators planned from day one, not added later
- Retention through persistence (UpOrbit): The My Jobs feature borrows UpOrbit’s approach to zero-friction saved state. No login required. Browser-local persistence
- Taxonomy before content (UpOrbit lesson): UpOrbit built 8 categories after dozens of articles existed. WireRef defined 14 content types and URL patterns before the first page was published
- Every page earns its existence (CheckMyTap): Each ampacity page has a unique derating chain. Each state page has a unique NEC edition and unique amendment data. No thin variation pages
How the System Shows Up
The framework is the same across every project. The execution is more precise each time because the mistakes are already documented and the wins are already measured. GageRef proved the reference-site model. WireRef tests how far that model scales when you add localization, interactive tools, and user persistence. This is one of several projects documented under my content strategy and architecture work, where every build compounds the lessons from what came before.
Calculators as Moat
GageRef is a static reference site. You look up an electrode spec, you leave. That is the limitation I wanted to fix. WireRef’s calculators take public NEC formulas and make them interactive, showing every step of the calculation so users can verify the math themselves. The calculations are not proprietary. The presentation and the workflow are.
Voltage Drop Calculator
Enter gauge, length, load, voltage. Returns single-phase and three-phase results with max distance tables. Cites NEC 210.19(A) Note 4 for the 3% and 5% recommendations. The tool that electricians run before pulling wire on long runs.
Panel Load Calculator
Can your panel handle an EV charger? Runs the NEC Article 220 demand calculation, shows remaining capacity, and flags whether a panel upgrade is needed. Answers the question homeowners ask before every major circuit addition.
Ampacity Calculator
Full derating with temperature correction, bundling adjustment, and the 110.14(C) termination limit that most competing calculators skip. Shows the complete chain from base rating to usable ampacity.
Conduit Fill Calculator
How many wires fit? NEC Annex C fill percentages for EMT, RMC, IMC, and PVC. Answers the question that comes up on every commercial pull and prevents the most common conduit sizing mistakes.
Inspection Checklist
Pre-inspection prep for residential rough-in, final, service upgrade, and EV charger installs. Not a calculator but a workflow tool that turns a stressful process into a verifiable sequence of checks.
License Reciprocity Explorer
Which states honor your electrician license? Transfer requirements and fees for all 50 states. Solves a real mobility problem for traveling electricians and apprentices planning their next move.
Anyone can look up NEC Table 310.16. The moat is the four-step derating walkthrough that shows you how to go from 75A base to 65A usable, with the specific inspection failure that happens when people skip step four. The moat is a panel load calculator that shows remaining capacity, not just total load. The moat is an ampacity calculator that includes the termination limit most tools ignore. Each calculator answers a question that competitors answer with a static table or not at all.
State-Level Strategy: The CheckMyTap Evolution
GageRef has no localization. AWS welding standards are national. But NEC electrical code is enforced differently in every state, and that creates a massive content opportunity GageRef’s domain did not have. CheckMyTap proved that local-level pages work when each page has genuinely distinct data. WireRef combined GageRef’s reference architecture with CheckMyTap’s localization strategy and built an entire state-level system from the start instead of adding it later. Every state has a different NEC edition, different licensing requirements, different permit costs, and different amendments. That is not duplication. It is real jurisdictional variation that affects every wire pull in the country.
- NEC edition in force: Which version of the code each state enforces, with effective dates. Not every state is on 2023
- Compliance score: WireRef’s 0-100 scoring system based on code recency, licensing stringency, permit enforcement, and active amendments
- License types and requirements: Master, journeyman, apprentice requirements with links to state licensing boards
- Permit cost ranges: What electricians and homeowners should expect to pay, by state
- State-specific GFCI rules: Older NEC editions mean different GFCI requirements. Each state page shows what actually applies in that jurisdiction
- EV charger permit info: State-specific requirements for the fastest-growing residential circuit type
Neighboring states with different NEC editions create real confusion for electricians who work across state lines. WireRef generates side-by-side comparison pages for every relevant state pair: Maryland vs Virginia, Texas vs Oklahoma, California vs Arizona.
Each comparison surfaces meaningful differences: NEC edition gaps, reciprocity agreements, license transfer fees, and GFCI scope differences. These are not thin content pages. They answer a specific question that a traveling electrician or contractor expanding into a neighboring state actually needs answered.
This is the same information gain approach that made CheckMyTap’s city pages work. The data is real. The variation is real. The page earns its existence because no other page on the internet shows these two states side by side.
On CheckMyTap, choosing a city shows local water data. WireRef borrowed that concept and extended it. The state selector at the top of the site follows users across every page. Pick California once, and every ampacity page, every GFCI guide, every project guide surfaces California-specific code context: which NEC edition applies, whether your kitchen GFCI scope is the 2023 expanded version or the older limited version, and what permit you need. This is the “save my stuff” pattern from UpOrbit applied to professional reference: once you tell the tool something about yourself, it remembers.
My Jobs: Building for Return Visits
GageRef is a lookup tool. You find the electrode spec, you leave. UpOrbit taught me that tools people return to daily create compounding value that single-visit reference cannot match. WireRef’s My Jobs feature is what happens when you take GageRef’s reference utility and add UpOrbit’s retention model. An electrician working on a kitchen remodel does not look up wire sizes once. They reference the same specifications across multiple visits over the duration of the project. My Jobs gives them a persistent workspace.
Users can save specific reference pages, calculator results, and project guides to a persistent job list. No account required. Everything stored locally in the browser, same as UpOrbit’s approach to zero-friction access. An electrician can create a job for “Smith Kitchen Remodel,” save the 20A circuit page, the GFCI kitchen requirements, the panel load calculation, and the inspection checklist to that job, then come back to all of it tomorrow from the same device.
This is the retention moat. A static reference site loses the user after one lookup. A site that remembers what you are working on becomes part of the workflow. The pattern was proven on UpOrbit with daily productivity use. WireRef applies it to project-based professional reference.
14 Content Types, One Taxonomy
GageRef had clean URL patterns but a limited number of content types. UpOrbit’s biggest mistake was building the content taxonomy after dozens of articles were already written. WireRef took GageRef’s URL discipline and applied it to a much larger scope: 14 content types, each with a defined template, a clear URL pattern, and an explicit relationship to other content types. All defined before the first page was published.
Ampacity Pages (97)
Every gauge and kcmil, copper and aluminum, THHN, XHHW-2, NM-B. Each page has a unique derating walkthrough. URL pattern: /ampacity/[gauge]-awg-[type]-[material]/
Wire Sizing by Circuit (26)
Reverse lookup: what wire for 15A, 20A, 30A through 400A. URL pattern: /wire-size/[amp]-amp-circuit/
Project Guides (35)
EV charger, kitchen remodel, 200A upgrade, hot tub, detached garage. Each scoped to a specific project with wire size, breaker, conduit, and code references. URL pattern: /projects/[project-slug]/
Appliance Wiring (62)
Wire size and breaker for 62 specific appliances: kitchen, HVAC, laundry, tools, outdoor. Each page has the exact NEC article, wire gauge, and breaker size for that appliance. URL pattern: /appliance/[appliance-slug]/
State Profiles + Comparisons (115)
50 state pages, 65 state-vs-state comparisons. NEC edition, licensing, permits, compliance scores, EV charger rules, and GFCI scope per jurisdiction.
Wire Comparisons (62)
Copper vs aluminum, gauge vs gauge, Romex vs conduit, GFCI vs AFCI. Each comparison page structured around a specific decision an electrician faces on the job.
Retrospective
I am still actively improving WireRef. The retrospective here reflects where the project is now, not a finished product. But this is the first project where I can point to specific decisions that worked because of previous failures rather than in spite of them. GageRef taught me the reference model. WireRef is where that model met the scale and complexity that exposed its limits.
- Taxonomy first: Defining 14 content types and URL patterns before writing a single page prevented the rework that plagued UpOrbit’s category system
- State architecture from day one: Building 50 state profiles into the URL structure from launch meant no retrofit, no redirects, no orphaned pages
- Calculators as first-class content: Six interactive tools were planned and built alongside reference pages, not bolted on later as an afterthought
- The 110.14(C) differentiator: Most competing ampacity tools skip the termination limit. Including it in every derating walkthrough became the single biggest credibility signal
- Answer-first on every page: The NEC Quick Answer box pattern, inherited from CheckMyTap, puts the usable value at the top before the full explanation
- State data maintenance: 50 state profiles means 50 pages that can go stale when states adopt new NEC editions. I should have built an update notification system from the start
- Comparison page volume: 65 state comparisons is a lot of pages to maintain. Some low-traffic pairs may not justify the ongoing update cost
- Search implementation timing: The search bar was planned early but the results experience needed more iteration before it felt useful at scale
- Affiliate integration depth: Product recommendations are present but could be more systematically tied to specific calculator outputs
What Went Wrong
Fewer fundamental failures this time, but not zero. And the site is still evolving, so there will be more. The pattern of mistakes shifted from “did not plan this” to “planned it but underestimated the maintenance cost.” That feels like progress, but I am not done making mistakes here.
Underestimated State Data Freshness
When a state adopts a new NEC edition, every reference on that state’s page needs updating. Several states changed adoption status between site launch and the first quarterly review, and the pages were stale for longer than they should have been. The data was right when published. It just needed a monitoring system I had not built.
Comparison Pages Created Maintenance Debt
65 state comparison pages seemed like a strong content play. And the high-traffic pairs like California vs Texas genuinely perform. But the long tail of low-traffic pairs like Montana vs Wyoming cost maintenance effort that could have gone elsewhere. Should have started with 20 high-value pairs and expanded based on search demand data.
My Jobs Was Invisible at Launch
The feature existed but the discoverability was poor. It was a footer link that users had no reason to click. Usage was near zero until I added contextual “Save to My Jobs” prompts on project guide and calculator pages. The feature was not the problem. The surface area was.
NEC 2026 Content Was Too Early
I published NEC 2026 preview content before the final version was ratified. Some values changed between the draft and the final. Early indexing of draft content meant Google cached incorrect values for a short period. Should have waited for ratification before publishing, or used a clearer “draft” signal that would have managed user expectations.
CheckMyTap’s failures were about scope and planning: fragmented data sources, retrofit state pages, no tools. UpOrbit’s failures were about execution: broken features, wrong platform, too much complexity at launch. GageRef’s failures were about depth: the reference model worked but lacked the localization and interactive tooling that would have made it stickier. WireRef’s failures are about maintenance and timing: data freshness, content volume management, and publishing too early. The mistakes are getting less fundamental. That is the only pattern I am confident about. I am sure the next project will teach me something I do not know yet.
These failures, along with the ones from CheckMyTap and UpOrbit, are documented in my decision log as part of the same evolving content strategy methodology.
Still building. Still learning.
WireRef is not finished. Neither are the other three. But each project makes the next one better, and the mistakes get smaller each time. GageRef proved the reference model. CheckMyTap proved public data at scale. UpOrbit proved tools drive retention. The case studies document what I am learning along the way.
Content Strategy Portfolio